Global Climate Change and the Coastal Areas of the Río de la Plata

Transcription

1 Global Climate Change and the Coastal Areas of the Río de la Plata A Final Report Submitted to Assessments of Impacts and Adaptations to Climate Change (AIACC), Project No. LA 26

2 (Page intentionally left blank)

3 Global Climate Change and the Coastal Areas of the Río de la Plata A Final Report Submitted to Assessments of Impacts and Adaptations to Climate Change (AIACC), Project No. LA 26 Submitted by Vicente Barros CIMA/Faculty of Sciences, University of Buenos Aires, Buenos Aires, Argentina 2005 Published by The International START Secretariat 2000 Florida Avenue, NW Washington, DC USA

4 Contents About AIACC...page v Summary Project Information page vi Executive Summary page vii 1 Introduction Characterization of Current Climate and Scenarios of Future Climate Change ACTIVITIES CONDUCTED DESCRIPTION OF SCIENTIFIC METHODS AND DATA General overview of the methodology Tides and the river level rise Storm surges Mean wind field The main Plata tributaries: Paraná and Uruguay rivers The greatest discharges of the Paraná River Geology, geomorphology and Delta accretion Topography Hydrodynamic modelling Flood modelling RESULTS Hydrologic scenarios Recurrent flood Maps Permanent flood Relative weight of the forcings of the Plata River level The wind influence in the levels of the Plata River CONCLUSIONS Socio-Economic Features ACTIVITIES CONDUCTED DESCRIPTION OF SCIENTIFIC METHODS AND DATA Delimitation of study area Available demographic information Critical review of the concept of vulnerability SOCIAL VULNERABILITY INDEX: SELECTED DEFINITION AND INDICATORS RESULTS CONCLUSIONS Impacts and Vulnerability ACTIVITIES CONDUCTED DESCRIPTION OF SCIENTIFIC METHODS AND DATA Socio economic vulnerability to recurrent floods Exposure of facilities to recurrent flooding Current and future damage costs Public services infrastructure Building infrastructure Economic quantification of flood damages RESULTS Socio economic vulnerability to recurrent floods...94

5 4.4 CONCLUSIONS Adaptation ACTIVITIES CONDUCTED DESCRIPTION OF SCIENTIFIC METHODS AND DATA La Boca neighbourhood Avellaneda Municipality RESULTS Other vulnerable areas of metropolitan area of Buenos Aires CONCLUSIONS Capacity Building Outcomes and Remaining Needs WORKSHOPS OTHER TRAINING ACTIVITIES SUPPORTED BY THE PROJECT COURSES Course for students Courses on climate change for journalists STUDENTS GENERAL CAPACITY BUILDING ACCOMPLISHMENTS REMAINING CAPACITY NEEDS National Communications, Science-Policy Link Ages and Stakeholder Engagement NATIONAL COMMUNICATION CONTRIBUTION TO UNFCCC ACTIVITIES IPCC NATIONAL POLICIES STAKEHOLDER ENGAGEMENT Outputs of the Project PUBLISHED IN PEER-REVIEWED JOURNALS OTHER OUTPUTS Policy Implications and Future Directions References...122

6 About AIACC Assessments of Impacts and Adaptations to Climate Change (AIACC) enhances capabilities in the developing world for responding to climate change by building scientific and technical capacity, advancing scientific knowledge, and linking scientific and policy communities. These activities are supporting the work of the United Nations Framework Convention on Climate Change (UNFCCC) by adding to the knowledge and expertise that are needed for national communications of parties to the Convention. Twenty-four regional assessments have been conducted under AIACC in Africa, Asia, Latin America and small island states of the Caribbean, Indian and Pacific Oceans. The regional assessments include investigations of climate change risks and adaptation options for agriculture, grazing lands, water resources, ecological systems, biodiversity, coastal settlements, food security, livelihoods, and human health. The regional assessments were executed over the period by multidisciplinary, multiinstitutional regional teams of investigators. The teams, selected through merit review of submitted proposals, were supported by the AIACC project with funding, technical assistance, mentoring and training. The network of AIACC regional teams also assisted each other through collaborations to share methods, data, climate change scenarios and expertise. More than 340 scientists, experts and students from 150 institutions in 50 developing and 12 developed countries participated in the project. The findings, methods and recommendations of the regional assessments are documented in the AIACC Final Reports series, as well as in numerous peer-reviewed and other publications. This report is one report in the series. AIACC, a project of the Global Environment Facility (GEF), is implemented by the United Nations Environment Programme (UNEP) and managed by the Global Change SysTem for Analysis, Research and Training (START) and the Third World Academy of Sciences (TWAS). The project concept and proposal was developed in collaboration with the Intergovernmental Panel on Climate Change (IPCC), which chairs the project steering committee. The primary funding for the project is provided by a grant from the GEF. In addition, AIACC receives funding from the Canadian International Development Agency, the U.S. Agency for International Development, the U.S. Environmental Protection Agency, and the Rockefeller Foundation. The developing country institutions that executed the regional assessments provided substantial in-kind support. For more information about the AIACC project, and to obtain electronic copies of AIACC Final Reports and other AIACC publications, please visit our website at v

7 Summary Project Information Regional Assessment Project Title and AIACC Project No. Abstract Global Climate Change and the Coastal Areas of the Río de la Plata (LA 26) The Argentine coast of the Plata River, including the metropolitan area of Buenos Aires, is subject to flooding when there are strong winds from the southeast, called sudestadas. As sea level rises as a result of global climate change, storm surge floods will become more frequent in this densely populated area. To investigate vulnerabilities of local populations to flooding, and examine potential adaptation options, present and future scenarios of recurrent flooding provoked by storm surges were developed for the Argentine coastal areas of the Plata River. Future scenarios were developed for the decades of 2030 and 2070 using a hydrodynamic model of the River forced by sea level, and surface winds. The model was calibrated and tuned to reproduce the statistics of the River level at the places with tide data and was used to the estimate of the maximum storm tide values along the coast of the Plata River, overcoming the lack of basic data. Sea level rise and meteorological fields to determine surface winds were taken from the IPCC Third Assessment Report. For calculating the reach of floods over land, it was necessary to construct a digital topographic map with enough resolution to describe the flooded areas. Results from the model and the digital model were combined in a geographic information system (GIS) to calculate the spatial reach of flooding at different return periods, both for current and for future scenarios. The areas at risk of flooding during this century in Plata River coasts are very small, but the social and economic impact of the increasing frequency of floods by storm surges will be important. The number of people facing a risk of at least one flood every 100 years would be about 1,700,000 in the 2070 decade, more than three times the present population. By the same decade, those that would suffer floods every year would be 230,000, about six times the population that are currently exposed to annual flooding. Under the assumptions that no adaptation measures would be implemented, the estimate of losses including real estate damages and the incremental operational costs of the coastal facilities for the period would range from 5 to 15 billion USD dollars depending on the speed of the sea level rise and of the rate of growth of the infrastructure. Administering Institution UBATEC SA, University of Buenos Aires, Buenos Aires. Argentina. Participating Stakeholder Institutions Fundación Ciudad, Buenos Aires, Argentina. Countries of Primary Focus Argentina and Uruguay Case Study Areas Metropolitan area of Buenos Aires, La Boca and Avellaneda neighborhoods. Systems and Sectors Studied Coastal zones, human settlements, infrastructure, real estate Groups Studied vi

8 Urban poor Sources of Stress and Change Sea level rise and change in surface level atmospheric circulation Project Funding and In-kind Support Investigators AIACC: US$ 115,000 grant; University of Buenos Aires: US$ 25,000 financial contribution and US$ 100,000 in-kind contribution; University of la República US$ 10,000 in-kind contribution. Principal Investigator: Vicente Barros, CIMA/Faculty of Sciences, University of Buenos Aires, Ciudad Universitaria, 1428 Buenos Aires, Argentina. Other Investigators: Susana Bischoff, Faculty of Sciences, University of Buenos Aires, Argentina; Walter Vargas, Faculty of Sciences, University of Buenos Aires, Argentina; Jorge Codignotto, Faculty of Sciences, University of Buenos Aires, Argentina; Angel Menéndez, Faculty of Engineering, University of Buenos Aires, Argentina; Claudia Natenzon, Faculty of Filosophy, University of Buenos Aires, Argentina; Rubén Caffera, Department of Physics, University of la República, Uruguay; Mario Bidegain, Department of Physics, Uruguay; Roberto Kokot, Faculty of Sciences, University of Buenos Aires, Argentina; Inés Camilloni, CIMA/Faculty of Sciences, University of Buenos Aires, Argentina; Enrique D Onofrio, Department of Hydrography, Argentine Navy; Mónica Fiore, Department of Hydrography, Argentine Navy; Moira Doyle, CIMA/Faculty of Sciences, University of Buenos Aires, Argentina; Gustavo Escobar, Faculty of Sciences, University of Buenos Aires, Argentina; Walter Dragani, Faculty of Sciences, University of Buenos Aires, Argentina; Silvia Romero, Department of Hydrography, Argentine Navy; Elvira Gentile, Faculty of Filosophy, University of Buenos Aires, Argentina; Julieta Barrenechea, Faculty of Filosophy, University of Buenos Aires, Argentina; Sebastián Ludueña, Sub Secretary of Water Resources, Argentina; Sergio Martin, Faculty of Sciences, University of Buenos Aires, Argentina; Ana Micou, Faculty of Filosophy, University of Buenos Aires, Argentina; Ana Murgida, Faculty of Filosophy, University of Buenos Aires, Argentina; Mariano Re, Faculty of Engineering, University of Buenos Aires, Argentina; Ezequiel Marcuzzi, Faculty of Sciences, University of Buenos Aires, Argentina; Victor Kind, Faculty of Engineering, University of Buenos Aires, Argentina; Silvia González, Faculty of Filosophy, University of Buenos Aires, Argentina; and Diego Ríos, Faculty of Filosophy, University of Buenos Aires, Argentina. vii

9 Executive Summary Research problem and objectives The Plata River is a fresh water estuary of unique features. At its source, it has already a width of 50 Km, and broadens up to 90 Km in the section Montevideo Punta Piedras (Fig. 1). The estuary, upstream of this section, is known as the inner Plata River. The front of salinity is downstream near this section, but at the outer part of the estuary the salinity is still lower than in the ocean, though gradually increases up to the Punta del Este Punta Rasa section where it reaches the oceanic values. In this section, considered the outer limit of the Plata River, the width of the estuary is 200 Km. The dimensions and shape of the Plata River together with its very slight slope of the order of 0.01 m / Km favour the propagation from the ocean of astronomic and storm surge tides without discontinuities. Because of the progressive reduction of the estuary s depth and width towards its interior, these tides increase in height as they propagate towards its interior. The strong winds from the southeast drag the waters towards the Plata River and produce very high levels, especially if they are simultaneous with high astronomic tides. These events are locally known as sudestadas and are the cause of floods along the low coasts of the Argentine margin. The typical duration of the flood caused by the sudestadas ranges from a few hours up to 2 or 3 days. These storm surges are higher on the Argentine coast than in the Uruguayan one due to the effect of the Coriolis force, but in addition, because of its lower coasts, the sudestada floods usually only influence the Argentine margin. The more affected areas are the south Samborombón Bay, the low coasts of the southern Great Buenos Aires, and the zones near the outlets of the Riachuelo and of the Reconquista River, as well as the front of the Paraná Delta. The coasts of the Plata River house near 14 million inhabitants, most of which live in the Metropolitan region that includes Buenos Aires city. As the sea level rises as a result of Climate Change, storm surge floods will become more frequent in this densely populated area. The following questions guided most of the Project activity: How many people are currently affected by recurrent floods and how frequent? What are their adaptation capabilities? What is the damage to the infrastructure and to the real-estate property caused by these floods? Considering the likely sea level rise in the twenty first century: How will it change the return periods of floods and consequently how much additional population will be affected? How and where will the social vulnerability to floods worsen and how much will it increase the cost of these events? Finally, it was intended to estimate if there were areas enduring floods, how many people will be affected by them? In brief, the objective of the Project was to assess the social and economic vulnerability to the water level rise in the Argentine coast of the Plata River that will be caused by the Climate Change. Approach Mean and storm surge levels were simulated by a two-dimensional hydrodynamic model with high spatial (2.5 km) and temporal (1 minute) resolution. The model is based on a finite difference implicit alternating direction method. It is forced with the astronomical tide at the southern border, the river discharges at the upstream border and the wind field over the whole domain. The model domain is large enough to include the fetch zone for the storm surges, which are then entirely generated within it. The model was calibrated to astronomical tides and storm surges for the period Then, it was verified that it reproduces the statistical distribution of the water level at the Buenos Aires port. After confidence in its capacity to reproduce the basic features of storm surges was acquired, the tuning of the model allowed estimating the maximum tide values of flood events along the coast of the Plata River, thus overcoming the lack of basic data. For calculating the reach of floods over land, for each level on a point of the coast, the surrounding area that is below this level on the land was assumed flooded. This method does not consider the backwater effect on the tributaries, and thus underestimate the flooded area near the mouth of them. To apply this viii

10 technique, it was necessary to have altitude maps with enough resolution to describe the flooded areas. The pre-existing documentation of land altitudes (maps or data at some locations) was old and outdated to take account of the man made modifications of the terrain during the last 20 years, especially on the Plata River coast. In addition, this documentation had a low vertical resolution that was in general insufficient to cope with the description of the changes that will be caused by mean water level rise scenarios of less than 0.50 m. Thus, it was necessary to develop a new digitized model of the land altitude with 0.25 m vertical and 1 Km horizontal resolution, using field measurements taken with a differential GPS and data from satellite interferometer radars. GPS measurements were taken at places that provided key information according to the geomorphologic maps that were constructed for this purpose. These data was complemented with the information of pre existing lower resolution topographies and with existing altitude measurements at certain points that were taken at request of the Buenos Aires City. The assessment of social vulnerability was made through the integration of the physical and social information in a geographical information system (GIS). This system facilitated the estimate of the affected population, the public service infrastructures and the real-estate damages under different possible scenarios. The social information for current conditions was taken from the 1991 census and the economic one was estimated, in case of the infrastructure of services, by the technical data provided by the companies: Real-estate values were estimated by current commercial values considering nine different zones of different socio economic status. For future scenarios, the socioeconomic conditions for future scenarios were considered constant in time, which is clearly a very strong simplification. On the other hand, the water level conditions were estimated using sea level and climate values of the A2 scenario reported by the IPCC (2001) as forcing entries to the hydrodynamic model of the Plata River estuary. The SRES A2 scenario makes socioeconomic assumptions and GHG emission trends that are relatively similar to the current ones and therefore its sea level rise and climatic scenario are approximately what could be expected if not drastic and rapid reductions of GHG emissions will be made in the next decades. The natural processes of erosion and accretion, as well as the geological conditions of the coast were evaluated, but they are not included in this report. They are considered to be processes of the second order in an environment under strong pressure due to the changes introduced by the direct anthropic action and the rapid increase of the level of the Plata River. The exception is the Paraná delta growth. Information and old maps of the Paraná delta since 1608 to the present were compiled to document the delta advance in the last years. The number of persons affected by floods was estimated from the spatial population distribution and the flooded area corresponding to every return period. Though population information is disaggregated in small areas, social indicators were only available at district level. Therefore, the social structural vulnerability, which is measured from the social economic indicators, was estimated at district scale, which is only an approximation of its spatial distribution. The indicators that are representative of the response capacity to cope with the different stages of the flood emergency were combined to develop a social index of structural vulnerability. These indicators include aspects of demography, quality of life and productive and consumption processes. An index of exposition to floods was calculated from the return period of floods calculated for every cell of 1 Km 2. It was assumed to be approximately the inverse of the return period. Then, an index of social vulnerability to the floods was developed combining this index of exposition to the floods with the social structural vulnerability index through the product of both and a normalization, so the final index ranges from 1 to 100 in the area of the study. The evaluation of the current and future costs includes the real-estate property and the infrastructure of the water supply, the sewage system, the plants of power generation, the highways and the railroads. For each one of these systems, the incremental costs were evaluated as a function of the level rise of the Plata River over its current value. The incremental costs were calculated by the effects of level rise of the river in every installation according to the technical data supplied by the operators of each facility. The damage to the real-estate property of each event was estimated as a percentage of its current realestate value, which includes the direct costs of repair, losses of furniture and costs of depreciation. The ix

11 areas under potential flood threat of present or future flood were classified in 9 different zones according to their current features and real-estate value. The real-estate cost in each zone for a given event was calculated in accordance with the percentage of the flooded area of the zone and with its total real-estate value. In order to estimate the mean annual value of the losses due to floods for the current scenario and for the future ones, the damages of each event to each of the components of the infrastructure and to both public and private real-estate property were added. Then, the losses of each flood-type event were combined with the recurrence of this type of event to estimate the annual average cost Two study cases were conducted to gain insight on the social and institutional responses to recurrent floods during the past and in present times. The selected cases were La Boca neighbourhood in the City of Buenos Aires, and the Avellaneda Municipality, in the Metropolitan Area of Buenos Aires. In both cases, their long tradition in dealing with recurrent floods gave indications on how the population of the Great Buenos Aires may respond when major adaptation will be required. In both cases, the Project conducted interviews with officials in charge of the institutions that deal with the different phases of floods (planning, disaster response, etc) as well as with key informers. Results The mean Plata River level rise will be a few centimetres greater than the sea level rise because of the wind rotation to the east, which incidentally is already taken place. For the same reason, the level rise in the Uruguayan coast will be higher than in the Argentine coast and more important towards the interior of the River. The areas with risk of enduring flooding during this century in Plata River coasts are very small. The southern coast of the Samborombón Bay presents the large area with such risk. In addition in that area, the characteristics of the soil that could be eroded in relatively little time may accelerate the permanent flooding. The growing front of the Paraná delta has also risk of enduring flooding. However, at present is scarcely settled, but may become an area of social vulnerability if it is occupied in the future. If the spatial distribution of the population in the metropolitan area of Buenos Aires does not change too much in this century, then the people to be affected by permanent flooding would be very small. Therefore, it is concluded that assessing climate change risk in the coastal areas of metropolitan region of Buenos Aires is more a matter of dealing with increasing inland reach of the storm surges than with permanent flooding. The areas that are now more exposed to storm surge floods are the coast of the Great Buenos Aires to the south east of the city, part of the district of Tigre and the coast of the Samborombón. According to the A2 SRES scenario, the social vulnerability to floods will become worst in this century along the margins of the Reconquista and Matanzas-Riachuelo rivers and in the south of the Great Buenos Aires in zones relatively far from the coast. In a scenario of 0.4 m sea level rise for the 2070 decade with a modest 1% annual rate increases in the population without considerable changes in its distribution and no new defences built, the population with risk of some flood (recurrence every 100 years) will amount to about 1,700,000, more than three times the present population in such conditions. Those with risk of flood every year will be about 230,000, six times the population that suffer now such recurrence. Under the assumptions that will not be adaptation measures and any change in real estate property, the estimate of losses including real estate damages and the incremental operational costs of the coastal facilities for the period would range from 5 to 15 billion US dollars depending on the speed of the sea level rise. Most of this cost originates in real-estate damage. Of course, these assumptions are unlikely to be fulfilled because it will be some sort of adaptation, but the estimates based on them allow assessing the economic burden of Climate Change in the Metropolitan region of Buenos Aires if not action is taken, from just only one of its impacts, namely the increasing inland reach of the storm surges. Regarding adaptation, in the areas with frequent flooding, the existence of informal alert networks among the neighbours tends to diminish the vulnerability to floods. However, in these areas, the x

12 increasing number of newcomers is reducing the collective cultural adaptation to floods. Other flaw in present adaptation is that defences against floods were designed without considering the future River level rise, what may reduce its efficiency in the future. The institutional responses to floods, although following a similar organization pattern differs from one district to another in its functioning and coordination. In some cases the lack of cooperation between the responsible institutions creates an additional source of vulnerability. In the past the occupation of the small areas of very low lands was avoided. This adaptation strategy is being ignored now, and the current trends of occupation of lands with flood risk, by both very poor settlements and gated communities of upper middle class people are not favouring the collective adaptation to present and future scenarios of recurrent floods. Scientific findings The hydrodynamic model of the estuary was in use to perform an analysis of sensitivity of the Plata River reaction to the changes in the variables that determine its level; that is, the sea level, the direction and intensity of the winds and the stream flows contributed by the main tributaries, that is to say the Paraná and the Uruguay rivers. Though the winds cause the greatest variations in the estuary level by generating important tides and are also the major cause of the seasonal variations throughout the year, the foreseen sea level changes during the twenty first century will be the principal factor of change in the mean level of the estuary. The rotation expected in the mean winds from the northeast towards the east will contribute also to the increase of the mean Plata River level continuing the trend observed in the last 30 years, adding in this century, probably 5 to 10 cm, but this rise will be lower than the contribution of the sea level rise. The discharges from the tributaries, just like in the past, will only contribute a few centimetres in cases of exceptional discharges and only near the delta front. Capacity building outcomes and remaining needs The active participation in the AIACC Workshops and in other training activities contributed to the capacity building of the Project participants. Other relevant activities directed to increase awareness in the society were a course at a Master level in the University of La República, Uruguay in cooperation with the Project Assessing Global Change Impacts, Vulnerability, and Adaptation Strategies for Estuarine Waters of the Rio de la Plata. This cooperation was extended to two courses for journalists in Buenos Aires and Montevideo. The Project supported the thesis of highly qualified students; five of them were either presented or are in the writing phase and were important contributions to the Project. A new and very important experience for most of the investigators of the Project was the work with stakeholder groups. Climate trends in Argentina were significant during the last decades. Thus, it is necessary to start or in some cases improve the current autonomous adaptation and consequently, it is necessary to build additional capacities in the study and development of adaptation to climate change. National Communications, Science-Policy Linkages and Stakeholder Engagement The Project results will be the basic input for the vulnerability study of the coastal area of the region of Buenos Aires this new study as an enabling activity for the second National Communication Argentina to the UNFCCC. This task to start on July 2005 was assigned to the Argentine Co PIs of the Project. The Secretary of Environment and Sustainable Development has developed the Environmental Agenda for Argentina. This planning was based on technical reports and workshops. The climate change issue was addressed in 14 reports of which 2 were on the Plata River coast and were based on the Project results. The project aimed to achieve an inter-consultation relationship with selected stakeholder organizations. This task focused in the evaluation of the project scope, development and results by the stakeholders. xi

13 Institutional actors were consulted both as privileged users of the information produced by the project and, as key informants due to their expertise and involvement in the issue of floods and their management. The methodology applied was the induction of a communicative process, with the finality to go beyond isolated consulting, establishing ruled and continuous mechanisms of association and interchange with stakeholders during the development of the project Policy implications and future directions The results of the Project and their dissemination between the stakeholders provide the basis for regulations of the coastal space favouring the use of activities compatible with recurrent floods. The lack of consideration of climate change in the planning and design of structural works that prevailed until now, it is likely to be abandoned, as the results from the Project become known by the specialists and the stakeholders. It is important that projects like the present one be undertaken to address the climate driven extreme events in the context of the climate change and of other changing factors in other systems and regions of the country. xii

14 1 Introduction The metropolitan area of Buenos Aires including the city of La Plata extends along the coast of the Plata River estuary for about 80 Km. The area is the residence to 12 out of the near 14 million inhabitants that live along both coasts of the estuary. Figure 1.1 shows the Plata River estuary and the main geographical references addressed throughout this report. The dimensions and shape of the Plata River together with its very slight slope of the order of 0.01 m / Km favour the propagation of level changes from the ocean into the River, being them either slow trends or rapid storm surge tides locally known as sudestadas. Thus, the coastal zones of the Plata River are expected to become more frequent flooded, as the sea level rises as a result of global warming. Fig.1.1: The Plata River estuary To address the additional effects of these floods as a result of Climate Change, it is necessary to know their present impacts. Therefore, part of the Project aimed to characterize with a quantitative approach the current effects of these flood events to assess how many people are currently affected in their homes by recurrent floods, in which zones they live, and with which frequency are affected. Other important aspects to characterize the present conditions were the social structure of the people perturbed and the damage to the infrastructure and to the real-estate property caused by these floods. All these impacts will be worsened during the twenty first century, raising several questions. Some of them were addressed in the Project; such as how the return periods of floods will change, how many additional people will be affected, how much the social vulnerability to floods will deteriorate and in which zones, and how much will increase the cost of these events if not additional defences are added. Other aspect was to determine if there will be areas with enduring flood, In brief, the objective of the Project was to assess the social and economic vulnerability to the Plata River level rise that will be caused by the Climate Change in the Argentine coast. The level of the shoreline at the port of Buenos Aires is 0.99 m over the mean sea level. The waters that exceed this value flood over the land. Nevertheless, because of certain step at the beach, in most of the 1

15 coast of the metropolitan area, the alert level is only at 1.80 m. Table 1 shows the time of recurrence for certain flood heights calculated with the record of tides of the last 50 years at the Buenos Aires port. Time of recurrence (years) Height over the mean sea level(m) Table 1.1: Level of the Plata River over mean sea level at the port of Buenos Aires for the indicated times of recurrence (D'Onofrio, Fiore and Romero 1999). The values on Table 1.1 are only valid for the port of Buenos Aires because the maximum of the storm surge tide intensifies as advances towards the interior of the estuary. The time of return of the floods in other localities along the coast of the Plata River cannot be calculated straightforward from long tide records because they do not exist. To overcome this obstacle, it was developed a method that calculate the time of return of extreme water levels along the right margin of the Plata river using a hydrodynamic model of the Plata estuary. This model was previously adjusted to simulate the level of the Plata River under a wide variety of conditions. The development and adjustment of this model was one of the main components of the Project. Future scenarios of mean water level were built forcing the model with the mean sea level taken from scenarios of the IPCC Third Assessment Report and with winds calculated from global climate models selected from the IPCC web page. The River levels corresponding to the different return periods were estimated for each scenario adding to the present value the estimated mean River level rise. The construction of a digital model of the land surface was another important element of the Project. This tool was necessary to calculate the return periods of floods over land, and it was constructed with 0.25 m vertical and 1 Km horizontal resolution, using field measurements taken with a GPS differential system and data from satellite interferometer radar. GPS measurements were taken at places that provided key information according to the geomorphologic maps that were constructed for this purpose. These data was complemented with the information of pre existing lower resolution topographies and with altitude measurements taken by the Buenos Aires City at certain points. Other component of the Project was the assessment of the socioeconomic conditions. For the present time, social data were taken from census and for future scenarios, socio-economic conditions were considered as in present time. Though this a simplistic approach, the history of Argentina during the last century, indicates that unthinkable socioeconomic scenarios had nonetheless taken place and there are few signs that the socioeconomic indicators may improve too much in the future. Demographic growth followed a more predictable path until now and therefore a simple hypothesis of a 1 % annual growth was adopted. The indicators for the social vulnerability index were selected according to its suitability for the construction of the index and their statistical availability. All the information, physical and social was included in a geographical information system (GIS). This system allowed a geographical and quantitative description of present and future scenarios. For each water height on the coast, the area that is below this height on the land was assumed flooded. Then, with the help of the GIS, it was calculated the distribution of two indexes, one of exposure to recurrent floods and other of social vulnerability to recurrent floods. The number of persons affected for every return period of flooding was estimated from the spatial population distribution and the flooded area corresponding to that return period. 2

16 Finally, with the recurrence of floods over land, the GIS facilitated the assessment of the costs resulting from the damage to the real-estate property and to the main components of the infrastructure by recurrent floods in present or future scenarios. Natural processes of erosion and accretion, as well as the geological conditions of the coast were evaluated, but they were considered as second order processes in an environment that is under the pressure of anthropogenic change and of rapid eustatic rise resulting from the sea level change. Other aspects more qualitative were analyzed separately from the GIS, like the strategies of adaptation adopted by the population and their influence on the future vulnerability, as well as the cultural and institutional factors that play an important role in present adaptation. 3

17 2 Characterization of Current Climate and Scenarios of Future Climate Change 2.1 Activities Conducted Since the purpose of the Project was to asses the vulnerability of the coastal areas of the Plata River to climate change, the activities were not oriented to the characterization of the current climate in general, but to those aspects of climate that were relevant to the hydrology of the Plata River and its trends. Therefore, the tides and the river level trends in the last century, the surge storms and its relation to the recurrent floods of the Argentine coast were analyzed. Since the wind field affects the River level, its trends were analyzed for both current and future scenarios. The other forcing of the Plata River is the discharge of the tributaries that only have a minor impact when attain extreme values. Thus, the extreme discharges of the two main tributaries were also studied. A new digital model of the topography with the required resolution to assess the flooded areas under different scenarios was developed. To construct this model, it was necessary to have geomorphology maps and as a base for them of the geology. Therefore maps of both aspects were also constructed. It was developed a two-dimensional hydrodynamic model to simulate the water level of the Plata River. The mode that was calibrated for astronomic and storm tides, and it was validated for both, its main fields and its statistical distribution of levels. The model was used to develop present and future scenarios of both recurrent and enduring floods. Other use of the hydrodynamic model was to assess the relative weight of the different forcings of the Plata River level, namely to the tributary discharges, wind and sea level. Finally, combining results of the model with wind trends on the region as well as level trends at Buenos Aires and Montevideo it was understood the importance of wind in the seasonal and spatial differences as well as in the different trends of the Plata River levels. 2.2 Description of Scientific Methods and Data This description is divided in many parts because of the complexity of the system under study. However, to facility the general understanding, the first section describes what can be considered the general methodology of the Project regarding the characterization of the present and future scenarios of floods. Then, the scientific methods and data of different and particular aspects are described in the following nine sections General overview of the methodology The integration of physical and social information in a geographical information system (GIS) facilitated the assessment of the geographic distribution of the population that would be affected by floods, as well as the public and services infrastructure and the real-estate damages under different possible scenarios. The methodology, used to generate the required physical information, is addressed in the next paragraphs. The integration with social information to assess potential damages and vulnerability to floods is discussed in chapter 4. Scenarios of the Plata river level were developed using sea level and climate scenarios SRES A2 of the IPCC Third Assessment Report (IPCC 2001) as input to a hydrodynamic model of the estuary of the RP. Mean and the extreme levels at the Argentine coast of the Plata River were simulated by a hydrodynamic model. The model is forced by astronomic tides in its south boundary, the discharges of the main PR tributaries near its northwester boundary and the surface wind field over its entire domain. This domain is big enough to simulate the generation and development of storm tides produced by the dragging effect of winds inside it. The spatial resolution of the model is 2.5 Km and the time step is of one minute. The 4

18 model was calibrated according to both the astronomic and storm tides for the period and it was verified that simulates adequately the statistical distribution of water levels at the Buenos Aires coast. Once the model achieved the ability to reproduce the basic features of storm waves along the coast, its final adjustment allowed the estimate of the maximum value of them all along the coast of the Plata River so overcoming the lack of basic information, which was available only in a few places along the coast. The relationship between winds and water levels is non linear, and therefore, it is not possible to obtain the mean water level field directly from the mean wind field. Hence, future scenarios of the mean level rise were developed forcing the model with the mean level of the sea according with the scenarios of the IPCC (2001) and with daily winds calculated from climatic scenarios taken from the same Report. For future scenarios, the flood levels corresponding to each return period were estimated adding the estimated mean increase of the level in that scenario to current levels. The spatial distribution of flooding over land was estimated using the water levels and a digital model of the surface altitude. The resolution of this model is 0.25 m in altitude and 1 Km in the horizontal scale. It was built combining measurements taken with a geo position differential satellite system (GPS) and satellite radar data. The measurements with GPS were taken in key places according to geo morphologic maps that were made for this purpose, section This information was completed with data from preexisting low resolution maps (Geographical Military Institute) and some previous existing measurements taken by the City of Buenos Aires at certain sites. The four sources of data have limitations that are discussed in section Nevertheless, they complemented each other and allowed to develop an acceptable digital model of the surface according with the purpose of this study. The digital model includes the Argentine coast of the Plata River from the Paraná Delta to Punta Rasa and extends within the continent up to the altitude of 5 m over the mean sea level. For each water level at a given coast site, it is assumed that the land of the surrounding area under this level will be flooded. This approach, neither considers the defended areas, which are relatively very small, nor the backwater effect in the tributaries that can produce floods over these levels. In the first case, errors in the region totals, either in the flooded surface, population involved or real estate damage, are not important. However, errors by underestimation in the second case are larger. Thus, the regional totals could result in some underestimation Tides and the river level rise Because of the shape and distribution of the depth contours that run parallel to the coastline, the Argentine Sea is open to the great surrounding ocean basin in which astronomical tidal waves propagate from SE to NW (Balay 1961). These waves refract on the continental slope, undergoing, from there on, all kind of transformations due to the progressive water shallowing, the meteorological action, and the Earth rotation effects. Because of these facts, tides in the Argentine coasts have very different forms and amplitudes. A way of classifying tides is the method suggested by Courtier (Defant 1961). This method characterizes four main forms of the tide, based on the result of dividing the sum of the amplitudes of diurnal waves K 1 and O 1 by the sum of the semidiurnal M 2 and S 2, the major constituents in each group. F = K1 + O1 M + S 2 2 The result (F) will be a number smaller than 1 if semidiurnal constituents predominate and greater than 1 when diurnal constituents predominate. If 0.25! F < 1.5, the tide is mixed and predominately semidiurnal. In most of the cases there are two high tides and two low tides per day, with strong diurnal inequalities, although occasionally there can be only one high tide and one low tide per day (Balay 1961). The latter takes place at maximum moon declination. This form can be observed in the Plata River, where the tidal component M2 represents more than 65% of the tide wave energy (D`Onofrio et al. 1999). From co tidal analysis it is known that wave tide propagation along the Argentine coastline up to the Plata River takes approximately 26 hours at a velocity of about 200 km/h, while it takes about 12 hours to 5

19 travel along the Plata River at a mean velocity of 30 km/h. The slow down at the Plata estuary is due to the smaller mean depth in the area of only about 5m. There are two amphydromic points, one at approximately 41 S and the other one at 47 S. From the equal amplitude charts it can be seen that the amplitude of M2 varies from about 4 m in the extreme south of Argentine to less than 0.30 m in the inner Plata River. Thus, the Plata River can be characterized by a micro tidal regime of a few centimetres of amplitude and then, the meteorological components; especially the wind regime acquires a decisive relevance in the River dynamics. The wind over the surface of the water influences its level, the vertical mixing and the wave tide velocity. The Plata River is under the influence of the South Atlantic subtropical High. Consequently, wind directions depend basically on the position of this pressure system prevailing from the northeast all over the year. In winter, the shift to the north of this system increases the frequency of winds from the west, while in summer, when it moves southwards, there are more frequent winds from the east and southeast. The average wind intensity in the region is fairly uniform, about 5 m/s. Sea level rise has been observed in most of the coasts of the planet during the last decades. Because of that, the level trends in the Plata River were explored. At the coast of Buenos Aires, there are hourly data recorded since Data were referred to a level called the zero of the Riachuelo that is about a meter below the mean maximum water level, Fig 2.1. There are periodic low-frequency astronomic contributions to tides, between 8 and 19 years, which might mask any trend in the scale of 50 years or less (Godin 1972). To attenuate these contributions a low pass filter was used. This filter was designed starting from the Kaiser Bessel window (Hamming, 1977) following Harris s (1978) technique. After that filtering, annual mean water levels were calculated as the arithmetic average of hourly tide levels. Figure 2.2 shows the series of annual mean levels, the filtered signal and the calculated linear regression for the period of the filtered series. The trend obtained was 1.7 ± 0.1 mm/year with a correlation coefficient of The rise in the twentieth century was about 17 cm, approximately 50 % of which occurred in the last 3 decades. 6

20 Fig.2.1: Relation between topographic (IGM) and tidal references in the Rio de la Plata 7

21 Fig.2.2: Annual mean water levels, filtered series and linear regression calculated over the latter, Buenos Aires Figure 2.3 shows the mean water levels corresponding to the trimesters, per decade, displaying the adjusted straight lines by least squares. Slopes are in mm/10 years. It can be seen that the highest mean levels correspond to the summer trimester, while the winter trimester has the lowest mean water levels. These differences in water level are mainly due to meteorological contributions, but can be minor effects from long-term tide constituents such as Sa (solar annual) and Ssa (solar semiannual) and variations in water density. The comparison with Montevideo, which is at the outer part of the estuary, and for this reason less affected by the mean wind variations indicates that the wind effect is important, Fig This aspect will be discussed in section , and a full discussion of the spatial and seasonal influence of wind on the Plata River level will be addressed in section Although, trends in Montevideo are qualitative similar to those of Buenos Aires, they are lower and the winter trend has a slightly lower slope than other seasons. The differences are larger in the last three decades, when the level at Buenos Aires augmented 12 cm, while at Montevideo increased only 5 cm during the same period. As will be seen in section 2.2.4, water level trends at the Rio de la Plata estuary are consistent with the low-level circulation trend and constitute additional evidence on the southward shift of the regional circulation during the last decades. 8

22 Fig. 2.3: Seasonal mean water levels per decade and corresponding linear regressions, Buenos Aires Fig 2.4: As in fig.2.3, but for Montevideo 9

23 2.2.3 Storm surges Major storm surges are the cause for floods in the Argentine coast of the Plata River. Some low areas of the city of Buenos Aires and its surroundings are affected by such events, associated with strong southeasterly winds over the Plata River estuary, and for that reason locally known as sudestadas. This phenomenon is normally accompanied with persistent, though in general not heavy rainfalls. The higher level registered at the coast of Buenos Aires was 3.90 m over the mean sea level in April 15, The increasing trend in the mean sea level in the context of a global warming process may lead to a rise of about 1 m during the present century. Under such conditions all the area of the metropolitan Buenos Aires that is below 5 m over the mean sea level would be potentially threaten by extraordinary storms. This area includes not only the coastal fringe of the Plata River, but also the populated margins of the Matanza-Riachuelo and Reconquista rivers. The first storm flood in this area that has been reliably recorded took place on the 5 th and 6 th of June, On that occasion strong southeast winds caused a river level rise that seriously affected the coastal area, and caused several ships to sink in the port of Buenos Aires. During the nineteenth century, many different intensity southeaster winds continued to produce considerable damage and even killed people. Unfortunately for these cases, there are no water level records available, as systematic measurements of the River levels referred to land benchmarks had not been initiated yet. Meteorological data used for the storm surge analysis was taken from the NCEP/NCAR reanalysis (Kalnay et al., 1996) for 1951/2000 in the domain between 20º S and 60º S and 80º W and 40º W. The hydrological data (1951/2000) was provided by the Naval Hydrographic Institute. There was a gap of missing data during 1963 and Storm surge levels were obtained from the difference between observed hourly levels and their corresponding astronomic tide predicted levels. Considering the length of the records and the possible modifications of astronomical tide by changes in the coasts and in the depth of the estuary, harmonic analyses were performed for periods of 19 years, using a least square method (Foreman, 1977, 1978). Estimated harmonic constants were calculated for its use in the tide predictions of each period. This allowed minimizing possible trends due to tide amplitude variations caused by changes in the morphology of the area between 1950 and In order to evaluate the quality of the obtained remainders, yearly spectral analyses were performed. From the spectra analysis, it appears that the energy present in the semidiurnal and diurnal tide bands is two orders of magnitude lower than the energy of the astronomical tide, what guarantees that storm surges were correctly separated from observed levels. The maximum value of the two components that define the level of the water, astronomical tide and storm surge, does not necessarily occur simultaneously. Therefore, since the warning level of the river is 2.50 m (Balay 1961) and the astronomical mean maximum tide is approximately 0.90 m (D Onofrio et al 1999), it was adopted as a criterion for defining the meteorological tide, the level of 1.60 m persisting for at least 24 hours. With this criterion an important percentage of the cases actually exceed the 2.5 m mark. There were 297 cases in the 50-year period that satisfied both height and duration thresholds. Hereafter, the term sudestadas will be restricted to these cases. 10

24 Fig.2.5: Annual distribution of sudestadas in the Plata River Sudestadas occur during the whole year, the least frequently in winter, Fig 2.5. However, the ones that do occur in winter have an intense and considerably developed low-pressure system in the north-eastern of Argentina or Uruguay characteristic of cyclogenesis, and produce, in the average, higher peak levels than the others. Fig.2.6: Distribution of maximum storm tide heights over the threshold of 1.60m The storm surge height distribution has a decreasing exponential shape, Fig 2.6. Storm surges with peak height higher than 3 m have near to 1 % probability. Although this percentage implies only 3 cases in 50 years, this level must however be taken as guidance for protection management. 11

25 Fig.2.7: Distribution of the sudestada durations Figure 2.7 shows the empirical distribution of the sudestada duration. It should be considered that according to the sudestada definition used here, their duration has to be longer than 24 hours. The mean value was 47 hours, the maximum 175 hours and the standard deviation 22 hours. Like in the previous case, the distribution has a decreasing exponential shape. Sudestadas ongoing for more than 60 hours are only reached with a probability lower than 20 %. Thus, approximately 80 % of all the sudestada events registered in this period persisted less than two days and a half (20 to 60 hours). This is the most likely range considered when estimating the costs associated with the risks of sudestadas in the Plata and therefore, although their impact can be acute, is short in time, thus moderating its social and economic adverse effects. The three principal components of atmosphere circulation field at 1000 hpa accompanying the sudestadas explain 75 % of the variance. These patterns show that sudestadas are associated with, either a combination of high pressure system to the south of the RP and a relative low pressure to the north (first and second PC modes not shown) or a very deep low pressure area to the north of the Plata River (third PC mode not shown). These three cases are associated with south-eastern winds, which produce important tide waves on the Plata River. 12

26 Fig. 2.8: Decadal distribution of sudestadas in the Rio de la Plata Almost all the sudestadas associated with the third PC occur in winter and they reach in the average greater peak levels than the others. These cases are associated with an intense low-pressure system north of the RP, which are typically due to the frequent cyclogenesis in that region. In many of these cases, intense precipitation occurs in Buenos Aires. The mean frequency of sudestadas occurrence for the period was about 6 events/year. Figure 2.8 shows the mean decadal frequency of sudestadas for the last five decades. During the period, there was a reduction in the absolute frequency respect to However, in the following decades, there was a positive trend in this frequency going from 44 cases in to 79 cases in The beginning of this trend at the 1970 decade was simultaneous with other climate and hydrological trends in the region Mean wind field Until about 1980, when satellite information became abundant, meteorological data over the South Atlantic heavily depended on opportunity observations made by merchant ships. This data were unfortunately very sparse, both in space and time and only restricted to the surface level. In the case of wind observations, the density and quality of observation were considerably worse than in pressure, but since sea level pressure (SLP) is strongly coupled with wind, it may be considered representative of the atmosphere low-level circulation. SLP analysis presents another advantage; spatial low frequency dominates the time-averaged features of SLP fields, and thus, data from continental synoptic stations that have been more systematically observed reinforce the credibility of SLP fields off the coast of South America. In addition, the ocean near the South American coast has had a better coverage of merchant ships than the middle of the South Atlantic Ocean. Therefore, for all these reasons, the study of a possible trend in the low-level atmospheric circulation of the western border of the South Atlantic high (WBSAH) during the second part of the last century was made analyzing SLP. SLP monthly means of the period were taken from the reanalysis of the National Centre for Environmental Prediction (NCEP). These reanalyses are available in 2.5 latitude by 2.5 longitude grid. The domain analysed was between 25 S and 45 S and 65 W y 45 W. Seasonal averages were calculated from monthly means as follows: December, January and February for summer, and so on for the rest of the seasons. NCEP reanalyses were run with a frozen model and a database that includes conventional surface observations and satellite observations and constitutes one of the most consistent atmospheric global data set (Kalnay et al. 1996) The shift of the annual mean field Decadal averages of NCEP reanalysis of SLP and of surface winds indicate a shift to the south of the pressure field over the area covered by the WBSAH. Fig. 2.9 illustrates this displacement, showing the 1951/1960 and the 1990/2000 mean fields, as well as their difference field a) b) 13

27 c) Fig. 2.9: Annual mean fields of sea level pressure and wind: a) , b) , c) difference between and The SLP difference field, Fig 2.9c, implies that between latitudes 33 S and 40 S, this change enhanced (reduced) the eastern (western) component of the mean surface wind and indicates a southward shift of the anticyclonic circulation. This shift resulted in increasing wind drag on the surface water that, because of the shape of the Rio de la Plata estuary, could have caused a water level rise in its inner part (Simionato et al 2003) Seasonal and interannual variability In order to express in a synthetic form the change of the SLP annual cycle during the period, it was performed a principal component analysis (PCA) on the matrix data composed of the SLP seasonal means of each year. Since for decades or even centuries, the seasonal variability (annual cycle) is expected to overweight the interannual variability, the first PCs are expected to represent roughly the seasonal variability while their loading factors might provide information on the interannual variability of the annual cycle. Following the precedent idea, a rotated PCA in T mode, with the correlation matrix as input allowed the identification of two SLP patterns that jointly explain 91.2 % of the variance. The other modes explain individually very small percentages of variance, i.e. less than 4 % each. The PC1 is characterized by a strong meridional gradient south of 35 S, typical of the mean westerly flow, Fig. 2.10a. North of this latitude, there is a pattern that resembles the WBSAH in the east, and the Chaco low in the west. The PC2 field shows an intensified circulation with respect to the PC1 pattern, both in the WBSAH and its ridge at 37 S and in the Chaco low centre to the west of 60 W, Fig. 2.10b. PCA 1 represents quite well the broad features of the mean winter SLP field over the region, while PCA 2 does the same with the mean summer SLP field (NCAR/NCEP reanalysis). Even without examining the SLP seasonal fields, the inspection of factor loading (FL) values, with FL1 close to one in winter and similarly with FL2 in summer, permits to regard them as the respective seasonal patterns. In winter, the first PC amounted 20.7 % of the total variance out of the 25.2 % explained by all the modes in that season, while the second PC only explains 2.9%, Table 2.9. This implies that the PC1 explains more than 80 % of the winter variance, while the PC2 does only 11.5 %. Similarly, the PC2 explains 73 % of the summer variance. In spring, as in summer the PC2 is the dominant pattern, while in autumn is the PC1. However, in both cases the dominant PC explains less variance than in winter or summer. From now on, these two first PCs patterns will be referred to as the winter and the summer modes. 14

28 a) b) -25 PC1 - NCEP % -25 PC2 - NCEP % Fig. 2.10: Principal components (PC) of annual mean sea level pressure a) PC1 b) PC2 MODE 1 MODE 2 TOTAL TOTAL SUMMER AUTUMN WINTER SPRING Table 2.1: Explained variance of the first two modes of the SLP The annual average series of the FLs corresponding to the winter and summer modes are consistently positive, and the winter FL has a negative trend while the summer one has a positive one. In both cases, these trends started at the early 70 s consistent with the observed trends of the SLP at the South American coast, where the maximum pressure along the coast according with the reanalysis shifted about 1.2 of latitude southward since the early 70s. This trend is significant at a 0.05 confidence level. This implies a growing predominance of the summer surface circulation type at expenses of the winter one, and therefore an intensification of the WBSAH circulation and its shift to the south GCM verification 15

29 The region surrounding the Plata River is the area of the present study, Fig. 2.11a. The skill of the GCMs to simulate the observed features of the mean SLP over this region was checked against the NCEP correlation coefficient HADCM3 CSIRO CCCMA NCAR ECHAM4 GFDL J F M A M J J A S O N D Fig. 2.11: Monthly spatial correlation coefficients between sea level pressure from the NCEP reanalyzes and six GCMs reanalyzes. Monthly spatial correlation coefficients were calculated between the GCM outputs and the NCEP reanalyzes. Most of the models had a poor correlation with the reanalysis data during the austral winter months of July, August and September (the only exception was the ECHAM4/OPYC3 model) and high correlation during the austral summer and autumn, Fig. 2.11b. Table 2.2 shows the four models that have the best correlation with NCEP reanalysis during the winter months in the selected area. From now on, we will restrict our analysis to these models. Model Institution Period HADCM3 Hadley Centre for Climate Prediction and Research CSIRO-Mk2 Australia's Commonwealth Scientific and Industrial Research Organization GFDL-R30 Geophysical Fluid Dynamics Laboratory ECHAM4/OPYC3 Max Planck Institute für Meteorologie Table 2.2: Global Climate Models considered in the analysis -25 NCEP -25 HADCM

30 NCEP - HADCM Fig. 2.12: Annual mean sea level pressure (hpa) for derived from the NCEP reanalyzes (right top), HADCM3 (left top) and difference between NCEP reanalyzes and HADCM3 (bottom) Table 2.2 also indicates the GCM experiment periods, which starts in the past and extends into the future forced by GHG concentrations resulted from the socio-economic SRES-A2 scenario. The fact that the other models do not simulate well the SLP of this region during winter months does not mean that they could not do better in other regions. As an example of how these models simulate the observed mean SLP field, Fig presents the NCEP and the HADCM3 model fields for the period, as well as their difference Seasonal variability and trends in the CGM experiments A further verification of the GCM experiments capabilities to simulate the regional SLP fields was made analyzing their SLP annual cycle modes. In addition, trends in these modes can also give some insight in 17

31 the nature of the observed trend discussed in section The ECHAM4/OPYC3 was not included in this analysis because its outputs were available only since The same PC technique as the one explained in section 4.8 was applied to the SLP field of each model. With some differences, the first two PCs of the three models reproduce the basic features of the respective NCEP reanalysis modes, Fig In the case of HADCM3 experiment, the order of the first two PC was permuted, but this is not important as the difference of explained variance between the first two modes is minimal as in the NCEP reanalysis, Table 4.9 and Table MODE 1 MODE 2 TOTAL HADCM3 CSIRO Mk2 GFDL R30 HADCM3 CSIRO Mk2 GFDL R30 HADCM3 CSIRO Mk2 GFDL R30 TOTAL SUMMER AUTUMN WINTER SPRING Table 2.3: Explained variance of the first two modes of the SLP for three GCM experiments On the contrary, in the case of the other two model experiments, the PC1 explained variance increases with respect to NCEP reanalysis at PC2 expense. To avoid confusion, hereinafter, we will refer to these PC patterns as mode 1 (winter mode) and mode 2 (summer mode). Table 2.3 shows the linear correlation coefficient between the respective spatial modes and the NCEP reanalysis modes. These correlation coefficients confirm the visual impression that comes from figures 2.10 and 2.13 that these modes are practically the same. 18

32 a) b) -25 PC2 - HADCM3-42.8% -25 PC1 - HADCM3-47.1% PC1 - CSIRO Mk2-52.3% PC2 - CSIRO Mk2-43.9% PC1 - GFDL R % PC2 - GFDL R %

33 c) d) HADCM3 CSIRO GFDL HADCM3 CSIRO GFDL Fig 2.13: Principal components (PC) of annual mean sea level pressure for three GCM experiments, a) winter mode, b) summer mode and their respective factor loadings c) and d) As in the NCEP reanalysis, the first two modes explain near 90 % of the variance, 89.9 % in the HADCM3, 86.2 % in the CSIRO-Mk2 and 88.5 in the GFDL-R30 experiments. These modes are even more clearly identified with winter (mode 1) and summer (mode 2) circulation as they explain only a minimal part of the explained variance of the opposed season, Table 4.9. It can be concluded that the three experiments simulate the more distinctive features of the SLP seasonal variability of the region studied. Moreover, they also reproduce the general trend of the SLP seasonal variability, as their first two FLs present trends similar in sign to those of the NCEP reanalysis Future scenarios The same PC technique used in section 2.10 and 2.12 was applied to the four MCG SLP field series during the period described in the right column of table Again, the first two modes were very similar to the respective modes corresponding to the past period of both the GCM and of NCEP, Fig and table As in the case of the NCEP reanalysis and of the MCG simulated fields of the last part of the twentieth century, the first two modes account for about 90 % of the variance with the exception of the ECHAM4/OPYC3 experiment in which these two modes account for nearly 80 %. In addition, they have similar partition of variance in seasons as before, Table It seems that the SLP changes in the future scenarios are not so important to substantially alter these patterns and their seasonal variability. The four models maintain the observed trends of the summer and winter modes into the future. The right column of figure 2.14 shows the mean explained variance of these two modes during each decade, as well as their sum. 20

34 Period Model or Reanaly sis NCEP HADC M3 Mode 1 CSIRO- Mk2 GFDL- R30 HADC M3 CSIRO- Mk2 GFDL- R30 ECHA M4/OP YC3 NCEP HADC M3 CSIRO- Mk GFDL- R30 HADC M3 CSIRO- Mk2 GFDL- R30 ECHAM 4/ OPYC Mode 2 Table 2.4: Correlation matrix between mode 1 and 2 of SLP in the NCEP reanalysis and in four GCM experiments 21

35 PC2 - HADCM3-40.4% PC1 - HADCM3-49.8% PC1 - CSIRO Mk2-49.4% PC1 - ECHAM4-41.4% PC2 - ECHAM4-38.3% PC2 -GFDL R % PC1 - GFDL R % PC2 - CSIRO Mk2-46.5% Fig 2.14: Principal components (PC) of annual mean sea level pressure for four GCM experiments, left) winter mode, centre) summer mode and their respective factor loadings 22

36 The MCG experiments present some differences on the initial partition of the percentage of variance between the summer and winter mode, but all four shows the same kind of evolution, with the summer mode growing at expense of the winter mode According to the SRES A2 scenarios of the four models, future changes will be more important in spring and autumn than in winter and summer, Table 2.5. However, it cannot be ruled out, some changes in the solstice seasons because, model experiments overestimate, with respect to the NCEP reanalysis, the explained variance of mode 1 in winter and mode 2 in summer. Nevertheless, changes in the transition seasons indicate a trend toward a longer period with prevailing summer low-level atmospheric circulation and shorter periods with dominant winter lowlevel atmospheric flow. The growth of the summer mode at expense of the winter mode implies both, an intensification of the anticyclonic circulation and a shift to the south of the axis of maximum pressure. The average shift for the four experiments is about 2 of latitude in 150 years. This is a considerable displacement when compared with the seasonal shift of about 10 of latitude between the extreme months of January and July. If this climate scenario will become real, this trend may produce important climate changes in subtropical South America. Presumably, these changes could have already started because the observed shift was almost simultaneous with the positive tendency in precipitation initiated during the sixties (Barros et al 2000). Other change, probably related to the observed shift in the SLP pattern, was the positive trend in autumn temperatures (Bejarán and Barros, 1998). The trend toward increasing (decreasing) predominance of the summer (winter) mode has a direct implication on the Plata estuary wind field. The SLP meridional gradient is proportional to the eastern wind component (geostrophic relationship) or more realistically because of friction effect to the southeasterly wind component. This wind component is directed to the Plata River mouth and. as explained before, because of the shape and very small slope of the River, contributes to rise its level in its inner stretch. Thus, it is possible that the level of the Plata River in its inner stretch has increased not only due to the sea level rise, but also because of the rotation of the wind field. In addition, according to the model results, it is likely that this effect will continue in the twenty first century. MODE 1 HADCM3 MODE 2 HADCM TOTAL TOTAL SUMMER SUMMER AUTUMN AUTUMN WINTER WINTER SPRING SPRING MODE 1 CSIRO Mk2 MODE 2 CSIRO Mk TOTAL TOTAL SUMMER SUMMER AUTUMN AUTUMN WINTER WINTER SPRING SPRING

37 MODE 1 GFDL R30 MODE 2 GFDL R TOTAL TOTAL SUMMER SUMMER AUTUMN AUTUMN WINTER WINTER SPRING SPRING MODE 1 ECHAM4 MODE 2 ECHAM TOTAL TOTAL SUMMER SUMMER AUTUMN AUTUMN WINTER WINTER SPRING SPRING Table 2.5: Explained variance of the first two modes of the SLP for four GCM experiments and for different periods The main Plata tributaries: Paraná and Uruguay rivers Most of eastern subtropical South America, about 3.1 x 10 6 km 2, constitutes the Plata Basin (Fig. 2.15). The two main tributaries are, by far, the Paraná and Uruguay rivers contributing with the 97% of the discharge into the Plata River. In turn, the Paraná has an important tributary, the Paraguay River. 24

38 Fig. 2.15: The Plata Basin and its main sub basins The main streamflow of the Plata River is about 23,000 m 3 /s. In this section, only the issue of the greatest discharges, which may cause some rise in the Plata level (see section 4.2.4) will be addressed. For this reason, it was studied the greatest discharges of the main tributaries of the Plata River, the Paraná and the Uruguay rivers and its climatic forcings. The following is the resume of the main results The greatest discharges of the Paraná River Without considering the Paraguay basin, the Paraná River basin covers about half the area of the La Plata basin. It is usually divided in three sub basins, the Upper Paraná (upstream of the junction with the Grande River), the Middle Paraná (between the junctions with the Grande and the Paraguay rivers) and the Lower Paraná (downstream from Corrientes) basins, Fig Most of the Paraná River streamflow comes from the upper and middle courses, having a relatively small contribution in its lower section. The high streamflows in the Middle Paraná causes flood over large areas of the Lower Paraná even without a significant local contribution in this sub basin. Due to the large size of the Paraná basin, its big discharges and floods persist for months, and they are not caused by single synoptic events. The greatest monthly-averaged discharge anomalies of the twentieth century at Corrientes (the outlet of the Middle Paraná) calculated with respect to the monthly means are shown in Table2.6. These discharges are considerably larger than any possible impact resulting from water management by the upstream dams since in the top ten peaks, anomalies more than doubled the mean annual discharge of the river, 18,000 m 3 /s. The table includes a classification of the events according to the season and the phase of El Niño-Southern Oscillation (ENSO), and the contribution of each of the sub-basins to these major discharge events, (Camilloni and Barros 2003). With few exceptions, the major discharge events in the Lower Paraná originate in the Middle Paraná basin. The only cases with important contribution from the Upper Paraná occurred during the extraordinary El Niño and a few months after its end. The contribution of the Paraguay River to the major discharges in Lower Paraná, although relatively lower than the contribution of the Middle Paraná, it always adds up to this one. 25

39 Corrientes Date and ENSO phase Upper Paraná Middle Paraná contribution Paraguay Jun 1983 Autumn (+) Jun 1992 Autumn (+) D Dec 1982 Summer (0) Mar 1983 Autumn (+) Jun 1905 Autumn (+) May 1998 Autumn (+) Oct 1998 neutral Spring Oct 1983 Spring neutral Jul 1982 Winter (0) Feb 1997 Summer neutral Sep 1989 neutral Spring N/A N/A N/A Table 2.6: Major discharge anomalies (m 3 /s) at Corrientes and the corresponding ones at the Upper Paraná. Paraguay and the Middle Paraná. (0) and (+) stands for El Niño periods as follows: (0) for the onset year of El Niño and (+) for the following year. N/A means no data. The middle Paraná basin is at the midpoint of the dipole structure of precipitation associated to the South Atlantic convergence zone (SACZ). Thus, the major discharges of the Paraná River are not associated to the intensification of any of the phases associated to this dipole; rather, they are associated to another forcings. In fact, according to table 2.13, El Niño (EN) is the most important forcing, although not the only one. The six greatest peaks occurred during EN events, five of them in the autumn of the year following the beginning of the events, autumn (+), (Camilloni and Barros 2003). Figure 2.16 shows the composite of the precipitation anomaly during the autumn (+) of El Niño events. The magnitude of the anomaly, centered at this basin, almost doubled the mean rainfall in this part of the Middle Paraná basin. In these cases, the warm anomalies in the tropical Pacific Ocean forces an atmospheric circulation that favours in the upper troposphere the advection of cyclonic vorticity and at low levels the advection of heat and moist from the tropical continent over the Plata Basin; both advections favour the precipitation processes, which originate great discharges in the Paraná River and floods in the lower section of this river. This type of floods occurred with more frequency since the change in the phase of El Niño events during the 1980 decade (1983, 1992 y 1998) coinciding with a great change in the atmospheric circulation observed during the middle of the 1970 decade that could be related to global warming. 26

40 Fig. 2.16: Anomaly precipitation (mm) for March-April-May (+) Floods in the Uruguay River Compared with the Paraná and Paraguay basins, the Uruguay River basin is relatively small, extending over an area of less than 0.4 x 10 6 km 2. Because of its size, its narrow transverse section and the stepped terrain, the lag between rainfall and the river discharge takes only a few days. Most of the large discharges persisted for only a week or less, with the exception of two events in 1983 and 1998 that were part of flood wave of about 2 months. Both occurred during the strongest El Niño (EN) events of the century. This indicates that the main discharges are usually caused by synoptic events or by a short succession of them, although these events could be modulated by some remote forcings as El Niño. Since the greatest discharges in the Paraná River lasts for many months, there is a fair chance that they can be superposed to a great peak of the Uruguay River, even more when the greatest discharges in both rivers tend to occur with higher probability during EN events. In fact, during June 1992, both rivers had great discharges that added up to 76,500 m 3 /s. In view that this was not the case of the greatest discharges at both rivers, an extreme total discharge of near 100,000 m 3 /s in the Plata River cannot be discarded. This implies an anomaly of about 75,000 m 3 /s. In section 2.2.4, it is explored the effect of such possible discharge on the level of the Plata River at several locations of both margins. Date Discharge anomaly (m 3 /sec) 9 June , April , July ,831 7 January , April ,779 5 May ,678 8 March , June , October , June , April ,187 27

41 9 September ,664 1 May , November , November , September ,913 Table 2.7: Extreme daily discharge anomalies (larger than 3") of the Uruguay River at the gauging station of Salto ( Geology, geomorphology and Delta accretion Geology The geological and geomorphologic aspects of the coastal area of the Plata River including its geomorphologic evolution and sediment transports were addressed during the Project. The coastal dynamics during past and present times was also discussed, making a special reference to the Paraná Delta advance. Two maps, based on the identification of the geological units that emerge at the La Plata River coasts, were constructed to help the description of the geology of the area. The identification of the geological units was carried out using aerial photographs of 1:40,000 scale, and satellite images of the Landsat 5 TM satellite (Thematic Mapper) which possess a spatial resolution of 30 metres and sensors in 7 spectral bands, from the visible to the infrared. Some specific geological features of interest for the coastal dynamics are pointed up in the following paragraphs. For description purposes the coast was divided in two zones, from the Paraná Delta to Punta Piedras, and the Samborombón bay coast from Punta Piedras to Punta Rasa Paraná Delta - Punta Piedras sector The general geology of the coastal area in the Paraná Delta is constituted by unconsolidated sediments corresponding to the sandy fraction in the levee areas and to clayey silts in the islands and submerged front, Fig There is also presence of sandy sediments in the submerged delta. Likewise, areas of sandy coastal ridges and clayey silty tidal flats constitute the coastal front between the Delta and Punta Piedras. In this part of the coast of the Plata River, there are deposits of estuary beach containing abundance of molluscs. Inside the tributary valleys, fine sediments, typical of swamps and coastal marshes ecosystems, replace the shell and sand deposits. About 2,000 years ago took place the stabilization of the sea level and thereafter some of the Buenos Aires territory was subject to light erosive processes. In the Plata River headwaters, the formation of the Paraná Delta continued with the advance of islands and bars, as well as of the great front barely submerged under the waters of the estuary. Currently, the coastal area located between the Paraná Delta front and La Plata City has been greatly modified by human activities. The analysis of outcrops should be done retrospectively because less than a hundred years ago it was still feasible to watch them, while today they have been removed or covered. The best watching places were on the scarpment as well as in the valleys of the streams that reached the coast. 28

42 Fig. 2.17: Geology of the coastal fringe of the Plata River between the Paraná Delta and Punta Piedras Samborombón Bay The data presented in this section were obtained from samplings made for this Project and from previous works (Tricart 1973; Fidalgo et al 1975; Parker et al 1990; Codignotto and Aguirre 1993 and Kokot 1999). The coast is constituted by clayey sediments corresponding to tidal flats deposits and a cheniers line, where a crab s community area is located. In the continental area, outside of the present riverside line, there are coastal ridges and barrier islands of the Holocene constituted by sand and containing abundant marine molluscs, Fig 4.2. Spalletti et al. (1987) studied the sedimentology, while Codignotto and Aguirre (1993) and Aguirre (1996) described the geomorphology, genesis and associated fauna of these deposits. 29

43 Fig. 2.18: Geology of the Samborombón Bay between Punta Piedras and Punta Rasa The outcrops between Punta Rasa and Punta Médanos (located southward the study area) correspond to dunes deposits and Holocene beach ridges deposits, constituted by medium and fine sands and containing bivalves and gastropod fauna, partially cemented with calcium carbonate. The area was formed during the last Holocene transgression (Dangavs 1983) and it has grown starting from a cape that was located southern to Punta Médanos (Violante 1988) where the deposits were studied by Teruggi (1949). 30

44 The beach ridges deposits corresponding to barrier spits that constitute the present coastal line between Punta Rasa and the south of Punta Médanos are composed by sands with fossil mollusc remains (Codignotto and Aguirre 1993; Kokot 1997). There is also find fine sandy deposits and organic remains of fossil island barriers. Clays, silts and fine sands constitute the tidal flat deposits, while in the main streams valleys there are alluvial deposits, mainly sandy ones Geomorphology The geomorphology of the area was interpreted from three sources: the satellite image Landsat 5TM 224/085 obtained at March the 3 rd of 1998 and provided by the National Commission of Space Research (CONAE) with a spatial resolution of 30 metres, aerial photographs of a scale of 1:40,000 and field tasks. Two maps were constructed, figures 2.19 and The digital treatment of the satellite image, combining bands and the application of filters allowed differentiating among the areas of interest, separating units that are enhanced in the image by the unequal presence of water and vegetation. As a result, the areas affected by tidal floods and storm surges are clearly distinguishable in the interpretation of the geomorphologic maps. The obtained information from satellite analysis also allowed determining the wetlands and the other geomorphologic units. Paraná Delta Punta Piedras sector From the front of the Delta towards the city of Buenos Aires, the coastal area shows the presence of a palaeocliff in whose base there is a estuarial terrace conformed by beach ridges cords, tidal flats and beaches. The group constitutes a low area subject to floods caused by storm surges. The coastal area of the city of Buenos Aires was completely modified by fillers. The city of Buenos Aires can also be divided in two areas of different characteristics because of its geomorphologic attributes, a high area, presenting altitudes above the 20 metres over the sea level, and a lower area, corresponding to the coast whose variable altitude is approximately less than 5 meters over the sea level (Fig. 2.19). The two areas, differentiated by their altitude, correspond to two geomorphologic features. The high area belongs to the geomorphologic province of the Pampa Undulate, presenting a relief formed mainly by fluvial action. The observed undulations correspond to a system of rivers and watersheds in an environment modified by human action. 31

45 Fig. 2.19: Geomorphology of the coastal fringe of the Plata River between the Paraná Delta and Punta Piedras The lowest area corresponds to the coast of the Plata River. It is an accretion area originated during the Holocene, corresponding to fine sand beaches and silt-clayey tidal flats lying on a compact material layer of calcareous crust. Artificial fills, port constructions and coastal defences currently expanded the area, as it is shown in the figure Deposits of clays, plastic silt-clayey floors and sandy floors also constitute these lands. Sometimes, the coastal area is flooded by the waters of the estuary, because of meteorological tides locally named sudestadas. The most recent deposits are those of the Paraná delta whose southern submerged border, the prodelta, is reaching the coastal area of Buenos Aires City Samborombón Bay sector The different geomorphologic units that compose this zone are shown in the figure The northern area corresponds to a higher zone, where the landscape was originated by fluvial action and it constitutes a plain area furrowed by some rivers that flows into the Samborombón Bay and others that drain to the coastal area located to the north of Punta Piedras. This relief presents a barely developed cliff that separates the Samborombón Bay continental lands from areas constituted by tidal flats, which are classified as follows: 32

46 Ascended tidal flats: Not attained to current marine action; they were originated during the Holocene marine transgression, ageing about 6000 years AP (Codignotto and Aguirre 1993). Due to the scarce slope of the area, there is not a well organized drainage. Nevertheless, the drainage is organized following trajectories that are controlled by the old tidal creeks. Extraordinary tidal flats: Area flooded at times by extraordinary tides. When the sea level rises because of storm surges, there is an immersion of a strip of about two kilometres wide, which is generally higher than the mean high tide level. Therefore, a sublitoral environment is developed constituted by an area of high crabs community (Tricart 1973) where a series of lagoons are flooded during the extraordinary storm tides. In general, the tidal creeks do not connect these lagoons and the water level is lowered slowly by evaporation. In this coastal area the waves are not very effective due to the control action generated by the contact with the scarcely leaned bottom. Nevertheless, along the coast it is noticed certain erosion effect. Semidiurnal tidal flat: Corresponds to the coastal strip that is exposed during the semidiurnal tidal cycle. It is a surface slightly leaning toward the sea and furrowed by tidal creeks, which reach great development between General Lavalle and San Clemente del Tuyú (the two urban areas indicated in Fig 2.20). Next to this last town, the tidal creeks orientation is controlled by the presence of beach ridges. 33

47 Fig. 2.20: Geomorphology of the coastal area of the Samborombón Bay Other geomorphologic units are: Cheniers: Ridge morphology area of scarce relief located in the central area of Samborombón Bay. 34

48 Beach Ridges: Placed in the north - centre of Samborombón Bay and in the southern area, conforming Punta Rasa and a spit that extends between Punta Médanos and Punta Rasa. Codignotto and Aguirre (1993) explained the genesis of the area, and Kokot (1997) studied the beach ridges deposits and explained the coastal dynamics. Dunes: Corresponds to the area of coastal dunes located in the eastern coast of the Punta Médanos - Punta Rasa area. Alluvial plains: The most important are those corresponding to the Salado and Samborombón rivers, located in the northern area of Samborombón Bay. This geoform can also be defined in the mouths of some smaller streams next to the General Lavalle locality Sediment transport The inner area of the Plata estuary, which extends from the Paraná Delta to the section that goes from Colonia to Buenos Aires, is influenced by the advance of the Paraná Delta represented by low bottoms, denominated Playa Honda. In this area, the average depth is 2.5 meters, increasing toward the outer zone. It is possible to recognize a flow with suspension materials almost continuous with a S-SE direction. It is observed at 700 / 800 m from the coast, and more intensively in the centre of the estuary and in the creeks. It transports silt-clayey sediment, contributed mainly by the Paraná River that flocculates when contacting the brackish waters. The concentrations are variable and present values between 15 and 250 mg/l. Besides this current, there are others of tractive type whose directions respond to the action of the waves on the coastline. These currents are denominated drift currents. The more important, though discontinuous, tractive drift current carrying on thick material coming from the erosion of the coast has a main course toward the N-NO and takes place in the interface between water and land. The origin and dynamics of this current responds to the storm surge pulses. The N-NO direction of the tractive current is evidenced by the orientation of the outlets of the small tributaries and creeks, and in the areas with deposits of removed materials, either of natural or artificial origin. These last ones include many types of residuals, deposited along the coast with the objective to win land to the River. It is important to point out that the direction of the tractive current sometimes shifts due to the incidence of waves from the north-east Recent geomorphologic evolution During the last 7,000 years there were relative sea level rises and descents in the area of the Plata estuary, which were accompanied with erosion and accumulation. To these processes, it should be added the advance of the Paraná Delta front. These factors coupled with the development of the human activities during the last 100 years determined the configuration of the coastline. Since the year 1907 human activities increased the land area of the coastal sector between the city of Buenos Aires and the Delta front in approximately 10 Km2. The coast of the City does not exhibit natural areas because all the sectors were modified by human action. Where the filling tasks advanced more, the depths offshore jump quickly to 1.3 m to 2 m because the natural slope of the bottom was broken by the filling and the construction of coastal defences. At the Samborombón bay, the present coastal forms have been developed over a Pleistocene erosive platform. Fig, 2.21 shows the coastal shoreline during late Pleistocene when the sea was approximately 10 m under the current level (A). Thereafter the Holocene transgression- regression took place and the sea level reached 5 m over the actual sea level (B) and during the subsequent Holocene regression the level lowered, reaching 2.5 m over the current sea level 3,500 years BP (C). Finally, the part D of the figure shows the current contour and the coastal dynamics. The incidence of the south easterly waves concentrated the erosion on a tip located between Mar Chiquita and P. Médanos. This process allowed the formation of beach ridges, which drifted toward the south and formed the Mar Chiquita lagoon; at 35

49 the same time, a barrier spit of great growth was generated to the north. The relict forms of both barriers are between 5 m and 2.5 m high over the present sea level, Fig Fig. 2.21: Evolution of the Samborombón bay during the high quaternary (Codignotto and Aguirre 1993) In the present, between Punta Médanos and Punta Rasa, the only natural accretion sector has however erosion processes caused by human activities. Also, in the Samborombón Bay coastal fringe there are clear signs of incipient erosion, mainly at the ascended tidal flats (1.80m-0.25m over the sea level) and at the tidal flat The Paraná Delta The Paraná Delta showed a striking advance inside the Plata estuary that has been known by some researchers, but it was not documented. Therefore, in view of the current Climate Change and its consequent sea level rise, it was studied if the Delta front advance has increased, diminished or maintained its growth rate. For this purpose, a map showing the mobility of the deltaic front through the time was elaborated making use of the historic information, Fig Cartographic information from 54 maps, which cover the period between years 1731 and 2003, was carefully analyzed. In addition, bibliographical information referred in Furlong (1936, 1963), in the Ministry of Public Works (1908) and in Outes (1930) was also consulted. A time consuming work was carried out to determine the value of these information. It should be pointed out that the oldest 36

50 cartographic information has very important uncertainties that had to be compensated by a critical and meticulous analysis The mouth of the Reconquista River was almost free of obstacles in 1731 flowing into the Plata estuary; between 1802 and 1829 was partially obstructed by islands, and finally since 1890 there was any longer a direct connection between the Reconquista and the Paraná de las Palmas rivers. With the Luján River happened something similar. It was free of obstacles in the maps corresponding to the years 1731, 1756 and 1783, but not in the outline made in 1731 and in the maps dated in the years 1762 and For these reasons there it is certain confusion around the eighteen century situation, although credit was given to a 1731 outline since was sketched by a resident or direct observer of the neighbourhood. In any case, it is clear that the mouth of the Lujan River at the Paraná de las Palmas was found it further from the delta front as more recent the maps were. The outline corresponding to the year 1731 shows that both the Lujan and Reconquista Rivers flowed into Paraná de las Palmas. There is some confusion between the maps of the nineteen century because there were islands at their mouths. On the other hand in later maps (at the end of that century), the continuous contribution of sediments made both rivers to flow directly into Río de La Plata estuary. Finally, since the beginning of twentieth century only the Luján River flows into the Plat River after receiving the waters of the Reconquista River. Fig. 2.22: Progress of the Paraná delta during the last 250 years The delta advance during the last 250 years can be summarised as follows: Between 1750 and 1800 the delta front advanced 9 km, that is to say, an average of 180 meters per year. 37

51 Between 1800 and 1850 the front of the delta advanced 5 km which results in a rate of 100 meters in average per year. During the period between 1850 and 1900 the delta prograded 5 km, at an average rate of 100 meters per year. Between 1900 and 1950 the delta advanced 3 km, at an average rate of 60 meters per year. Between 1950 and 2002 the delta advanced 4.5km resulting in a rate of 90 meters per year. Finally the current data indicate an advance between 60 and 70 meters per year. The growth rate decrease of the Paraná delta can be observed more clearly in the figure Fig. 2.23: The Paraná Delta Front advance The reduction of the speed in the progress of the Delta is compatible with an increment of the river level, which was 17cm in the Buenos Aires port during the twenty century. On the other hand, the extraordinary and long lasting flood of the Paraná River in 1982/1983 deposited an unusual amount of heavy sediments on the bottom of the Rio de la Plata, and thus, favored the later progress of the delta front. The 1982/1983 flood was unusually long because started some months before the strong El Niño event, and remained during six months after the end of this event. It seems that the long lasting flood of 1982/83 was the consequence of three consecutive climate forcings, only one of them was a very strong El Niño event (Camilloni and Barros 2000, Barros et al 2004). Therefore, it seems unlikely the occurrence of another long lasting flood such as that of 1982/83 during the next decades. If this were the case, the rate of growth of the delta front will diminish again, as the water level of the Rio de la Plata continues to increase. Other less likely scenario could be the occurrence of another long lasting flood as the one of 1982/83. In such a case, after this episode, it could be anticipated a faster growth of the delta front Topography For calculating the return periods of floods over land, a digital model of the surface altitude over sea level was constructed with 0.25 m vertical and 1 Km horizontal resolution, using field measurements taken 38

52 with a GPS differential system and data from satellite radar. GPS measurements were taken at places that provided key information according to the geomorphologic maps that were constructed to help the construction of the digital model. These data was complemented with the information of preexisting lower resolution topographic maps and with altitude measurements taken by the Buenos Aires city at certain points. The four sources of data present difficulties. The topographic maps have an altitude resolution of 1.25 m and in some areas 2.5 m, too low for the study of present floods and for assessment of future changes in the affected areas. The GPS measurements, taken as unique source, would require too lengthy and costly campaigns. In addition, access to certain areas is very difficult and furthermore this system presents difficulties when used in urban areas due to interferences caused by trees, buildings, and cars that interrupt the satellites reception. This was a problem in part of Buenos Aires City, but a few hundreds observations of altitude data were available from former measurements ordered by the Buenos Aires administration. However, these data was restricted to a certain areas within the city boundaries. Radar data are very accurate, but the built areas and trees introduce errors. The first step was the cartographic compilation of available information and construction of a digital map in Autocad format. The coordinates were Gauss-Krügger, a system used by the local Military Geographical Institute. The map includes the Argentine coastal area of the Plata River from the Paraná Delta to Punta Rasa and extends inland to the altitude of 5 m over mean sea level. Field measurements with a GPS differential system were obtained during 12 campaigns. The measurements were taken at places that bring key information. The data were processed with the software Ashtec. It indicates the height of the ellipsoid according to the system WGS 84. The digital model of the surface altitude over sea level was started with the initial input of radar data with a horizontal resolution of 90 m. The data were filtered, eliminating noises caused by the presence of buildings, vegetation and small ponds. For this purpose, the other three sources of data were used. The initial radar data of 90 m horizontal resolution were used to produce an altitude model of a 1 km cell size that includes the maximum, average and minimum value. In urban and suburban areas, the minimum value is likely more representative of the real mean altitude because it may stand for the areas without or with less buildings and trees. In addition, these minimum values are more functional to the purpose of the map, which is to asses the frequency of flooding. Finally, a 1000 m horizontal resolution grid cell was built. The constructed digital model covers the Argentine coastal area of the Plata River from the Paraná Delta to the Cape Rasa and extends inland up to 5 m over mean sea level, Figs and Although the digital map has a resolution of 0.25 m, figures 2.24 and 2.25 show only a 0.5 m resolution to avoid a confusing depiction. 39

53 Fig. 2.24: Topography of the coastal fringe of the Plata River between the Paraná Delta and Punta Piedras. Altitudes over the zero IGM; see Fig. 4.1 As it will be seen in section 4.3.4, the sea level rise will propagate almost 100 % into most of the inner Plata River. However, as long as this rise will be not more than 50 cm, almost no land would be permanent flooded in the sector between the delta front and Punta Piedras because as seen in figure 2.24, only very small areas are below the 1.5 m mark over the IGM zero. This altitude (1.5 m) would be reached by the shoreline after a 0.5 m rise in the water level at Buenos Aires, Fig On the other hand, in the Samborombón Bay, a 0.5 m rise will lead to some permanent flood in some areas that are below 0.50 m, Fig

54 Fig. 2.25: Topography of the coastal area of the Samborombón Bay. Altitudes as in Fig

55 Wave climate The purpose of this activity was to produce a quantitative estimate of the wave climate in the inner Plata River considering a possible future change in local winds. First, sea and swell climate in the outer Plata River were statistically analyzed from direct observations obtained at the outer Plata River. Second, propagation and transformation of sea and swell, from the outer Plata River throughout the intermediate and inner regions, were computed and analyzed. Present wave climate (directional wave heights and periods) was estimated by a hindcasting methodology based on ten-year statistics of winds measured at the Aeroparque meteorological station. There are no direct measurements of waves in the Plata River except for a single 5-year record of wave data gathered in the outer area. Based on these data, Anschütz (2000) showed that wave climate in the outer Plata River is a combination of swell (wave generated far away, not related to local wind) and sea (wave generated by local wind). The analysis of the waverider data revealed predominant heights between 0.5 to 1.5 m. When sea prevailed, periods were between 4 and 6 s, when swell prevailed they were between 10 to 12 s. In the inner Plata River, wind waves have been neither measured nor modeled. Consequently, a realistic wind wave climate in this area was not available Sea and swell climate in the outer Plata River A single series of 11,297 records, gathered from June 1996 to November 2001 was obtained with a directional wave recorder Datawell Waverider in the outer Plata River at latitude S and longitude W, Fig Table 2.26 shows the number of occurrences for the eight directions analyzed. SE direction, followed by E and S are the main directions of propagation with 41%, 28% and 14% of occurrences, respectively. Frequencies for the rest of the directions are equal or lower than 5%. The bi-dimensional distribution for the direction with largest number of occurrences, i.e. SE, is shown in Fig Two domains with very high number of events around periods of 10 s and heights of 0.8 m (swell) and periods of 5 s and heights of 1.25 m (sea) can be clearly identified. Fig. 2.26: Hindcasting point and directional wave recorder location are indicated. Depth contours in meters are depicted 42

56 Considering the general orientation (NW-SE) of the Plata estuary and its very shallow waters, only those waves propagating from the southeast could reach the inner part. Thus, swell and sea propagating from the outer to the inner part of the estuary would be the long and short swell, respectively, within the intermediate and inner regions. Direction Number of events Percentage % N NE E SE S SW W NW Table 2.8: Number of events for each of the eight analyzed wave directions at the outer Plata River Fig. 2.27: Bidimensional distribution of heights and periods. Mean sea and swell characteristics are indicated Swell climate in the inner Plata River Wave propagation and transformation from the outer Plata River towards the coast of Buenos Aires were analyzed by means of a numerical model of wave transformation (Dragani and Mazio, 1991). The wave number, a parameter which must satisfy the modified dispersion relation at any point (Watanabe, 1982), 43

57 was computed considering realistic bathymetry and stationary, but spatially variable current field corresponding to flood and ebb conditions. Refraction, shoaling and friction effects along the wave ray were computed based on a bathymetric grid of 1 Km spatial resolution obtained from the nautical charts (SHN 1999a and SHN 1999b). The refraction coefficient was computed by the classical methodology given by Griswold (1963). The shoaling coefficient was based on wave velocity at any point by applying the linear wave-theory. To obtain the rate at which energy was removed from waves we adopted the formulation given by Putnam and Johnson (1942). Energy dissipation was computed along the ray using an integral expression given by Vincent and Carrie (1988). Fig. 2.28: Refraction diagram corresponding to wave coming from Southeast. Refraction, shoaling and dissipation coefficients are indicated. An intermediate ray (dashed line) reaching the inner Plata River is included We herein assumed that waves propagate as swell from the outer Plata River towards the inner part of the river under non-locally generated wave conditions. Two different swell conditions were analyzed: short swell, associated to sea in the outer part of the river and long swell, associated to swell in the outer Plata River. Results show that shallow water effects (especially refraction) are less evident over short swell, thus being the most likely one to reach the intermediate and inner Plata River. Outputs corresponding to short swell coming from the southeast are shown in Figure The figure depicts a refraction diagram for twenty rays directed towards the NW from the outer part of the river. Shallow waters and banks were identified as areas where wave braking (rays caustics) occurred. Caustics seem to have an important role in wave propagation and transformation towards the inner Plata River, thus producing the consequent wave attenuation. Figure 2.28 also shows that rays strongly divert at the intermediate part of the river producing two areas of caustics: one located between Punta Piedras and Atalaya, (near the coast of Buenos Aires) and another southeast of Colonia (near the coast of Uruguay). In order to analyze the wave transformation between the outer and inner regions, an additional ray, able to reach the inner region was computed (dashed line in Figure 2.28). Heights of waves propagating from the outer to the inner part of the river were attenuated 95 % by refraction, shoaling and friction effects. Consequently, predominant wave climate in the upper Plata River can be described considering only wind waves locally generated (sea). 44

58 Sea climate in the inner Plata River The Wave Hindcasting Method (CERC 1984) and the improvements given by CERC (2002) were applied at a coastal point near Buenos Aires City (Fig. 6. 1). The methodology was applied in shallow waters, off the surf zone, where local depth is of approximately 2 m. Monthly wind statistics (SMN, 1992) were obtained from hourly data gathered at the Aeroparque meteorological station, located on the coast of Buenos Aires City (Fig. 6.1). Based on these data, it was estimated mean wave heights. Directional mean winds used in the wave hindcasting are presented in Table Index Direction Frequency ( %) Mean wind ms -1 Fetch (Km) H MO (m) 1 N NE E SE S SW W NW Calms T P s (sec.) Table 2.9: Directional mean wind and probability of occurrence at Aeroparque on the coast of Buenos Aires City, e directional fetch, heights and periods obtained by hindcasting Fetch-limited conditions have been considered herein given the nature of the meteorological data available and considering the geographical characteristics of the inner Plata River. Under these conditions, the supposition is made that winds have been blowing long enough for wave heights to reach equilibrium at the end of the fetch. The parameters required for hindcasting are fetch and wind speed, the latter being representative of the average value over the fetch. The wave parameters computed are the energy-based wave height and the peak spectral period. In shallow waters the deep-water methodology is applicable CERC (2002). Shallow-water formulae are quite close to the ones of deep-water wave growth for the same wind speeds, up to a point where an asymptotic depth-dependent wave height is attained. In light of this evidence it is convenient to disregard bottom friction effects on wave growth in shallow waters (CERC, 2002). Table 4.2 shows that the East direction presents the highest frequency (18.4% of cases), N, NE, SE and S directions present similar frequencies (ranging from 11.5 to 15.8%) and SW, W, NW directions and calms are the least frequent ones, with values ranging from 6.5 to 7.8%. Directional mean wind intensity (from 4.6 to 6.1 ms-1) is quite uniform for all directions. Table 4.2 also presents mean wind intensity, frequency, and fetch for the eight analyzed directions and the heights (HMO) and periods (TP) estimated by hindcasting. The largest heights and the longest periods resulted for the Southeast 1.22 m and 5.5 s and for the East 0.9 m and 5.3 s, respectively. The shortest heights and periods resulted for the West with 0.12 m and 1.4 s, for the Southwest with 0.13 m and 1.4 s and for the South with 0.15 m and 1.4 s. 45

59 Effect of the local wind change on the wave climate of the inner Plata River The western border of the South Atlantic High is moving southward, see section This displacement has produced increased frequencies in east winds over the Plata Estuary. It was assessed the changes in the wave climate in the inner Plata River considering a possible future change in local winds. Since wave climate in the inner Plata River can basically be represented by sea, the hypothesis is that in this case, the small changes both in frequency and intensity of local winds should affect the wave climate. Therefore, present and future mean directional wave heights are estimated and compared. Based on the results described in section 4.2.4, a possible scenario with an increase of 30% and 10% respectively in the east wind frequency and intensity was analyzed. In this hypothetical scenario it is assumed a decrease of the same magnitude would occur in frequency of the west direction. Results of the hindcasting show that, mean East wave height and mean total wave height will increase within the inner Plata River. Present mean East wave heights (0.90 m) will increase by 0.12 m (13 %) and their frequencies will increase by 30 % (from 18 to 24 %). The mean period for East waves will not change significantly (less than 4 %, from 5.3 to 5.5 s). The ratio between the future and the present total average energy would be about 1.3. This means that, an increase of 30 % could occur under this future scenario of climatic change. An increase in the mean wave energy would implicate a higher capability of water to maintain sediments suspended within the water column and thus an increase in the suspended sediment transport which could modify the sediment transport dynamics and would affect the deposition rates. Another conclusion is that the coast of Buenos Aires City will be more frequently exposed to wave effects, giving rise to intensified associated littoral processes. Given the predominant orientation of the coast of Buenos Aires, the tractive currents carrying on thick material, north-westward will be increased as well as the drift currents along the shore, see section Therefore, it should be expected an increment of both the accretion of sediments upstream the structures and of the erosion of the shore downstream the coastal emplacements. Although there are many uncertainties in these predictions, the potential impacts of future changes in the wave climate need to be assessed. Likewise, decision-makers should thoroughly consider the possible impacts on the coast of Buenos Aires City Hydrodynamic modelling There are several relatively recent papers that offer a physical characterization of the Plata River, including the continental platform (Framiñán et al. 1999, Campos et al. 1999, Piola et al. 2000, Menéndez 2001). On the other hand, there is more experience in the development and application of numerical hydrodynamic models to the Plata River being treated as a shallow water system, for which is sufficient a two-dimensional description in the horizontal plane. It is important to stress that, in order to simulate storm waves, the two-dimensional horizontal analysis turns out to be sufficient, since its wave length is quite larger than its depth (Whitham 1974, Menéndez and Norscini 1982). Though there were initial attempts in the decade of 1970, the first systematical development of a hydrodynamic model was presented in 1986 (Molinari 1986), that used the software HIDROBID II (Menéndez 1990). Improvements and applications of this model continued in successive theses (Albarracín 1987, Olalde 1988). Since then, many other models of the river have been developed (Simionato and others 2002). A new version of the HIDROBID II, named RP200 has a great spatial resolution (mesh of 1 Km per side) and has been carefully calibrated (Jaime and Menéndez 1999) The RPP-2D model Hydrodynamic model RPP-200 was taken as the basic model to develop the RPP-2D model, based on software HIDROBID II, a 2.5 km x 2.5 km resolution 2D-Horizontal model. This software is based on the numeric resolution of the shallow water equations. Its performance was shown to be comparable to more sophisticated, but much higher time-computer demanding 3Dmodels like the HANSOM-CIMA. 46

60 Fig. 2.29: Calculation domain of model RPP-2D Given its barotropic nature, the software cannot represent the vertical stratification due to the effect of the salinity. The domain of the RPP-2D model is large enough to include the Plata River and an extended area of its maritime front to simulate the generation of storm waves, Fig, The theoretical model considers as driving forces the gravity, the Coriolis acceleration (inertia force due to the rotation of the Earth) and the superficial tensions due to the action of the wind. On the other hand, it includes the resistance to the movement resulting from the generation of turbulence at the bottom (historically named "friction"). It can have any form at the bottom, but constant in time (fixed bottom). The hypothesis of quasi-two-dimensional flow means that the movement is essentially bidirectional and the speed is practically uniform along the vertical direction and, consequently that the vertical acceleration is negligible with respect to gravity, resulting in a hydrostatic distribution of pressures. The shallow water equations results after filtering over the statistical ensemble of Navier-Stokes' equations (Reynolds' equations), followed by the vertical integration and the application of the simplifying hypotheses of the theoretical model (Abbott 1979) are: ()()000()11()()0 ()11()()0 fxsxgxxxyfysygxyyyhhuhvtxyhzuuuuvfvghthttxyxhhhxhyhzvvvu where x and y are the spatial coordinates, u and v the mean vertical speeds in those directions, respectively, f g the Coriolis factor, # sx y # sy the sliding tensions on the bottom and T the tensor of the effective tensions (includes the effects of viscosity, turbulence and differential advection). The numerical scheme of resolution of these equations used in the software HIDROBID II is based on the method of the finite differences. The grid is of the alternated type (the two components of the speed and 47

61 the water level are centered on different nodes). The method is implicit with alternated directions (Menéndez 1990). The domain of the model RPP-2D is delimited by physical and mathematical contours. The physical contours are the Uruguayan and Argentine coasts. The mathematical contours are on the Maritime Front: the parallels 35.8 S in the north and 40.5 S in the south and the meridian 51.5 W in the west. The Front of the Paraná Delta is considered to be also a physical contour, with the exception of the mouths of the Paraná and Uruguay rivers, which constitute mathematical contours. The bottom depth information was obtained from the combination of two data bases, one provided by the Service of Naval Hydrography (SHN) of Argentina (Dragani 2002) about the Plata River and its Maritime Front and the other one supplied by the same SHN consisting of information of Plata River digitized depths (CARP 1989) In the numeric model of the terrain a spatial rectangular discrimination grid was adopted, with cells of 2500 m per side ($x = $y), on a system of coordinates orientated according to the cardinal directions. It resulted 382 cells grid in east-west direction and 408 cells in the direction north-south, of which about 55 % falls over the continent, so that they do not intervene directly in the calculation. Depth values were assigned to each of the cells of the model grid by a process of interpolation with the "krigging" technique. The depth so generated is referred to the chart of local reduction, with distance to the geode surface that is variable; therefore the surface of reference of the depth is not an equipotential one. However, it was verified that this systematic error has no quantitative significance in the results. The drag of the bottom is significant only in the inner Plata River where depths are low, losing importance on the outer part of the river, and at the Maritime Front. Thus, it was adopted a uniform value for the drag coefficient of Manning s for the whole domain of the model, namely 0,015, which was obtained in the calibration of the RP2000 (Jaime and Menéndez 1999). The discharge of each of the two large tributaries (Paraná and Uruguay) was forced as a boundary. It can be a constant discharge in time, if the interest is to represent mean conditions, or a variable one depending on the process to be simulated. It is not necessary to include the modulation effect caused by the tide wave in the discharge (a priori unknown), since it only affects a very short area near the boundary (Jaime and Menéndez 1999). The model has three oceanic borders (East, North and South), which constitute mathematic contours. The East border was considered as impenetrable on the basis that the wave energy that crosses it is very low in relation with the one that propagates along the continental platform (tests carried out imposing the tidal wave showed that this approach is satisfactory). The north edge was treated as a not reflecting contour, allowing the exit of the waves that affect it without reflecting information. Astronomic wave tides are imposed as a contour condition in the south edge of the model, on the basis of the existing knowledge that in this region tide waves effectively propagates from south to north. The tide wave is built by combining the information registered in the Mar del Plata station (since it is the closest station with reliable historical records), suitably corrected in amplitude and phase, to represent the oscillation on the coast, and the information obtained from the global model of tides RSC94 (Cartwright and Ray 1990), to represent the oscillation off-shore. The latter tool comes from a model of generalized response and from the utilization of the weight of its responses, derived from Proudman's functions, calculated for a 1 º grid that covers the area located between the latitudes -68 º and 68 º. The solution of the tide is based on the contribution of the measurements of the altimeters TOPEX-POSEIDON and information of about twenty stations of tide observation. The combination between the coastal oscillation and the off-shore wave was made adjusting the incoming wave with a certain angle respect to the normal to the contour and with an exponential decay of its eastward amplitude, compatible with its Kelvin wave character. The wind fields forcing the water surface were generated from the NCEP/NCAR's reanalyses (Kalnay et al 1996). These have a space resolution of 1,9048 of latitude and 1,875 of longitude (Fig. 2.30) and a time resolution of 6 hours. The data base matches with a grid T62 Gaussian with 192 x 94 points located inside the latitudes 88,54N-88,54S and 0E-358,125E. 48

62 Fig. 2.30: Example of the wind field from NCEP/NCAR Since NCEP/NCAR's wind fields underestimate the intensities of the observed winds, following the experience of the model HANSOM-TOP (Simionatto et al 2002) these intensities were increased in a factor of the form 1+exp [-(W/X) m], where W is the module of the wind speed, X a speed value (of the order of the larger intensities of the data base) and m an exponent. The utilization of this factor seeks to duplicate the values of the very low intensities of winds and to keep the most intensity winds unaltered. In the RPP-2D model, routines that take the information of the NCEP/NCAR fields and perform a bi-linear interpolation in the whole domain were implemented. Since the software HIDROBID II is based on an implicit scheme of finite differences, it does not have serious limitations for the time step value. Therefore, the election of this step is mainly conditioned by the precision criteria required. As the phenomenon of the most rapid scale of the present problem are the superficial waves, which move on the much slower flow at Lagrange's speed, the temporary step of calculation $t should be chosen on such a way that represents adequately the displacement of these waves along the domain. Then the following condition can be imposed on the temporary step: maxxtcδδ: where cmax is the maximum speed intended to be adequately solved. Then, since the wave energy is concentrated basically in the continental platform, the wave speed in that zone has been considered the maximum speed (the oceanic depth is significantly greater than the Plata river one). Since the maximum speed in the platform is of about 30 m/s, then $t % 80 seg. Thus, a 60 seconds step was used and consequently 720 steps were needed to represent a 12 hours tidal oscillation Calibration of the model First, the model was calibrated to a pure astronomical tide scenario (only the oscillatory component was present) aiming to reproduce the waves predicted in the Tides Tables of the SHN and of the SOHMA (Service of Oceanography, Hydrography and Meteorology of the Navy, Oriental Republic of the Uruguay), from the classic harmonic analysis of the records, for all the interior monitoring stations at the calculation domain. The south edge contour condition was adjusted, establishing the criteria to correct the amplitude and phase of the tidal wave to the Mar del Plata station, representing the variation of the water level on the coast, but assuming the entry angle and the parameter of exponential decay eastward in order to be compatible with the tide wave of the global model. The application of a low pass filter to the data provided by the tides tables allowed to distinguish an oscillation of low frequency (period of around 14 days) in almost all the stations, of variable amplitude from one station to another. The above mentioned oscillations were not considered for modeling, and therefore, were eliminated. Results for many places at the coast showed a quite satisfactory agreement. As an illustration, Figure 2.31 shows the comparison between water level Tide Table data and model results. 49

63 a) b) Fig. 2.31: Comparison of water level from tide tables and model for pure astronomical tide in Buenos Aires a), and Montevideo b) A second step in the calibration aimed to simulate the conditions of monthly mean level. The decade that goes through 1990 and 1999 was taken as representative of the present conditions. a) 50

64 b) Fig. 2.32: Statistics of Buenos Aires water level a) mean and b) frequency distribution To save computer time, an analysis was performed to determine the year that better represented the hydrological characteristics of the decade, in order to be used as representative. For this purpose, statistics corresponding to the city of Buenos Aires were used. In the figure 4.2, both the annual average levels and the decadal ones are shown; it is observed that the years 1992, 1997 and 1999 have practically the same mean value than the decade. In addition, figure 4.2 (b) shows the curves of level distributions of the year 1997 and of the decade. In view of this very good matching, 1997 was selected as the representative year. The average monthly levels for the year 1997 corresponding to the city of Buenos Aires were used to calibrate the adjustment on the average level of the sea at the south edge (the only boundary where a mean level is forced) and the drag coefficient that parameterize the surface tension due to wind (W): 51

65 sxdxsydycwwcw The value 0.77 m with respect the reference plane for the mean sea level was selected. As for CD, the law of variation as a function of wind speed shown in the figure 2.33 was chosen. Fig. 2.33: The adopted drag coefficient C D In this form, the agreement shown in the figure 2.33 was obtained. It can be observed that the annual mean level can be calculated with great accuracy whereas the mean seasonal levels reproduce the same cycle as the observed ones, with a maximum difference of about 10 cm (for the summer). This shows how adequate is to utilize the NCEP/NCAR wind fields for the representation of the mean seasonal conditions, in spite of their relatively low resolution Fig. 2.34: Mean water level at Buenos Aires from model and records In a third step, the objective was to simulate the curve of frequency of water levels in the city of Buenos Aires, considering again as representative of the 1990 decade, the year In this case, the coefficients X and m of alteration of the winds provided by NCEP/NCAR were adjusted, selecting the values X = 54 Km/h and m = 1, obtaining the agreement shown in the figure 4.2 (the class interval is 10 cm), which can be considered satisfactory. This indicates that NCEP/NCAR wind fields also are adequate for the representation of the level statistics. Furthermore, this result is of importance for the simulation of the extreme levels reached during storm surges. 52

66 Fig. 2.35: Frequency distribution of water levels in Buenos Aires for year 1997 observed and simulated by the RPP-2D model Finally, the final adjustment of the parameters were simultaneous with the comparison of the performance of the model in the simulation of large storm waves as a way of guaranteeing that the model is capable of dealing with these extreme events. Events of storm of varied levels of significance and different characteristics in the period were identified, and used for comparison: 06/Dec/1982, 06/Mar/1988, 12/Nov/1989, 31/Aug/1991, and 16/May/2000. The comparison for a storm event is presented in Figure , showing a good agreement taking into account the relatively poor wind information. The agreement is considered to be acceptable bearing in mind that with the NCEP/NCAR wind data it is not possible to expect a precise representation of isolated events. Nevertheless, it is verified that the extension of the domain of the model RPP-2D is sufficient to include the fetch of the storms. Fig. 2.36: Observed and simulated water level at Torre Oyarvide station for the November 1989 storm 53

67 Model verification Once calibrated the model, its verification was made through its application to situations different from the ones used in the calibration. As an illustration, figure 4.2 shows the results for four years of the 1990 decade. It is observed that, in all the cases, the trend of seasonal variation of the water level is correctly represented, but the quantitative differences between observed and simulated seasonal levels widen, now, to a maximum of 20 cm. a) 1990 b) 1991 c) 1993 d) 1994 Fig. 2.37: Observed and RPP-2D model simulated annual and seasonal mean levels in Buenos Aires for four different years In addition, information of the sea level with respect to the geode of reference provided by satellite TOPEX-POSEIDON for the period was used for verification of levels. The measurements belong to 148 stations located inside the domain of the RPP-2D model. The number of data of each station is variable. For information sake, in the figure are shown the measurements corresponding to station 3092, located in the neighborhood of Punta del Este, on the Uruguayan coast, which is the one with the greater number of observations. It is interesting to see the large variability of values observed. To compare this information with results of the RPP-2D model, the data in the direction towards the Plata River was analyzed, to check the gradient in this direction, Fig

68 Fig. 2.38: Level data from the TOPEX-POSEIDON satellites for station 3092 Fig. 2.39: Sampling points The observed data were averaged at every point according to the season. Only were considered the points with more than 200 observations and those that were not in the inner Plata River, where too marked variations were detected. Besides, it was obtained the mean seasonal simulated levels making the average of the series of instantaneous levels. In the figure 2.29 (a) and (b) are presented the comparisons between information and results of the model for mean summer and winter conditions during the 1990 decade along the axis of the data cloud. It is observed that the model turns out to be a suitable interpolator of the satellite information, identifying clearly the trends of the superficial gradient, which indicates major (minor) levels of the River in relation to the ocean conditions for summer (winter). a) 55

69 b) Fig. 2.40: Mean levels a) Summer, b) Winter Comparisons of the curves of level frequency for the rest of the decade of 90 were performed. As illustration, figure 2.41a shows the results for four years of the decade. It is observed that the level of agreement is similar to the one obtained for the calibration year. 56

70 a) 1990 b) 1991 c) 1993 d) 1994 Fig.2.41: Comparison of the frequency distributions of water levels at Buenos Aires obtained from measurements and simulated by the RPP-2D model Finally, speed records of currents obtained by the company Hidrovía S.A., concessionary of the dredging maintenance of the navigation channels, were used for further comparison. They correspond to 10 stations located in the middle stretch of the river, relatively close from one another. Figure 2.29 presents the comparison between two components of the horizontal mean speed according to the records of one of these stations. Bearing in mind again that, with the base of wind information of NCEP/NCAR, the model cannot represent the detail of events, the agreement is considered to be highly satisfactory. a) 57

71 b) Fig. 2.42: Comparison between recorded and calculated flow velocities, for a normal tide scenario: a) component west-east and b) component south-north. Red observed, blue model Flood modelling As explained before, floods in the coasts of the Plata River are caused by storm surges, forced by strong south easterly winds. The return period or recurrence time for each maximum level is a common tool for assessing the flood risk at coastal locations. Fig, 2.43 shows the recurrence period for annual peak levels in Buenos Aires city, according to the series data. 58

72 Fig. 2.43: Return period of the annual maximum level in Buenos Aires port The only available long record of water levels on the Argentine coast of the Plata River is Buenos Aires. This is a serious limitation to asses flooding return periods, because, the wave tides increases its height as propagates from the ocean into the Plata River. This effect appears because the depth and the width decrease upstream from the ocean. To extend the results shown in fig 2.43 to all the right coast of the Plata River, it was developed a methodology that makes use of the RPP-2D model. The methodology also permits to avoid running the model for considerable simulated time, say 50 to 100 years, which is not always possible when there are only limited computer resources. The method consists in defining a typical storm surge that can be representative of the greatest storm surges, both in its space and time pattern, even with different maximum heights. For this storm surge, the maximum height at each point on the coast has a value that determines its constant scale factor that results from its rate with the maximum value at a reference place. In our study the reference place is the port of Buenos Aires where given a level height, its recurrence is known. As typical storm surge was selected the one of May 2000, whose wave tide had a well defined peak. The model was run during 15 days starting in May 8 th and finishing in May 22 nd. The storm peak happened in May 16 th. Along the Argentine coast were defined 24 locations as shown in Fig In each of them was determined from the model simulation, the maximum level produced by the wave storm passage. This value, when compared with the respective one of Buenos Aires, gives the constant scale factor. 59

73 Fig. 2.44: Locations for the determination of flood level As an example, Fig presents, the wave storms resulting at each location for the 100 years recurrence period. As expected, the amplitude of the wave amplifies as the wave propagates to the inner part of the River. This is even more clearly seen in the curve of the maximum levels at each location corresponding to the same wave, fig Fig. 2.45: Storm waves for the 100 year recurrence period at each location along the coast, constructed from their respective scale factor. 60

74 Fig. 2.46: Maximum heights calculated for the storm surge tide with return period of 100 years 2.3 Results Results that are consider central to the Project in connection with present and future scenarios are presented in the first three sections Other results, mainly of scientific interest are presented in sections and Hydrologic scenarios In order to address the questions concerning climate change, it is necessary to specify some type of future climatic scenario. In this project, the scenario SRES A2 defined by the Third Report of the Intergovernmental Panel for the Climatic Change (IPCC) was chosen. Socioeconomic trend assumptions in this scenario are similar to the current ones, and therefore the resulting trends of greenhouse effect gases emissions would reflect what would come if the humanity does not take rapid and drastic measures to reduce these emissions. Nevertheless, some other estimate were also made with a greater rise that are compatible with the SRES A2 uncertainties and with the current rate of sea level rise as calculated from the Topex- Poseidon satellite complex. The model RPP-2D was used to develop water level scenarios for the twenty first century. Scenarios were developed for the 2030 and 2070 decades, preserving the same astronomic tides and mean discharges of the tributaries as in the 1990 decade and changing the winds fields and the mean level of the sea. As will see in section 2.3.4, these are the key variables for future change in the Plata River levels. In the case of the winds, there are two problems, in general GCM underestimate the frequency and intensity of perturbations of the zonal flow. In addition, the available climate scenarios from IPCC 2001 did not have daily data, but only monthly averages. On top of that, not even monthly averages of surface winds were available. To find a way to assess wind change, it was made the assumption that although mean winds would change, their variability will remain the same as well as their relationship with the pressure field. Then the following procedure was followed. First, it was found a multiple regression adjustment between surface daily pressure and the wind at each point of the NCEP/NCAR grid over the hydrodynamic model domain and for every month. This was made with NCEP/NCAR reanalysis data, and almost without exception the regression explained more than 95 % of the variance of both, the zonal and the meridional wind components. The second step was to choose the model that best reproduces the present surface pressure climate in the region of concern. Although, regarding the surface pressure field, the ECHAM4 had as good performance 61

75 as the HADCM3, the last was chosen because best reproduces other features of the regional climate, like precipitation and temperature, (Camilloni 2004). The next step was to calculate the surface pressure mean difference between the field given by the model for each month for the 2030s and 2070s and the 1990s. These differences were assumed to be the same as the one between the true fields in the given scenario at the 2030 and 2070 decades and the 1990 observed decade. Then the daily surface pressure scenario for each month of the future decades was calculated adding the difference so calculated to the daily surface pressure of the NCEP/ NCAR reanalysis. Finally, with the daily surface pressure field, the daily wind field was calculated according with the regression equations in the 16 points located between the latitudes 32.5 º S and 40 º S and the longitudes 50 º W and 60 º W, all of them over the water domain of the model RPP-2D. The sea level rise was taken from the SRES A2 mean scenario (IPCC 2001) for every decade, but a higher value was also considered in the case of the 2070 decade, more consistent with present trends, Table 2.4. Decade Scenario Sea level rise (cm) Mean Mean High 40 Effect of the wind change Table 2.10: sea level rise scenarios considered The first experiments only included the wind field change to quantify its relative contribution. Figure 2.47 shows the mean level changes in future scenarios with respect to present (1990 decade). In both scenarios the level increases all over the Plata River. For the decade of 2030, a significant rise, greater than 5 cm, is only at the sources the river, in the neighborhoods of the Paraná Delta front. On the other hand, for 2070 decade, this or a greater rise takes place in the whole inner part of the River and the over the Uruguayan coast. a) 2030 b)

76 Fig. 2.47: Level rise for A2 mean scenario only forced by the change in the wind field as predicted by the RPP-2D model Effect of the whole change Future scenarios were developed including the combined effect of the wind change and mean sea level rise. The tributary changes would have a minor effect as will be seen later in the sensitivity study and were not included. Figure 4.13 shows the average annual and seasonal present levels, as well as the future levels for Buenos Aires in the mean A2 scenarios of the 2030 and 2070 decades. The annual mean rises are slightly larger than those corresponding to the mean sea level due to the effect of the wind change, in agreement with the experiments made with only wind change. 63

77 Fig. 2.48: Mean level rise in Buenos Aires for the A2 mean scenario as predicted by the RPP-2D model Figures 2.49 show the annual mean level rise for both decades in the mean A2 scenario. In both cases, the rise at every section of the river is higher on the Uruguayan coast. Figure 2.50 shows the mean seasonal rise for the same scenario in the 2070 decade, but for every season. a) 2030 b) 2070 Fig. 2.49: Level rise for the A2 mean scenario as predicted by the RPP-2D model 64

78 a) Summer b) Autumn c) Winter d) Spring Fig.2.50: Level rise for the A2 mean scenario as predicted by the RPP-2D model for the 2070 decade The frequency distributions of levels for both decades in the mean A2 scenario are shown in the figure Besides the shift expected in the mode, there is a growth in the dispersion of the frequencies to both higher and lower levels and a reduction of the mode frequency. 65

79 Fig. 2.51: Frequency distribution of levels for current conditions and for the A2 mean scenario as predicted by the RPP-2D model Fig shows the flood levels at Buenos Aires as a function of the return periods for current conditions as well as for the A2 mean scenario in the 2030 and 2070 decades. The typical storm surge was calibrated to attend the present levels corresponding at the return periods depicted in Fig Then, this typical storm was run forced by the corresponding 2030 and 2070 sea level scenarios. The increment of the water levels in each case is about the same as that of the sea level rise indicating that there are not important non linear effects. Therefore, the levels that correspond to each time of return were estimated for each scenario adding to the levels of the current scenarios the increase of the mean estimated level. As will be seen in the sensitivity studies (section 2.3.4), this is consistent with the fact that most of the level response would come from the sea level rise. Fig. 2.52: Return period for the maximum annual level in Buenos Aires for present and future mean A2 scenarios 66

80 2.3.2 Recurrent flood Maps For every recurrence time, the corresponding level along the coast can be obtained from the scale factor with respect to Buenos Aires (as illustrated in figure 2.46 for the 100 year return period). Then, for each site along the coast, the surrounding land area below this level is considered flooded. With this approach, the backwater effect on tributary rivers and brooks is ignored and therefore, the estimated flooded areas on their valleys are underestimated. The combination of the recurrence levels and the digital surface map of the coastal zone is performed in a GIS. Fig 4.17 shows the flooded areas for present conditions (1990 decade) and for different return periods as well as for the SRS A2 scenario in the 2030 decade. Figures 2.54 is similar, but for SRES A3 scenario at the 2070 decade. The case of the 2070 sea level rise of 0.4 m with respect to present will be referred to as 2070 max. The same information is presented in figures 2.55 to , but now for each return period (1, 5, 20, 50, and 100 years), making more clear the changes from present to future scenarios. The black and red colours in figure 2.53 that are indicative of the areas flooded at the respective return periods were superimposed on a satellite image where the huge metropolitan area of Buenos Aires and the smaller one of La Plata can be identified. Fig. 2.53: Flooded areas corresponding to return periods in years as indicated 1990 decade conditions (upper panel), 2030 decade in the SRS A2 scenario (lower panel) 67

81 68 Fig. 2.54: As in figure 2.53, but for the 2070 decade in the SRS A2 scenario

82 Fig. 2.55: Flooded areas with an expected mean recurrence of 1 year for different scenarios. Black area corresponds to the 1995 scenario. Other colours indicate the incremental areas added to the former scenario Fig. 2.56: As in figure 2.55, but for an expected mean recurrence of 5 years 69

83 Fig.257: As in Fig.2.55, but for an expected mean recurrence of 20 years Fig. 2.58: As in Fig.2.55, but for an expected mean recurrence of 50 years 70

84 71 Fig. 2.59: As in Fig.2.55, but for an expected mean recurrence of 100 years.

85 On the average, floods can be expected every year over a large coastal area in the south of the Metropolitan Buenos Aires, as well as in the coast of the Samborombón bay. The valleys of the Reconquista and Matanzas- Riachuelo rivers have also risk of floods, but only every 20 years. At the south of the Samborombón Bay there is a large area, well inland that surrounds the city of General Lavalle that presents return periods of floods between 50 and 100 years. In the future, the return periods of floods will become shorter in the mentioned valleys, which are now densely populated areas. The most striking change in terms of flooded area is on the south of the Samborombón Bay, as expected according to its altitude over mean sea level, fig Permanent flood As anticipated in section 2.2.8, the areas that will result with enduring floods in the studied scenarios are relatively small and only constrained to the Samborombón Bay, especially in its southern part. Fig 2.53 depicts the places that will suffer permanent flood in the 2070max scenario. Fig. 2.60: Area of enduring floods in the 2070 max scenario. Red as calculated from the GIS, pink as likely to be partially flooded. Since the terrain is rather flat, with lagoon and tide channels, the horizontal resolution of 1 Km may not be adequate to describe the real situation, where could be small marsh areas and isolated small islands. Therefore in the figure, the area that could be in such state is depicted in pink colour. At the same time, since the soil in this area is not composed of well consolidated elements, it is very likely that will be eroded in relatively few years. 72

86 2.3.4 Relative weight of the forcings of the Plata River level As explained above, the forcings of the dynamics of the Plata River are the astronomic tide, the input of the principal tributaries, the winds and the sea level. Once implemented, calibrated and verified, the hydrodynamic model was able to represent independently the effect of all these forcings, and therefore was used to assess the influence of changes in each of them on the mean level of the river. Thus, a sensitivity analysis to changes in the forcings was performed. The differences of mean levels of the water were calculated in 6 control stations. They are Martin García, Buenos Aires, La Plata and Colonia in the Inner River, and Montevideo and San Clemente in the Outer River. The baseline conditions were as follows: the main tributaries with their mean volume (the Uruguay River 5200 m 3 /s, the Paraná River 18,000 m 3 /s, amounting a total input of 23,200 m 3 /s); a wave of astronomic tide corresponding to summer (the month taken was February, 1997); a uniform wind field of 3 m/s and direction of 70 º clockwise respect to the north, that is to say, approximately from the E-NE. These conditions outline a typical mean summer scenario Tributary discharges Two conditions of streamflow input increase to the Plata River Plate was tested: 30,000 m3/s and m3/s, distributed (among) the tributaries in equal proportion as in the base condition. The case of 30,000 m 3 /s means an increase of 30 %, similar to many cases registered in the last three decades (See section 2.2.5). The results are shown in the figure a. The effect is almost imperceptible at Buenos Aires and downstream, producing a significant change only in Martin Garcia, with about 8 cm of level rise. The case of 75,000 m 3 /s corresponds to a case where the maximum observed at both rivers would occur simultaneously. As discussed in section 2.2.5, this is not an event that can be discarded to occur in the future. The figure b presents the results. This streamflow affects considerably the inner part of the Plata River: 65 cm at Martin Garcia and about 20 cm at Buenos Aires. But its effect is already almost imperceptible at Montevideo. a) b) Fig. 2.61: Mean water level increase due to: tributary discharge increase of (a) 30,000 m 3 /s and b) m 3 /s.mg: Martín García; BA: Buenos Aires; CO: Colonia; LP: la Plata; MO: Montevideo; SC: San Clemente 73

87 Winds A test scenario was run with an increase of 33 % in the intensity (4 m/s) and a change in the direction of the winds, which rotate eastward (reaching 90 º), compatible with the climatic trend observed for summer (See section 4.2.4). The results of the simulation are shown in the figure Increases of the level are observed in the inner Plata River, just as expected, reaching values of 4 cm in La Plata and Colonia and about 8 cm in Buenos Aires. Fig. 2.62: Change in the mean level due to an increment and rotation of wind to the east as described in the text Changes in the levels of the outer Plata River are also sensitive to the direction of the winds, changing from the decreases observed in figure 2.62 to almost not changes when the wind turns slightly more towards the North-East Mean sea level change A situation with an increase of 25 cm of the average level of the sea with respect the present conditions was tested. This value is representative of the order of magnitude of the expectable increase during the first part of the twenty first century. Figure 2.63 shows that the response all over the estuary is practically the same, with a very slight reduction of the rise, scarcely perceptible in Martin García, which is 3 cm lower than the sea level rise. Fig. 2.63: Change in the mean level due to an increment of mean sea level of 0.25 m Response comparison According to the expected possible variations in the forcings of the Plata River level, the mean sea level rise would be the prevailing mechanism of change of the mean level of the Plata River for the present 74

88 century. The mean sea level rise not only would be the more important in magnitude, but it will affect the whole estuary. It follows in importance the wind effect, which according to its likely expected changes would generate mean level rises of the order of 10 cm in the inner part of the river. Finally, the minor effect is that of the changes in the tributary discharges, which only for very exceptional and extreme events would produce important increases in Buenos Aires. More regular extreme events only would cause a few centimeters rise up to Martin García The wind influence in the levels of the Plata River In the preceding section was shown that changes in wind intensity and direction can modify the mean level of the Plata River. According to the model sensibility study, an augment of the easterly component is expected to rise the mean level of the inner Plata River, especially at the Argentine coast and decrease this level at the outer part of the estuary, especially in the Uruguayan coast. In the Argentine inner coast there are long records at Buenos Aires and at the Uruguayan outer coast, there is equally long records at Montevideo. In section 2.2.4, we have seen that the easterly component increased since the 1950 decade. Consistent with this wind trend, and with the sensibility results of the model, the level trend in Montevideo was lower than in Buenos Aires, section The difference in both trends was greater in the last three decades, as the level at Buenos Aires augmented 12 cm, while at Montevideo only increased 5 cm. The difference can be attributed to the wind change. Another indication of the influence of the wind field on the mean level of the Plata River is the different seasonal behavior of this level at Buenos Aires and Montevideo. The seasonal variability of the level at Montevideo is such that the maximum level is attained in autumn and the minimum in spring, and summer level is higher than winter. This cycle corresponds to the water density variation that accompanies the annual cycle of temperature in the sea. Because of its inertia, the sea reaches the warmest temperature, and consequently the maximum expansion, during the early autumn and the coldest temperature and minimum expansion in the early spring. On the other hand, in Buenos Aires, the maximum level is in summer and the minimum in winter, indicating that the wind effect overcomes the density effect in the summer/autumn and in the winter/spring parts of the year. Indeed, the easterly component of wind in the Plata River is considerably stronger in summer than in autumn, while the same happens with the westerly component in winter with respect to spring. The future evolution of the regional sea level pressure (SLP) fields was discussed in section The trend toward increasing (decreasing) predominance of the summer (winter) mode will affect the Rio de la Plata estuary wind field. The SLP meridional gradient is proportional to the eastern wind component (geostrophic relationship) or more realistically because of the friction effect to the southeasterly wind component. Table 2.1, shows the mean meridional SLP gradient across the RP estuary for different periods and model experiments. This gradient is different between models and between them and NCEP, but all have the same positive trend including the same decline in the nineties. This might imply that the water level of the RP estuary in its inner stretch has increased not only due to the sea level rise but also because of the rotation of the wind field, and that this effect could continue in the future. 75

89 Model/ Reanalysis NCEP HADCM CSIRO-Mk GFDL-R ECHAM4/OPYC Table 2.64: Average ( W) SLP difference (hpa) between 32.5 and 37.5 S 2.4 Conclusions The mean level rise of the Plata River will be a few centimetres greater than the sea level rise because of the wind rotation to the east that is already taken place. For this reason the level rise in the Uruguayan coast will be higher than in the Argentine coast and more important towards the interior of the River. Since, in general the coast in the Uruguayan margin is high; the only prejudice of this rise will be in the reduction of the shores, that however is an important asset because of the economic profile of Uruguay as a destination from tourists of the neighbouring countries. The areas with risk of enduring flood during this century in the Argentine coast of the Plata River are very small. The southern coast of the Samborombón Bay presents the large area with such risk that could be enhanced by the characteristics of the soil, which is composed of not well consolidated elements and therefore could be eroded in relatively few years. Other area of permanent flood risk is the front of the Paraná delta, which may become an area of social and economic vulnerability if it is occupied in the future. Therefore, the major impact of the Climate Change regarding coastal flooding will be in the increasing frequency of floods caused by storm surges. These floods can be expected now every year over a large and wide fringe in the south of the Great Buenos Aires, as well as in the coast of the Samborombón bay. The valleys of the Reconquista and Matanzas- Riachuelo rivers have also risk of these floods, but only every 20 years. In the future, the return periods of floods will become shorter everywhere, but especially in these valleys, which are densely populated areas. The most striking change in terms of land areas to become recurrently flooded will be on the south of the Samborombón bay. According to the expected changes in the forcings of the Plata River level, the mean sea level rise would be the prevailing mechanism of change during the present century, being the more important in magnitude and affecting the whole estuary. Wind changes would generate mean water level rises of a few centimeters in the inner part of the river. This rises can be matched by very exceptional and extreme discharges of the tributaries, but only for few days or eventually months. The wind is however, the most important forcing in causing the recurrent floods on the Plata coast. In addition, its annual cycle is also responsible for most of the seasonal changes in the mean River level and for the differences between the observed water level trends in Montevideo and Buenos Aires. 76

90 3 Socio-Economic Features 3.1 Activities Conducted We made a critical review of the concept of vulnerability, considering the need of defining indexes that reflect the conditions a priori of catastrophic events, which however condition the capacity of response and adaptation during and after these events. Then, it followed a delimitation of the area of the study. For current conditions, social data were taken from census, and for future scenarios, as a first approach, socio-economic conditions were considered as in present time. Though this a simplistic approach, the history of Argentina during the last century, indicates that once unthinkable socioeconomic scenarios had nonetheless taken place and there are few signs that the socioeconomic indicators may improve in the future. Demographic growth followed a more predictable path and therefore a simple hypothesis of a 1 % annual growth was adopted. The last step was to map the indicators and social indexes in a GIS 3.2 Description of Scientific Methods and Data Available demographic and social information was taken from the national census of Though there was a more recent census in 2001, most of the social variables were not yet processed at the time this activity was undertaken. The companies in charge of public services provided the technical data that permitted to assess the cost of floods in their facilities. In the case of real-estate property, the cost assessment was carried out according to the mean value of each zone. The socioeconomic conditions for future scenarios were considered equal to the present ones, which is clearly a very strong simplification. It was considered that it is practically impossible to make projections of these conditions to 30 or 80 years ahead in such changeable world and country. Anyhow, the results are indicative of the impacts that the Climatic Change would cause in the current social-economic conditions and are useful to show the trends and the principal aspects that would be necessary to attend Delimitation of study area To develop a social vulnerability index and its expression in a GIS context, it was necessary, as a first step, to choose the geographic area of study. The delimitation of the study region and the politicaladministrative units involved were done using two conditions: that the administrative units were located on the coastal zone of the Plata River, and that part of it was below 5 m above mean sea level. The 5 meter mark was based on the assumption of a maximum scenario of mean sea level rise in the year 2100 of about 1m and considering that the maximum tidal peak registered until now was near 4 m. It should be noted that the application of an exclusively physical-natural criterion in the delimitation of the study area for the characterization of social vulnerability was not possible for two reasons: i) When the littoral area is affected, the rest of the territory that is part of the political-administrative units involved will also suffer the socioeconomic effects of the phenomenon. Political decisions that may be proposed and eventually taken on the potentially affected area are largely circumscribed to the politicaladministrative units involved. ii) The socioeconomic and demographic information compiled from the National Censuses of Population and Housing (CNPyV), necessary for the characterization of social vulnerability is consolidated in political-administrative units (Municipalities/Districts) and it is available at smaller units (census fractions and radii) only for population data. Low areas bordering the Reconquista River in San Fernando County, and low areas in the floodplain of the Matanza-Riachuelo basin, located within Buenos Aires city were included, although they have no coast over the Plata River. The low areas of the continental sector of Tigre County, which clearly illustrate 77

91 the growth of gated urban polderized neighborhoods, were also included. Fig 3.1 illustrates the selected area. a) b ) Fig. 3.1: The study area, a) the north sector and b) the south sector Available demographic information Information of the political-administrative units and their corresponding census fractions and radii, according to the 1991 CNPyV, carried out by the National Institute for Statistics and Censuses (INDEC) 78

92 was processed in a GIS program ArcView 3.1 and interpolated to the same grid of 1Km2 used for the digital model of the topography. Figure 3.2 shows the population density. It is seen the maximum density over the City of Buenos Aires and the shape of the metropolitan area of Buenos Aires as well as the city of La Plata to the southeast of it. Fig. 3.2: Population density (hab/km 2 ) Critical review of the concept of vulnerability For this research, social vulnerability has been defined based on the conditions of the social group (social, economic, cultural, political dimensions), prior to the occurrence of the catastrophic event, in terms of its capacity to face it and recover from it. The social ensemble -all those who are subject to being potentially affected by a possible disaster - should be identified. The members of this group share certain features defined in terms of exposure (territorial and material aspects) but are however heterogeneous in terms of response capacity (economic, cultural, and political aspects). This ensemble is consequently heterogeneous. The differences within it must be taken into account when establishing priorities in a context of resource shortage. Some authors view heterogeneity in a dichotomous way, linking it to a poverty or non-poverty situation and consequently, to a situation of social inclusion or exclusion. Other authors identify a series of nuances and grades, in which multiple intermediate situations exist between both extremes. Who is to be included, or not, within the vulnerable group depends on the criterion to be applied. The first vision (dichotomous) considers a group and excludes another. But in reality, social vulnerability is multidimensional. Therefore, the nuances and grades that express this multi-dimensional condition should be retained in the analysis. 79

93 3.3 Social vulnerability index: Selected definition and indicators. An index of social vulnerability makes possible to identify situations of greater social vulnerability within a given group of units. Its scope and limitations are related to the objective of identifying units in which the process is apparently more intense. In this case, the index includes indicators related to the following aspects: a) demography b) living conditions of the population and c) structural production and consumption processes. The demographic sub index includes the following indicators: total population, population's density, index of potential dependence (children and elderly). The conditions of life sub index includes the following indicators: population s percentage of homes with unsatisfied basic needs (NBI), percentage of homes with women in charge, total rate of infantile mortality, and population s percentage without access to health services. The work, production, consumption sub index includes unemployment rate, aggregated gross product, registered cars rate (inhabitants / car rate), and percentage of workers without social benefits. These indicators were chosen on the basis of data availability for all the administrative units under survey. The data employed were those corresponding to the Buenos Aires province for the year 1991, the last census available to date for most of the indicators and administrative units. In all cases, a classification in five categories was made, using the system of natural breaks provided by the GIS. Then the data were plotted and classified in five categories, and the breaks were analyzed according to the curves. In most cases, the categories proposed by the GIS were accepted, while in some were adjusted, changing the (upper or lower) limit so as to obtain a more significant variation of the data, and consequently, the highest possible heterogeneity. Each sub index results from the sum of the values assigned to its four indicators. Finally, the values were grouped in four categories as depicted in table 3.1. Index Classes Demographic Sub index Conditions of life Sub index Productive Sub index Very low Low High Very high Social Vulnerability Index Table 3.1: The composition of the Social Vulnerability Index 3.4 Results The geographical distribution of the sub indexes can be seen in Table 3.2. Figure 3.3 presents also a mapping of the social vulnerability index. According with this social vulnerability index, E. Echeverria in the south of the Great Buenos Aires and General Sarmiento in the north are the two districts with higher social vulnerability. However, since they are districts that only have a minor exposure to floods, they are not highly vulnerable to them, although in the second case would be vulnerable in the future because of the River level rise. If this index were combined with physical conditions, Berazategui and Berisso would be the districts with higher social vulnerability to floods. On the other hand, General Lavalle, Magdalena y Tordillo, districts of definite rural profile rural, present the lowest indexes of social vulnerability. The only urban district with a comparable low index of social vulnerability is Vicente López, next to the city of Buenos Aires, a residential area of predominantly high income population. 80

94 Departments Demographic Sub index Conditions of life Sub index Productive Sub index Esteban Echeverria General Sarmiento Berazategui Berisso Castelli Ensenada General San Martin La Costa La Matanza Lanas Lomas de Zamora Quilmas San Fernando Tigre Avellaneda Capital Federal Chascomus La Plata Moron San Isidro Tres de Febrero Dolores Maipú General Lavalle Magdalena Vicente Lopez Tordillo Social Vulnerability Index Table 3.2: Social Vulnerability Index of the administrative units of the study area. References: 1=very low; 2= low; 3=middle; 4=high 81

95 Fig. 3.3: Index of social vulnerability 3.5 Conclusions The coastal area of the Plata River under risk of potential floods during this century is heterogeneous in terms of social vulnerability. E. Echeverria and General Sarmiento are the two districts with higher social vulnerability. However, since they are districts that only have a minor exposure to floods, they are not highly vulnerable to them. The other two districts that follow in social vulnerability, Berazategui and Berisso, are at the same time highly exposed to recurrent floods as will be seen in the next chapter. 82

96 4 Impacts and Vulnerability 4.1 Activities Conducted Assessment of vulnerability followed a geographical approach integrating physical and social information in a GIS. The GIS was used to estimate the population affected and the public service infrastructure and real estate property damage for different return periods of flooding. Based on the areas flooded at given return periods, it was calculated the social and the economic impacts for current conditions as well as for future climate and sea level scenarios. The first was done through the construction of an index of social vulnerability to recurrent floods combining an exposition to the floods index with the social vulnerability index. Economic impact followed two approaches, one that produced an inventory of the facilities on the area that will become exposed to flood during this century, and another that assessed the costs of the damaged caused by the recurrent floods in the public services facilities along the coast and in the real estate property, 4.2 Description of Scientific Methods and Data The methods and data used were different for the assessment of the social vulnerability to recurrent floods and for the economic damages of them Socio economic vulnerability to recurrent floods An index of exposure to floods was calculated with the minimum return period of flood that corresponds to every cell of 1 Km2. It was calculated as approximately the inverse of the minimum return period (MPR). Its formulation is 20/ (MPR) +1. The idea behind this formulation is that although the exposition index should somehow be inverse to MPR, the implications of a flood that happens every 20 years is larger than one twentieth of the one that takes place every year since it affects zones where the phenomenon is less expected and generates negative expectations that are already incorporated in the other case. On the other hand, there is not much differences in the perception of the flood risk, if this has a recurrence of 20, 50 or 100 years, since in any of these cases, some precaution has to be considered. For this reason, the index reflects little changes for areas with expected recurrence values greater than 20 years. Finally, an index of social vulnerability to recurrent floods was developed combining the index of social vulnerability with the one of exposure to the floods through the product of both. Later the obtained indexes were normalized in order to rank them from zero to 100 in the area of the study. The number of persons affected in their households by a flood that happens with a certain return period was estimated according with the spatial distribution of the population on the area that would be flooded with that time of recurrence Exposure of facilities to recurrent flooding The exposition of the public buildings were estimated by a survey performed in 27 administrative units, the City of Buenos Aires and in 26 municipalities, which has all or part of its territory below 5 m over the mean sea level. The survey was conducted by internet, mail, telephone and fax. In some rural districts, it was necessary personal interviews. With the obtained information, a preliminary database was constructed that was 83

97 checked and adjusted with personal interviews at each of the 28 administrative offices of the studied region Current and future damage costs An identification of the most relevant infrastructure was undertaken to assess the first order effects of the Plata level rise. The most relevant infrastructure is that of the public services and the real estate property Public services infrastructure Water supply The water supply service in the metropolitan area of Buenos Aires is operated by the private company Aguas Argentinas, who has two water supply plants, the San Martín Plant, located in the city of Buenos Aires and the Belgrano Plant, located in the southern area of the Great Buenos Aires. The helpful and adverse effects of the Rio de la Plata level variations in the operation of both plants are summarized in Table 2.5. LOW LEVELS Water supply plants Helpful Adverse Low performance of raw water pumps,,,,,,,,,,,,,,,,,,,, Coastal pollution Rise in chemical doses HIGH LEVELS,,,,,,,,,,,,,,,,,,,, Hydraulic limitations in drainage... Coastal pollution after flood - Rise in chemical doses Save pumping energy,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Table 4.1: Effects of the Plata River level variation over water supply plants In case of low levels, the water pumps show a lower hydraulic yield. At the San Martín Plant, the main elevating pumps are the most sensitive to low waters, being the first ones to go out of service. Excess vibrations are also a common difficulty. These problems lead to higher energy consumption and lower water elevation volumes. Regarding the quality of water during low waters, there appear higher concentrations of pollutants (organic matter, chlorides, ammonium, conductivity, and alkalinity). This requires an increase of the chemical doses. In case of high levels, a saving in water elevation energy is obtained, but hydraulic difficulties appear at the drainages, as conduits start to work under pressure conditions instead of as an open channel. As for the quality of water, the main difficulty associated with a significant and long lasting flood is the blockage effect exerted on the drainages, which prevents its normal discharge. The problem manifest when the flood ends, as the retained drainage volume discharges abruptly, transporting its high load of pollutants into the Plata River. The resultant worsening of the water quality, which lasts several days, implies the increase of chemical doses. For scenarios of mean level rise in the Plata River, the most significant effect, from the economical point of view, is the saving in water elevation energy. Hence, curves that represent this saving, for different 84

98 water level increases, were determined for each plant based on average daily volume, variation of pumping height, pumping yield, and unit cost Sewer Outlets The sewerage system in the metropolitan area of Buenos Aires is also operated by the private company Aguas Argentinas. There are two elevating stations, Wilde and Boca-Barracas, and two treatment plants for sewer liquids, Sudoeste and Norte. The probable effects of the increase in the Plata River level on these plants are summarized in Table 4.2. LOW LEVELS Elevating stations Treatment plants Helpful Adverse Helpful Adverse No effects HIGH LEVELS Increase of mean pumping energy,,,,,,,,,,,,,,,, Increase of pumping energy during extreme events,,,,,,,,,,,,,,,, Increase of streamflow,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Overflow of sewer liquid during electrical failures,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, Table 4.2: Effects of the Plata River level on the sewerage plants For the Plata River levels higher than IGM 1.97 m, the consumption of electric power increases due to higher elevation levels, especially at Wilde station. The cost of the extra pumping due to the elevation of the Plata River level has been determined based on the mean daily volume, the pumping level, the equipments performances, and the unit cost. As an example, for the Boca-Barracas station an increase of 0.20 m in water level would mean an increment of 2.5 % in the pumping cost. In the case of the treatment plants, the energy consumption would not change Electric power production and transport facilities The existing utilities in the area that could be expose to floods are Central Costanera, Central Puerto and Central Dock Sud, Fig Their inputs for the electric power generation are basically natural gas and occasionally fuel oil. Central Costanera is located at the Matanza-Riachuelo outlet. Central Puerto is within the Buenos Aires harbor, and consists of two plants Puerto Nuevo and Nuevo Puerto. Central Dock Sud is to the south of Buenos Aires City; their total installed capacity is MW. 85

99 Fig. 4.1: Location of power plants Central Dock Sud has eight gas fed turbo generators, with a total power of 211 Mw. The increase of pumping energy costs for Central Dock Sud, due to the Plata level rise was determined based on the available data. The corresponding costs for Central Costanera and Central Puerto were estimated according to their relative rate of discharge with respect to Central Dock Sud. Central Dock Sud Helpful Adverse LOW LEVELS No effects HIGH LEVELS Increase of mean pumping energy,,,,,,,,,,,,,,,, Increase of pumping energy during extreme events,,,,,,,,,,,,,,,, Table 4.3: Effects of the Plata level on the on Central Dock Sud utility The electrical service supply in the metropolitan area of Buenos Aires is provided by the private companies Edesur and Edenor. Fig shows the location of the Edesur transformation stations that could be exposed to floods as River level rise. They were divided into different groups, according to their topographic level, starting from the most vulnerable to floods. It has been assumed the existence of transformation stations from middle to low voltage, with a densification of one every eight blocks. Typical costs of these installations were estimated together with their percentage of damage as a function of the flood height. Based on that information, curves of cost due to damages were obtained as a function of the water level rise. 86

100 The curve of cost corresponding to Edenor facilities was obtained from the previous one, by assuming to be proportional to the number of transformation stations that each company has at the area with risk flood. Fig. 4.2: Location of the Edesur transformation stations Roads The likely effects on the road system because of the Plata River level rise are: Flooding by superficial waters. The associated damages are proportional to the water depth, its permanence and the density of traffic during the event. Rising of the water table. This weakens the base and sub-base, and slows down the runoff. The road system in the metropolitan area of Buenos Aires is formed by highways, fast lanes, and the urban road network, Fig. 4.3 shows those that could be exposed to floods. The main highway is Buenos Aires-La Plata Highway, shown in Fig About 13 km are below IGM 5 m, of which 4 km are below IGM 4 m, which means that will be increasingly vulnerable to floods. The fast ways are Acceso Sudeste, Lugones Ave and Cantilo Ave. The Acceso Sudeste is the most exposed to future Plata River floods. Damage curves were obtained for the combination of highways and fast ways for every flood level, estimating the length of flooded roads and the repairing costs. 87

101 Fig. 4.3: Location of highways and fast lanes Fig.4.4: Buenos Aires La Plata Highway To estimate the damage of floods to the urban road infrastructure, the following zones were defined: Buenos Aires City, the Matanza River basin, the coastal fringe between Avellaneda and Berazategui and 88

102 the Reconquista River basin. For each of them, the following variables were quantified: percentage of urbanization, density of kilometers of road per unit area, percentage of roads or streets made of concrete, asphalt and bare soil. Based on these parameters, the flooded road lengths were calculated for different River levels, as well as their repairing cost, thus finding the damage curve as a function of River level Railway system The metropolitan area of Buenos Aires has 850 km of railways, most of them of double track. In the area below 5 m over mean sea level, there are about 115 km of railroads. Damages related to the River level rise are caused by the elevation of the water table, the presence of superficial waters that weakens the embankments, and, for the higher levels, damage of the structure itself. The corresponding curve of damages was obtained based on an indicative value for the cost of the railway kilometer, the length of railway at different altitude bands and the percentage of damage according to the water depth Building infrastructure Housing The damages on the building infrastructure are the most significant from the economic and social point of view. Because, the areas exposed to floods have different real estate features, some level of desegregation was necessary to make an evaluation of the costs of floods. Therefore, a zoning of the vulnerable area was performed, resulting in the identification of nine units, Fig For each zone, several hypotheses were made to estimate housing damages as a function of the following parameters: costs of housings per unit area, percentages of area flooded by each River level, and population density. The damages of houses were estimated as a percentage of its real estate value. When the flood level reaches less than 0.5 m at the houses, this percentage varies from 12 to 16 % according with the zoning. If the water level reaches more than 0.5 m, then this percentage augments to 25% or 30 % according to the zoning. Based on these assumptions and the mentioned parameters, a curve of damages was built, Fig

103 Fig.4.5: Zoning of the vulnerable building infrastructure ,50 1,00 1,50 2,00 2,50 3,00 3,50 Incremento nivel Río de la Plata (m) Fig. 4.6: Curve of damages in $ (pesos) to the building infrastructure as a function of the Plata River level rise in m (1 $ = 0.33 US$) Public buildings For each zone, the damages to health, safety, and education buildings were calculated according to the number of these buildings that was estimated according to the population density. Hence, the damages for each River level were estimated based on the percentage of area affected, as for the previous case Furniture Furniture losses were estimated as a percentage of the damage to building infrastructure. Then, its curve depicting the cost as a function of the River level was obtained from the corresponding housing damages curve Economic quantification of flood damages As explained in section 2.3.3, the rise in the mean level of the Plata River will not produce any significant permanent flood in the metropolitan area of Buenos Aires. Hence, the impact on the infrastructure will be linked to the greater duration and frequency of the recurrent floods associated to storm events. The costs were estimated in Argentine pesos, which during the time of the present study was stabilized at the exchange rate of one American dollar (1 US$) for three Argentine pesos (3 $) Curve of total damages The curve of total damages was obtained as the sum the individual curves of damages mentioned in section This is shown in Fig. 4.9 for the range of level rise in the Plata River going from 0.25 to 3 m. 90

104 Up to approximately 2.25 m there is an exponential growth, which turns into a quasi-linear response for higher level rises ,5 1 1,5 2 2,5 3 3,5 Fig. 4.7: Curve of total damages in $ (pesos) to the infrastructure as a function of the Plata River level rise in m. (1 $ = 0.33 US$) Frequency of water levels Based on recorded hourly data for the 1990 decade at Buenos Aires, the curves of frequency of occurrence and of time over each level for the Plata River were built. Using model RPP-2D, the same curves were obtained for scenarios with 0.50 m and 1 m mean sea level rise. These results are shown in Fig. 4.8 and Fig. 4.9, referred to MOP zero (0.56 m below the IGM zero). Curvas de frecuencia de niveles del Río de la Plata en Buenos Aires 0,200 0,180 0,160 0,140 0,120 Frecuencia 0,100 0,080 0,060 0,040 0,020 0,000-2,00-1,00 0,00 1,00 2,00 3,00 4,00 5,00 Nivel MOP (m) Condición actual Nivel mar m Nivel mar más 1.00 m 91

105 Fig. 4.8: Frequency of occurrence of the Plata River level at Buenos Aires for current conditions (1990 decade) in blue, for a scenario of men sea level rise of 0.50 m in pink and for a scenario of mean sea level rise of 1m in yellow. Duración de niveles del Río de la Plata 1,00 0,90 0,80 0,70 Frecuencia de excedencia 0,60 0,50 0,40 0,30 0,20 0,10 0,00-3,00-2,00-1,00 0,00 1,00 2,00 3,00 4,00 5,00 Nivel MOP (m) Fig. 4.9: Frequency of time over each Plata Presente River level at m Buenos + 1 Aires m for current conditions (1990 decade) in pink, for a scenario of men sea level rise of 0.50 m in light blue and for a scenario of mean sea level rise of 1m in orange Curvas de recurrencia de niveles extremos en el Río de la Plata en Buenos Aires 6,00 5,50 5,00 4,50 Nivel MOP (m) 4,00 3,50 3,00 2,50 2, Período de retorno (años) Nivel Actual Escenario + 0,50 m nivel mar Escenario + 1,00 m nivel mar Fig. 4.10: Recurrence period in years (abscise) for extreme high levels in m in the ordinate (MOP level) in the Plata River for current conditions (1990 decade) in blue; for a scenario of men sea level rise of 0.50 m in pink and for a scenario of mean sea level rise of 1m in orange 92

106 To complement the information provided by figures 4.9 and 4.9, figure 2.12 present the curves for the recurrence period of annual maxima peak water levels associated to storm surges. Again, the curve for current conditions is based on recorded data, while the inferences for 0.50 and 1 m mean sea level increase were obtained with model RPP-2D Current and future damages Combining the obtained information, it was assessed the annual damage for the current condition and for two future scenarios, with 0.50 and 1.00 m mean sea level rise. Previously, it was necessary to merge the results from the recurrence period of extreme values and from the frequency of occurrence of ordinary water levels. For the case of extreme events we have: CA = CU * P Where CU: Cost per event; CA: Cost per year; P = 1/T: Number of events per year; T: Return period On the other hand, for the case of ordinary water levels we have: f: frequency of occurrence associated to a water level range then D = 365 * f is the number of days per year with water levels within that range. To link this continuous analysis to an event analysis like the previous one, the parameter D was interpreted as a succession of events of mean duration d. Hence, P = D / d = 365 * f / d To make the transition smooth, the event which occurs only one day per year, i.e., that with f = 1/365, should have P = 1. Hence, it was taken d = 1 day. The annual damage, for each level of the Plata River, is the sum of the damages produced by all the events occurring during one year. In turn, the annual damage associated to an event of a given water level, is the unit cost associated to that event times the number of times it happens during the year. In this way, it was obtained the curve shown in Fig The low annual damage observed for the three curves for the highest levels is due to the very low probability of occurrence of these levels. The total annual damages can be computed through integration of the curves of figure Results

107 Fig. 4.11: Total damage in $ (pesos) to the infrastructure as a function of the Plata River level in m (MOP level) for current conditions (1990 decade) in blue; for a scenario of men sea level rise of 0.50 m in pink and for a scenario of mean sea level rise of 1m in orange. (1 $ = 0.33 US$) Socio economic vulnerability to recurrent floods The areas that are more exposed to storm surges are the coast of the Great Buenos Aires, to the south east of the city, a coastal fringe in the district of Tigre in the extreme north of the Great Buenos Aires and all the coast of the Samborombón Bay Fig. 4.12a. Fig. 4.12: Exposure index to recurrent floods. Current conditions in the upper panel, and 2030 scenario in the lower one 94

108 95 Fig. 4.13: As figure 4.12, but for 2070 scenarios. SRES A2 in the upper panel and in the case of a sea level rise of 0.40 m in the lower one (2070 max )

109 Changes in the index of exposure to floods, in the RP coast of the city of Buenos Aires and in the north of it, would not be considerable even in the scenario of 2070max. On the contrary, to the south of the city, and in the valleys of the Matanzas Riachuelo and Reconquista rivers, populated predominantly with low income and high structural social vulnerability people, will be a rise of exposure to recurrent floods Fig. 4.12b and The figure 4.14 shows the index of social vulnerability to recurrent floods. The qualitative geographic pattern is similar to the one of the exposure index because of the coincidence of the areas of maximum exposure with those of high structural social vulnerability. However, there are some differences like in the north the highly exposed coast of the district of Tigre that only has in the average a medium structural social vulnerability. The areas with maximum social vulnerability to floods are to the south of this district in the Reconquista valley as well as in the south of the Great Buenos Aires. 96

110 Fig. 4.14: Index of social vulnerability to recurrent floods. Current conditions Fig. 4.15: Differences between the indexes of social vulnerability to recurrent floods. Scenario 2030 minus present in the upper panel and 2070 max minus present in the lower panel 97

111 Changes in the index of social vulnerability to floods in the 2030 and in both scenarios in the 2070 worsen the situation of the already more vulnerable areas along the valleys of the Reconquista and Matanzas- Riachuelo rivers and in the south of the Great Buenos Aires in zones relatively far from the coast and where because of that cannot be expected an expansion of affluent population on gated communities, Fig If it is assumed that that population density and distribution will not have considerable changes, in the scenario of maximum sea level rise in the 2070 decade, the people living in the area with flood risk with a return period of 100 years will be about , almost doubling the present population with such risk. The relative increment of affected population is even larger for recurrence time of 1 to 5 years, which will triple in the 2070max scenario, Table These figures were calculated without considering the very likely growth of population. With a modest 1% annual growth of the population during the next 70 years, maintaining the present geographical distribution, the number of people affected for each return period in the year 2070 would double the values of. This means that the population with risk of some flood, recurrence every 100 years will amount to about 1,700,000. Return Period (years) / / / Table 4.16: Present population living in areas that are, or will be, flooded under different scenarios 98

112 Exposure of facilities to recurrent flooding Through a survey conducted in 27 administrative units, it was identified the facilities that are sited in areas with altitude below 5m over mean sea level. As it was explained before, these areas have a potential risk of flooding sometime during the present century. Table shows the number of facilities, disaggregated according to their type. Almost all public offices buildings correspond to municipality or provincial facilities. The districts that are more exposed with respect to these buildings are Berisso and Ensenada, because all of them are in areas below 5m over mean sea level, followed by the districts of Tigre (72 %), San Fernando (53%) and Avellaneda (47 %). In the rest of the districts, most of the public offices are over the 5m level (75% or more). The exposure of the heath centres has similar distribution. Ensenada and Berisso have the greater percentaje of health centres below the 5 m level, 89 % and 94 % respectively, followed by San Fernando (65%) and Avellaneda (51%). Public offices 125 Welfare 17 Heath centres 205 Education 928 Police and security 92 Transport 41 Industries Recreation 306 Table 4.17: Amount of facilities below 5m over mean sea level in the Argentine coast of the Plata River In Ensenada, Berisso and Avellaneda there is a large percentage of schools and other buildings devoted to education that are in areas below the 5 m level, that is 96, 82 and 67 respectively. However, the city of Buenos Aires has the greater number of these buildings in such situation, namely 182, but its percentage in the city is only 10 %. With respect to the industries, again Ensenada and Berisso have the totality of them below the 5m level. Other districts with considerable exposure are Avellaneda (82%), San Fernando (61%) and Lanús (51%). In absolute values, San Fernando has the greater number of exposed industries, followed by Lanús and Quilmes. It can be concluded that Ensenada and Berisso appears as the most vulnerable districts because most of the public offices, health centres and industries could be flooded sometime this century. Other areas with important vulnerability are Avellaneda and San Fernando Costs of Current and future damages The frequency of events which produce damages increases with the sea level rise as shown in Fig

113 1,00 0,90 0,80 0,70 Frecuencia de eventos 0,60 0,50 0,40 0,30 0,20 0,10 - Actual 0,50 m 1,00 m Condición Fig. 4.18: Frequency of events producing damages The total annual damages can be computed through integration of the curves of the curves of figure 4.2. They are shown for current conditions and for the scenarios of 0.5 m and 1m of sea level rise in figure Actual 0,50 m 1,00 m Condición Fig. 4.19: Total annual damages for current conditions (1990 decade) in $ (pesos) and for scenarios of mean sea level rise of 0.50 m and 1m. (1 $ = 0.33 US$) Future scenarios Scenarios corresponding to 2030, 2070 and 2100 were considered. For each year, two values for sea level rise were considered. One value corresponds to the SRES A2 scenario as reported by IPCC The other, correspond to a scenario of maximum rise that considers the uncertainty of the socio economic scenarios. In the case of 2070, this scenario gives a rise of about 0.40 m. Since this rise was also considered for the A2 in the year 2100, it was consider an even more extreme rise of 0.50 m. Model RPP-2D provided the corresponding water level rise in Buenos Aires, which are shown in the following table. Year Mean sea level rise (m) Mean level rise at Buenos Aires (m) 100

114 A2 Extreme high A2 Extreme high Table 4.3: Mean sea level rise and mean River rise at Buenos Aires in m Damages considering different scenarios of growth in the infrastructure Table 4.4 present the damages associated to the above described scenarios, both if no change of the infrastructure value is considered, which constitutes a lower bound to expected damages and for different growth rates of the infrastructure. This growth and of the associated value can be expected due to the following factors: population growth, increase of urbanization, increase and/or improvement of public infrastructure and evolution of the Gross Product of the area The evolution of each components of the infrastructure (railways, roads, water supply plants, electrical supply, housing, etc.) can obviously have its own characteristics, in response to the needs of different sectors. Instead of analyzing the probable evolution of each component, it was simply assumed different rates of increase of the total infrastructure value. Thus, the total damage was calculated multiplying the damages corresponding to the present infrastructure by the percentage of increase in the future. A2 scenario Year Current conditions Mean level rise at Buenos Aires (m) Mean level at Buenos Aires (m at MOP) Infrastructure growth rate % Extreme rise scenario Current conditions Mean level rise at Buenos Aires (m) Mean level at Buenos Aires (m at MOP) Infrastructure growth rate % Table 4.4: damages in M of U$S associated to different mean sea level rise and different scenarios of growth in the infrastructure Annual growth rates of 0.5 %, 1 % and 1.5 % of the value of infrastructure were considered to estimate the probable damages for the period in both A2 and the extreme sea level rise scenarios. Table shows the results, which are illustrated also in the figures 4.4, and

115 Millones Fig. 4.20: Annual damages in millions of $ (pesos) as a function of time in a scenario of no change of the infrastructure with time. (1 $ = 0.33 US$). Climate A2 scenario in blue, extreme scenario in pink 1.000,00 900,00 800,00 700,00 600,00 500,00 400,00 300,00 200,00 100, Año Fig. 4.21: Annual costs of damages in millions of $ (pesos) up to year 2100 for the A2 climate scenario of mean sea level rise considering different annual growth rate in the infrastructure. Light blue 0%; blue 0.5 %; pink 1.0 % and orange 1.5 %.(1 $ = 0.33 US$). 102

116 2.500, , , ,00 500, Año Fig. 4.22: Annual costs of damages in millions of $ (pesos) up to year 2100 for the extreme mean sea level rise scenario considering different annual growth rate in the infrastructure. Light blue 0%; blue 0.5 %; pink 1.0 % and orange 1.5 %.(1 $ = 0.33 US$). The costs of damages in the A2 scenario are almost linear with time, except for the case of the growth of the infrastructure. In the case of the extreme mean sea level rise scenario, there is an abrupt increase in the cost trend after If the growth rate of the infrastructure is assumed to have at least a minimum annual growth of the order of 0.5 % or more, the accumulated costs in the second half of the century will range from 5 to 15 billion dollars depending on the climate scenario and the growth rate of the infrastructure. 4.4 Conclusions The areas with population vulnerable to storm surge floods are the coast of the Great Buenos Aires to the southeast of the city, particularly in the districts of Ensenada and Berisso, and to the north, in part of the districts of San Fernando and Tigre. In Tigre there is a mixture of vulnerable population of low socialeconomic level and closed middle-high class neighborhoods. Other area with great exposure to floods is part of the district of Avellaneda, close and south of the city. The neighborhood of La Boca in the city was exposed to floods in the past, but as will be seen in chapter 7, the defenses built to contain the floods from the Plata and its tributary, the Riachuelo have reduced the flood risks. According to the A2 SRES scenario, the social vulnerability to floods will become worst during this century along the valleys of the Reconquista and Matanzas-Riachuelo rivers and in the south of the Great Buenos Aires in zones relatively far from the coast. These are already areas oh high social vulnerability that will be worsened by the increasing recurrence and spatial reach of the storm surges. On the other hand, it is expected that the coastal zone of the Buenos Aires City and the districts located to the north of it will not suffer important changes in the danger to flood exposition. In a scenario of 0.5 m sea level rise with a modest 1 % annual rate increases in the population without considerable changes in its distribution and without new defences, the population with risk of some flood (recurrence every 100 years) in the 2070 decade will amount to about 1,700,000, more than three times the present population in such conditions. Those with risk of flood every year will be about 230,000, six times the population that suffer now such recurrence. These figures indicate that, although slow, the river level rise will create severe socio economic problems if not early planning is undertaken. The mean cost of the current damages to the coastal infrastructure was estimated in the order of the 24 millions of American dollars per year. Most of this cost originates in real-estate damages. On the other 103

117 hand, damages were only estimated for the zones of greater value in the city of Buenos Aires and the nearby districts, not including the zones with lower real-estate value in the flood valley of the Reconquista River, and in districts of Tigre and San Fernando, or their gated neighbourhoods. Therefore, the calculated values represent a minimum estimate of the total cost. This cost of the recurrent floods will increase during this century because of both, the sea level rise and consequently the greater frequency and reach of the floods, and the growth of the infrastructure and its value. If a moderate growth rate of the infrastructure is assumed, the accumulated costs in the second half of the century will range from 5 to 15 billion dollars depending on the actual climate scenario and growth rate of the infrastructure. 104

118 5 Adaptation 5.1 Activities Conducted Study cases were performed with the intention to gain insight on the different dimensions (socioeconomic, institutional and cultural) of the responses to recurrent floods in the past and present. The selected cases were La Boca neighbourhood, in the City of Buenos Aires, and the Avellaneda Municipality, in the Metropolitan Area of Buenos Aires. In both cases, the social vulnerability index has low values, section 5.3, but they were selected because their long tradition in dealing with recurrent floods can give indications on how the population of Buenos Aires may behave when a major adaptation response will be required. In both cases, the Project conducted interviews with officials in charge of the institutions that deal with some of the aspects of floods (planning, disaster response, etc) as well as with key stakeholders. It was also analyzed the past and current trends of occupation of lands subject to recurrent floods in other areas, in view of the growing tendency to build gated communities on the shore of the Plata or of its small tributaries. 5.2 Description of Scientific Methods and Data In the past, La Boca neighbourhood has suffered periodic floods due to storm surges in the Plata River and /or to intense rainfalls. Part of the Avellaneda Municipality, separated from La Boca by the Riachuelo, also suffers floods due to the same causes. The recurrent floods and, because of them, their relatively small real estate value turn these neighbourhoods in marginal areas, even considered dangerous in the urban social imaginary. In both case studies, the flood danger is seen as the most important among of the natural threats. They originate either in intense rainfall, which provoke the inundation of the streets or in storm surges in the Plata River, known as sudestadas. When both phenomenon occur simultaneously (section 2.2.3), the sudestadas block the normal drainage and worsen the flood caused by the intense rainfall. In those cases, polluted waters add nuisances to the already stressed population. In the neighbourhood of La Boca, for example, the contaminated waters of the Riachuelo sometimes spring to surface through the storm outlets of the streets and drainpipes, reaching in the lowest sectors more than 1.50 meters height. The capacity of response has to do, among other aspects, with the knowledge, the values, the perceptions, etc. that the civil society and the government institutions have of their potential dangers. These aspects were study in both cases with the same general methodology, but with different specific approaches according to the nature of the problems and the sources of information. Past and current trends on the use of low lands subject to floods in other metropolitan areas of the Great Buenos Aires were discussed in connection with their implications on present and future adaptation La Boca neighbourhood La Boca is one of the oldest neighbourhoods of the City of Buenos Aires. It is located at the southeast of the city on the low lands of the flood valley of the Riachuelo. The historic chronicles describe La Boca as a low, swamp zone, with lagoons and tall grass. This morphology disappeared under the asphalt, though the area continued to be exposed to floods. This neighbourhood was considered to be marginal from its beginnings. With regard to some indicators that reflect the conditions of life in the neighbourhood, according to information of 1991, La Boca has the highest percentage of tenants of the Federal Capital (44 % of the homes). The homes with critical overcrowding (more than 3 persons per room) were 5 %, whereas the average for Buenos Aires city is 2 %. Its high exposition to floods and pollution from the Riachuelo has kept this neighbourhood as 105

119 marginal area. However, at the beginning of the 1990 decade there were some changes aiming to improve the conditions of life in this neighbourhood. By the end of 1993 (year in which occurred ten floods), it was announced the construction of a coastal defence to avoid the flooding from the Riachuelo waters. The project included the reconstruction of the existing net of rainfall outlets. The coastal defence was inaugurated in 1998 whereas the new rainfall outlet network has not been yet finished. Within the context of this defence construction, the floods due to sudestadas seem to be "forgotten", which leads to a reconfiguration of the situation of risk in the neighbourhood. The defences generated a feeling of confidence, which may be a disadvantage in the future, if the sea level rise would make this defence insufficient. To identify floods and their impacts, the DesInventar data base of journal information on catastrophes, developed by the Network of Social Studies of Latin America was used. This consult was complemented with own searches in local newspapers. Both sources, permitted to collect basic information of the critical moments of floods, the impacts on the neighbourhood (persons, goods and services affected), the type of actions started from the government (local, national) and from the civil society in order to respond to the emergency, and the actions of prevention that were implemented. The approach for the diagnosis of the cultural adaptation was based on the identification of some key stakeholder that was interviewed, providing information that complemented other sources. All specific bibliography and documents related to floods in La Boca and in the city of Buenos Aires were consulted to trace institutional responses. The institutions acting in the different phases of the disaster (prevention, response, rehabilitation) during the catastrophic floods of the decade of 1990 that affected La Boca neighbourhood were identified as well as the direct and indirect actions, both of structural and non structural type Avellaneda Municipality The rapid process of urbanization provoked a fragmentation of the Avellaneda Municipality in neighbourhoods and slums, which were planned in agreement with the interests and possibilities of the owners and the real-estate companies who advanced on low and easily flooded areas. In some cases, the land was refilled to raise the level of some of areas. These types of interventions were carried out by the neighbours, as well as by economic groups without any regulation. Later, some infrastructure planned to modify the original runoff of the district was built with public funding, including three important channels that solved some of the flood problems in certain areas. As in La Boca, the pollution from the Matanza-Riachuelo River amplifies the harmful effect of the floods. In addition, there are overflows of small streams and brooks heavily contaminated. The adaptation responses were analyzed around three selected dimensions: socio-economic, institutional and cultural. The socioeconomic information was taken from the national censuses of population of 1991 and 2001, the national economic census of 1994 and the statistical Buenos Aires yearbook of For the institutional responses, regulations associated with the management of the disasters, especially floods were surveyed and analyzed. In addition, interviews with the competent institutions were realized. The institutions involved in disaster activities are specified in a law of the province of Buenos Aires, who determines the integration of the Civil Defence system at the county level. The more active institutions in the management floods in the case of Avellaneda were the Civil Municipal Defence, the Voluntary Firemen of Avellaneda, the Marine Police and Red Cross Argentina. These institutions were interviewed with specific guidelines. The questions were concerning the following aspects: Description of general aspects of the institution; Actions performed by the institutions before, during and after the occurrence of the disaster; The form in which they are articulated with other institutions before, during and after the occurrence of the disaster; 106

120 Their opinions with regard to other institutions, The aspects that they would improve to optimize the management of the flood-related disasters. The interviews had individual character and the executives of the four mentioned institutions were interviewed during November and December 2003 in their respective offices. Results of their views on the role of the institutions were compared with the sensu stricto duties that appears in the regulations ("the must be ") as well as with the effective form in which they are accomplished by the institutions. To approach the cultural responses, primary sources of information were analyzed, obtained through interviews to neighbours of the floodable areas. The interviews to these neighbours were made at their homes or at community centres where they develop their activities. Some of them were involved in the neighbourhoods community centres; others were store-keepers or just neighbours with an extensive history in the place. Among the aspects that were surveyed from the people interviewed, are the following: Their history as neighbours of the County; Experiences concerning the floods; Strategies elaborated to mitigate the harmful effects; Form of organization among the neighbours as a result of the floods; The way of perceiving other affected neighbours; Total or partial solutions that they know or would take in relation to the problem of floods; Social representations concerning floods. 5.3 Results La Boca neighbourhood Most of the measures to attenuate impacts of floods come from the governmental sector. The management involve the phases of prevention, response and rehabilitation. Table 5.1 shows a synthesis of the institutions involved, according to the corresponding phase. Table 5.2 shows actions related to the floods, considering the type of measure and its state of execution. Government of the City of Buenos Aires Prevention Response Rehabilitation Department of Logistics and Emergencies, Direction of Social Emergencies and Civil Defence Department of Public Works and Services, Hydraulics Direction Medical Care Emergency System Department of Health 107

121 Department of Logistics and Emergencies, Public Spaces Emergency Corp Department of Public Works and Services Table 5.1: Institutions of the City of Buenos Aires involved in flood management Policy / Action Plan of flood control: Coastal defence Plan of flood control: Renewing the rainfall drainage Hydraulic Master Plan Type of measure Structure No structure Current conditions Projected In Preparation In process Alert system for sudestadas and severe storms Civil Defence Metropolitan Master Plan Table 5.2: Actions and policies; in Green for La Boca only; in light blue for the city of Buenos Aires including La Boca The prevention phase In the national sphere the institutions that perform actions in the prevention phase are the National Meteorological Service (NMS), the Hydrographic Naval Service of (NSH) and the National Direction of Civil Defence. At the city level, the most out-standing institutions are the Department of Public Works and the Department of Logistics and Emergencies. Besides, and in case that a flood overcomes the level of crowning of the coastal defence, the Firemen are in charge of raising the level to a safety height Flood managements measures The first direct actions tending to solve the problem of floods were of construction type. It was a construction of a partial network of rainfall outlets made between 1874 and In 1934 levels of guarantee were established for the streets and for the houses (filling the low areas) that were surpassed a few years later with the major flood registered up to that moment in April, These were the only structural interventions until the decade of 1990 when the defence was built in La Boca and Barracas. This defence consists in a wall of concrete that prevents the inflow of the Riachuelo water. The height above sea level of the wall is 4.22, which corresponded to a recurrence of 245 years. Since the defence constitutes an obstacle for the evacuation of the rainfall waters, 7 stations of pumping and a collectors' system that gathers the rainfall water towards the pumps were constructed. The opening and the closing of the hatches, as well as the functioning of the pumps, is controlled by a computer that receives the information transmitted by sensors that detect the level of the water. Since its inauguration, the defence resisted several floods, the greatest of which took place on May 2000 with a peak of 3.05 meters. It is necessary to emphasize that only the record of historical sudestadas was 108

122 taken into account in the design of this defence without considering the future rise of the River level. Meanwhile, the neighbourhood continues being flooded by intense rainfalls because the renewing of the rain outlets is not yet finished. The implementation of direct non structural measures is more recent. The system of alert of sudestadas and severe storms in the City of Buenos Aires and the province of Buenos Aires was formalized in 1987 through an agreement between technical institutions (the NMS and the NSH) and the Civil Defence. The agreement was reviewed in 1993 and since then, the SHN issues the hydrological warnings and the SMN the meteorological alerts. More recently the Main Metropolitan Plan of Civil Defence was appointed by law to establish a "Basic Norm of Planning" for disaster situations, which must contain, coordinate and regulate all the specific plans of the different sectors, institutions and organisms involved with the occurrence of a catastrophe. The Plan constitutes a set of general guidelines for what has to be done by every institution involved in each phase (prevention, response, and rehabilitation). Apart from these aspects, the Plan contemplates the dissemination of information and the training of the population. As synthesis of the indirect actions led by the local government, it is possible to say that, after a long tradition of exposure to floods, the construction of the coastal defence seems to have opened the doors for the neighbourhood progress. Anyhow, it is necessary to emphasize that all these actions are taken without bearing in mind that the risk of sudestadas still exists and will growth as the sea level will continue rising during this century. There was also a permanent fragmentation in the decisions on urban policies that seems to be reversed recently The response phase With an alert notice of sudestada by the SHN or one of a severe storm by the SMN, the Department of Logistics and Emergency starts systematic actions to protect the neighbours. These actions consist of surveys in the more critical sites of the neighbourhood, in order to anticipate the problems and solve them. Simultaneously, the communication of the risk to the population is started. The warnings are spread through the massive means of communication (plates in television, warnings in the radio broadcastings). There are also recommendations and advices that the population should follow. The helpfulness of the system of communication is weakened by the lack of previous preparation and training of the population (what to do, where to meet in case of need of self-evacuation, etc.). Being the sudestada a hydrometereologic phenomenon a unique forecast would avoid some confusion among the population due to the emission of partial forecasts by the NMS and the SHN. On the other hand, and for the case of the severe storms, it is not enough with two daily reports. Apart from the system of official alert, the affected population has its own shock-absorbing networks of alert and of self-help and evacuation. The neighbours established information links, spreading the news about the alert state and have their own perception to anticipate the flood and their own strategy on how to protect their personal goods for different flood levels. Faced to the flood, the settlers appeal fundamentally to their own relatives and friends. In general, it is possible to say that in La Boca there was always a local organization to cope with floods. Nevertheless, the population at risk is seldom consulted and planning is still restricted to the design by specialists The reconstruction phase In the case of Buenos Aires, every catastrophic flood starts claims from the neighbours and the mass media, which is answered by the public officials with announcements of new infrastructure or the finishing of those that are already under construction. In addition, the answers from the public administrations (especially those depending from the Government of the City) do emphasis the extraordinary characteristics of the natural event (whether rainfall or sudestada). This reaction does anything but to outplace the responsibility, trying to associate it with climatic unforeseen factors. 109

123 Cultural capacities for adaptation The oldest residents of the neighbourhood have some knowledge of the dynamics of the river, and they pay attention to the signs that precede a flood. They developed strategies of response that complement the managing of the flood through the official alerts. The same thing happens with measures corresponding to the preventive phase, among which is frequent the elevation of their houses, even after the ending of the defence wall. The neighbourhood is having a rapid renovation of its population. The newcomers are people of low income, in most of the cases emigrants from neighbour countries. It is quite possible that the cultural capacities of adaptation will decrease, since the new comers do not have the local experience to face the floods and even, in some cases, tend to ignore the advices from the institutions involved in the management of the floods The Avellaneda Municipality The areas of the County most vulnerable to floods are occupied by very low economic income people who live in precarious settlements. The comparison of the census information, between 1991 and 2001 indicate that there was a significant demographic expansion of the precarious and illegal settlements in the floodable areas of the Avellaneda's Municipality. This trend clearly complicates the social adaptation to recurrent floods. Other vulnerable areas are some neighbourhoods of average and average-low sectors in the towns of Dock Sur, Sarandí and Villa Domínico. Most of the population of these neighbourhoods rarely get self evacuated or requires official help during the floods Institutional responses In Avellaneda, the institutional aspects of the disaster management are the major weakness in the response and adaptation processes to floods. The activity of the institutions, both at the prevention and at the emergency or response phase, as well as the interrelations between them is described in the following paragraphs. In the province of Buenos Aires, of which Avellaneda is part, there is a legal framework by which at each municipality, the Municipal Commissions of Civil Defence (MCCD), under the dependence of the Provincial Direction of Civil Defence, is the maximum organism in the management of disasters. In Avellaneda the MCCD is formed as follows: Major (Chairman), Civil Defence Director (CDD) (Secretary), Government Secretaries Health, Welfare, Public Works and Services, etc. Avellaneda s Voluntary Firemen, Dock Sud Naval Police Command, (PNDS) Red Cross Villa Domínico Office (CRAFVD) And leaders of civil organizattions as it is the case of the Scouts. These institutions, with the exception of the Voluntary Firemen of Avellaneda who possess an automatic warning system, act under request and under the coordination of the MCCD in the moment of the disasters. The CDD is seen by authorities as a political position, and it is renewed with the local government change every four years or sometime more frequently. This attitude goes against the need to keep the policies 110

124 with certain continuity. In addition, many of the members of the MCCD are also part of the political staff of the Municipality. Concerning the practices associated with the prevention phase, one of the principal functions of the MCCD is to develop a Municipal Plan of Emergencies, in which emergency hypothesis are formulated and developed. The floods appear as one of the principal hypotheses of emergency, but the plan is not well disaggregated to cope with the different situations in which this type of disasters appears. The MCDD uses a map of exposure in which the areas considered more exposed to floods are shown. The map is very precarious and does not contain any type of gradient by which some differentiation is established with regard to the degree of exposure. The major weakness is that the Municipal Plan of Emergencies is not well known by the people or the organizations of the civil society. Up to the moment, there is no public easily accessible information for the public on how to act before, during and after the disaster. During the emergency, the CDD is in charge of coordinating the rest of the institutions that form a part of the MCCD. In this instance all the secretariats of the government are at his disposition (available means and personnel). There is practically no involvement from the civil society Institutional conflicts One of the main points of conflict between the institutions involved in the management of the flood caused disasters in the County is related to the very existence of the MCCD. Although, the MCCD was constituted in conformity with the Provincial Law, the CRAFVD and the PNDS that are part of it are not called to participate in the meetings. The view of these organizations is that the MCCD meetings look more like a cabinet meeting of the municipal government than an authentic MCCD. The Voluntary Firemen of the County, in turn, have a troubled relation with both Civil Defence and the Government of the Municipality in general. This tense situation between the Municipality and the Voluntary Firemen of Avellaneda was amplified due to the mutual lack of fulfilment of assumed commitments as well as the lack of transference of information needed for planning. Finally, another example of the disarticulation between the main institutions in charge of the management of the disasters is associated with the practices of prevention. Periodically the CRAFVD, the PNDS and the Voluntary Firemen stations of Avellaneda perform joint emergency practices without the participation of the Municipal CDD, the designed coordinating institution Cultural adaptation All the interviewed social actors were born and currently live in Avellaneda and many take part in NGOs (Non-Governmental Organizations) that work on local problems. This circumstance favours their interaction with a wide range of people, and therefore resulting efficient collectors of the local experiences. The building systems adopted by the neighbours expose material signs that express the recognition of the inherent risk of flood. Among them, the more obvious is the elevation of the level of the houses. Often, these practices are transmitted from generation to generation with the aim to safeguard the lives and the material goods at the moment of the flood. Another factor that enhances the danger of floods is the rise of the phreatic layer. Several of the interviewee assured that in many zones of the County the layer is at less than 0.5m from the surface. This situation has repercussion on the building structures and in the building technologies. For instance, no water is added to the mortar for the construction of the columns and foundations because the water is absorbed from the soil by the materials spilt in the columns. In case of the most affected zones, the extension of networks of cooperation and of self-help depend on the number of relatives and friends in the neighbourhood, which may offer their houses for shelter or the materials for the elevation of the housing, etc. The establishment of the networks of solidarity and the relevance given to them by the neighbours themselves also shows the insufficient participation of the State in the prevention and mitigation of the recurrent floods. Another preventive measure adopted by the neighbours is associated with the alert notice, even before its public announcement. The warning is 111

125 often mouth to mouth transmitted, after telephonic consultation to the PNDS regarding the height of the Plata River Other vulnerable areas of metropolitan area of Buenos Aires Past adaptation strategies and their influence on future vulnerability Before about 1950, the areas exposed to frequent floods and no close to the downtown city like La Boca and Avellaneda, either remained not inhabited or in some cases were scarcely occupied by poor settlements where the people intruded the land without formal property or permission. Thus, very low areas, which will be likely permanently flooded by 2070/2080, are still scarcely populated because they are frequently flooded by storm surges or are in process of being elevated to be used as gated communities. As a result of this adaptation to current storm surge conditions, the social impact of future permanent flooding will be small. Therefore, climate change vulnerability in the coastal zone of the Plata River will come from future increase of the exposure to extreme surges. Neighborhoods that currently have relatively low recurrence of floods, and because of that are densely occupied, are those where the storm surge recurrence changes will create the greater impact. They are mostly in the valleys of the Reconquista and Matanzas rivers and are now occupied by different social stratus, ranging from middle class to socially vulnerable population. Thus, these changes will lead to social damages as well as to important real state losses as was discussed in chapter Present trends and their influence on future vulnerability During the 1950 decade started the occupation of low lands with precarious and usually illegal settlements. Obviously, this trend went against the collective adaptation to recurrent flooding. These lands were occupied by a population with unsatisfied basic needs, higher than national average child mortality rate and a high percentage of population without access to social security. In many cases, women are family heads. Social vulnerability of these settlements is worsen by the floods and will deteriorate more in future as the sea level rise will increase the frequency of floods in these areas. The most socially vulnerable areas that are at the same time affected by floods are not over the coast of the Plata rover, but on the flood valleys of two tributaries of it, the Reconquista and the Matanzas-Riachuelo rivers. Other two areas of the Great Buenos Aires that have large social vulnerability and flood exposure are the southern coast of the Plata River, 20 to 50 Km to the southeast of the Buenos Aires city and the county of Tigre, immediately to the south of the Paraná Delta. Starting in the eighties but with definite momentum since the nineties there was a dramatic change in the urban tendencies that affected partially these zones. New highways and increased demand for private gated towns are making these areas attractive as new settlements for the upper middle class. In this new process, the drive to gated community come from two main social perceptions: the fear linked with the increasing lack of security, and the idea that nature, country and green scenery are better conditions of life (Ríos 2002). At the beginning of the nineties, gated communities had an area rounding 34 square kilometers; while by the 2000 this area has grown nearly ten times: 306 square kilometers (Maestrojuan et. al. 2000). Argentina does not escape to the global phenomenon of population migration towards the coasts. Thus, it is very common that many gated community were localized in initially cheap suburban lands, as those that are frequently flooded. To have an idea of the momentum of this process until 1998, only in the district of Tigre, 90 gated communities had been authorized, out of which, 50 are already constructed. This trend is likely to increase in the next years. New projects spring all along the coast, both in the south eastern and in the northern extremes of the Buenos Aires metropolitan area, and even in the front of the Paraná delta (Rios 2002). The modification of the low level environment by closed towns has many effects, mainly in the hydrological drainage, which affects the people living around the gated communities. The urbanization of initially low sectors, historically frequently flooded, requires a massive transformation of the terrain and of the surface drainage with the destruction and replacement of the original ecosystems in order to obtain an assumed secure height. However, before the AIACC results were made public, most of the 112

126 gated communities considered adequate the height of 4.4 m over sea level, which may not be so safe in the future. With the spread out of these gated urbanizations a new situation of vulnerability and risk has been aroused. Habitants inside the urbanization have seen their crime insecurity mitigated, but they have now the flood threat. Population outside these gated towns, while living with crime insecurity, unemployment, violence, etc. is in top of that damaged by the lost of drainage due to the land elevation of the gated towns (Rios 2002). 5.4 Conclusions In the areas with long tradition of coexistence of population with floods, such as La Boca and Avellaneda, the existence of informal networks of alert, self-help and evacuation among the neighbours themselves, added to the practice of own strategies to anticipate the arrival of the flood, tend to diminish the vulnerability to floods. However, in both areas as well as in other areas where the occupation of lands with risk of floods is more recent, the increasing number of newcomers is reducing the collective cultural adaptation to floods. After its completion in 1998, the works of coastal defence in the city of Buenos Aires have mitigated successfully the last floods. Anyhow, La Boca neighbourhood continues being flooded as consequence of severe storms, since the renovation of the rainfall drainages has not been yet concluded. To this, must be added that the defence was designed without considering the future River level rise, what may reduce its efficiency in the future. In this city, although there are measures and plans of flood management, they are separated from the urban global environmental policy of the city, which can help to promote the vulnerability in the future. The institutional responses to floods, although following a similar organization pattern differs from one district to another in its functioning and coordination. In the case of Avellaneda, the lack of cooperation between the responsible institutions creates an additional source of vulnerability and it illustrates what happens in some other districts. Except in La Boca and Avellaneda, where the people found that its proximity to downtown compensated the annoyance of the recurrent floods, the rest of the coastal areas subject to floods were little populated until recent decades. This past adaptation prevented the occupation of the small areas of very low lands that will result enduringly flooded sometime this century. However, the current trends of occupation of lands with flood risk, by both very poor settlements and gated communities of upper middle class people are not favouring the collective adaptation to present and future scenarios of recurrent floods. 113

127 6 Capacity Building Outcomes and Remaining Needs 6.1 Workshops Investigators and Students participate AIACC workshops and several other meetings in order to gain more knowledge and experience. Following is the list of meetings attended: AIACC Global Kick-off Meeting11-15 February 2002, Nairobi, Kenya. Participants from the Project: Vicente Barros (PI), Claudia Natenzon (Co PI) and Angel Menendez (Co PI). AIACC Project Development Workshop: Development and Application of Scenarios in Impacts, Adaptation and Vulnerability Assessments15-26 April 2002, Norwich, UK. Participant from the Project: Inés Camillon.i AIACC Project Development Workshop: Climate Change Vulnerability and Adaptation 3-14 June 2002, Trieste. Participants from the Project: Claudia Natenzon (Co PI), Jorge Codignotto (Co PI), Mariano Re (Student) and Julieta Barrenechea. Climate Change in the Plata River. Joint workshop with AIACC Project Assessing Global Change Impacts, Vulnerability, and Adaptation Strategies for Estuarine Waters of the Rio de la Plata September , Montevideo, Uruguay. Participants from the Project: Vicente Barros (PI), Angel Menendez (Co PI), Claudia Natenzon (Co PI), Roberto Kokot, Mariano Re (Student), Inés Camilloni, Walter Vargas (Co PI), Susana Bischoff (Co Pi) and Gustavo Escobar. First AIACC Regional Workshop for Latin America and Caribbean May 27-30, 2003, San Jose, Costa Rica. Participants from the Project: Vicente Barros (PI), Claudia Natenzon (Co PI), Roberto Kokot, Mariano Re (Student), Carlos Rinaldi (Director of the Argentine second National Communication) and Andrea Ferrarazo from Fundación Ciudad. 2nd AIACC Regional Workshop for Latin America and Caribbean August 24-27, 2004, Buenos Aires, Argentina Participants from the Project: All investigators and students of the Project. Bellagio Synthesis report on Vulnerability March 7-12, 2005, Bellagio, Italy. Partipant from the Project: Vicente Barros (PI). 6.2 Other training activities supported by the Project Moira Doyle and Inés Camilloni participated of the PRECIS workshop in the CPETEC (Brazil) organized by CPTEC, CIMA and MET OFFICE. November 1 5, Courses Course for students The course was on Assessing Global Change Impacts, Vulnerability, and Adaptation Strategies for Estuarine Waters of the Rio de la Plata A course on Climate variability and anthropic influences was lectured at the University of la República (UdlR) in Montevideo during October 2002 as a joint activity with AIACC Project Assessing Global Change Impacts, Vulnerability, and Adaptation Strategies for Estuarine Waters of the Rio de la Plata. The course has credits for the Master of Science program on Environmental Sciences of the UdlR. Professors from the Project were V. Barros (PI) and S. Bischoff. 114

128 6.3.2 Courses on climate change for journalists Two short courses were offered for journalists, one in Buenos Aires (Argentina) in July 5, 2004 and the other in Montevideo (Uruguay) in July 13. They were designed to help the participants to cover the climate change issue and the results of both Projects. In Buenos Aires, the participants were 13 journalists from different media, about half of them from the two most important newspapers of Argentina, Clarin and La Nacion, which sell daily 400,000 and 170,000 newspapers respectively. This activity was very opportune as was done in advance to the COP X that was held in December in Buenos Aires and helped to have better and wider coverage of this event. There was an immediate response with three articles in the leading newspapers of the country, two in La Nación and one in Clarín. The three articles together amounted more than 500 cm. The notes in these newspapers leaded to a wave of comments on the radio and in some cases to interviews to the scientists of the Project. The wave reached even to the media in Uruguay with comments in television and newspapers. In Montevideo, (Common activity with Project AIACC Assessing Global Change Impacts, Vulnerability, and Adaptation Strategies for Estuarine Waters of the Rio de la Plata ) the participants were journalists from magazines and students of Communication. Between the journalists, there were some from the important magazine Busqueda. There were two notes in Busqueda and in Montevideo Digital, as in the case of the course in Buenos Aires, these notes leaded to a wave of comments and interviews on radio and television, and even in two newspapers, including the most important of the country, El Pais. 6.4 Students The Project supported the development of highly qualified students, who made their thesis in different aspect of it.victor Kind attained his degree of Hydraulic Engineer with a thesis on the salinity front of the Plata River. Diego Ríos obtained his degree of Licenciate in Geography with his thesis on gated communities. Three other theses are at its final stage, i.e. they are in the writing phase. One is for a doctor degree in Geography, Lic. Silvia Gonzalez (Buenos Aires flooding), the second is for a Master degree in Environmental Sciences, Eng. Mariano Re (The hydrodynamic modeling of floods) and the third for the degree of Licentiate in Atmospheric Sciences, Ezequiel Marcuzzi (El Niño and extreme precipitations). 6.5 General capacity building accomplishments Since the Project integrate research from climate, oceanography, geology, geography and social sciences, the Co PIs of the Project gained experience in multidisciplinary work and learned to synthesize results from different disciplines. During the Project, the participants developed some models and tools and learn to use others. These activities resulted in increased individual capacities in many techniques. As an example should be mentioned the development of the hydrodynamic model of the Plata Estuary, the analysis of regional results from GCM, the development of a high resolution topography and the development of social vulnerability indexes. A new and very important experience for most of the investigators of the Project was the work with stakeholder groups. Institutional capacity building resulted from the strengthening of three groups for further investigations in climate change Two of these groups are the at the University of Buenos Aires, one devoted to regional climate at the Center of Research on Sea and Atmosphere (CIMA), and the other to human geography and social sciences (PIRNA) at the institute of Geography. The third is at the National Institute of Water (INA) working in hydrodynamic modeling. It is important to stress the establishment of networks of persons and institutions between these three mentioned groups and others at the University of Buenos Aires and other institutions as the University of la República in Uruguay and the Department of Hydrography of the Argentine Navy. 115

129 Not less important was the contribution of the project to increase public awareness and understanding of climate change and related issues. This was done through workshops for stakeholders, conferences by the Co PIs and numerous notes and interviews in the media. 6.6 Remaining capacity needs In these last three years, the requirements of information on climate change, and in the related regional impacts from both, the government and civil society have created a growing demand of trained personnel in different aspects of climate change. Since, climate trends in Argentina were very important in the last decades, it is necessary to start adaptation or in some cases improve the current autonomous adaptation. Thus, it seems important to build additional capacities with focus on adaptation to climate change. 116

130 7 National Communications, Science-Policy Linkages and Stakeholder Engagement 7.1 National Communication The activities of the second National Communication of Argentina to the UNFCCC were assigned to consultants, scientific groups or institutions according to bids that were defined according to quality. The Argentine Co PIs of the Project gained the vulnerability study of the coastal area of the region of Buenos Aires. The Project results will be the basic input for this new study for the second National Communication that will start on July Recently, the third National Communication of Uruguay was started. The Uruguayan Co PIs of the Project are working in the climate scenarios for this Communication. 7.2 Contribution to UNFCCC activities In the COP-10, Buenos Aires December 2004, Vicente Barros (PI) was part of the Argentine delegation. In this COP, in the side event on Science in Support of Adaptation to Climate Change organized by START and UNEP, the PI of the Project made a presentation on Key Messages for Adaptation from Recent Assessments. Inés Camilloni, Jorge Codignotto (Co PI), Vicente Barros (PI) and Angel Menendez (Co PI) presented results of the Project in another side event on the Argentine Agenda on Climate Change organized by the Di Tella Foundation. 7.3 IPCC Jorge Codignotto (Co PI) is participating in the Fourth Assessment Report as lead author in the group 2, chapter 5 coastal systems and low lying areas. 7.4 National Policies The Secretary of Environment and Sustainable Development has developed the Environmental Agenda during This planning was developed through technical reports and workshops that were held in all the regions of the country. The Climate Change vulnerability section was developed by the Di Tella Foundation Reports in a series of 14 reports, of which 2 were on the Plata River coast and were almost completely based on the Project results. 7.5 Stakeholder engagement The Project and its objectives were initially presented to a small number of key stakeholders during the first year. Four of them answer a detailed questionnaire in connection with the objectives and activities of the Project They were City Foundation, Redes, the Federal Emergency System (SIFEM), and the Defendant of the People of Buenos Aires. The first two were NGOs that helped to enlarge the number of stakeholders in touch with the Project, while the remaining two were key instances in the administration of floods as coordinator, the first, and as control instance, the second. Some of the suggestions received in this inter-consultation process helped to reshape the tools to be developed by the Project. Example of this was the risk maps that were asked by the SIFEM. In addition, the Project was presented to a larger number of stakeholders in the workshop on Social Pertinence of the research developed at the University of Buenos Aires (UBA) in the area of floods. This meeting was part of the process of external evaluation of UBA activities. The Project was invited to participate because it received collateral funds from the UBA in the framework of a special program. The aim of this meeting was to the assessment of the research work in terms of its socioeconomic application. The modality adopted was to send the documentation of the projects selected to the participants. The 117

131 meeting took place on Tuesday 8 October Its objective was to evaluate, through a participatory process, the pertinence of the scientific research and its social benefits from the point of view of the stakeholders. The public institutions that participated invited by UBA were the Department of Water Resources Management from the National Secretariat of Water Resources, the Civil Defense Department of the Buenos Aires Province, the Secretariat of Environmental Policy and Food Security of the Municipality of Avellaneda, the Sanitation and Hydraulic Works Department of the Buenos Aires Province and two NGO, GAO (Associated Management of the West) and Pro Tigre. The participants found useful the Project objectives and the Department of Water Resources Management from the National Secretariat of Water Resources asked for continue support to this type of projects. In March 5, 2003, a workshop to present the initial results of the Project to stakeholders and to receive from them suggestions and demands was held. There were forty participants from 30 governmental and non-governmental organizations. In the same workshop, it was also presented results from the University of Buenos Aires Project on Floods on the Paraná and Uruguay Rivers that provide the collateral funds to this Project. As a result of this workshop, the Project received a considerable help in information from many stakeholders, and it was started a close work with the City Foundation. The City Foundation is an important institution that works on the City of Buenos Aires and Great Buenos Aires issues. It has a program called Buenos Aires and the River. This NGO gathers many of the stakeholders of coastal risks. It participated actively in the Project through two activities. It organized and moderated the final workshop with stakeholders and helped in the design and edition of paperback material oriented to the general public for dissemination of the project results. Final Workshop on Climate Change and the water level rise of the Plata River The City Foundation provided the methodology, co-organized the meeting with the Project and chaired the discussion sessions. The workshop with the personnel of the Project and 96 participants from NGOs, technical public officers, executives and technicians from the private sector was held in July 27, 2004 in Buenos Aires. The workshop had three sessions; in the first, the final results of the Project were presented by a panel of researches in a plenary. These results were discussed by three separate groups during the second session. Finally, in the third session, again in plenary, questions and recommendations from each group were addressed to the panel, for answers and discussions. The three groups were integrated by technical public officers, executives and technicians from the private sector, and NGO and conspicuous stakeholders. The workshop recommended further dissemination of the information produced by the Project and the revision of the regulation on soil use in the coast of the city and the Province of Buenos Aires to avoid the urbanization of the coastal low lands and to encourage the use of this space in recreational activities compatible with recurrent floods. It was also recommended a brief dossier for decision makers and legislators to alert them about the damage that could happen if no adaptation actions were considered on time. Finally, it asked for the elaboration of a communication document for the general public to alert them about the inconvenience of investing, building or settling in areas that could possibly get more frequently flooded in the future. This document was elaborated and it is referred in the next paragraph Paperback material for dissemination of the project This material was elaborated by the Project and was adapted by specialists in communication of the City Foundation, who also edited and published it. The document has 42 pages and includes about 30 illustrations. The engagement with stakeholders opened numerous opportunities for further work either to the core group of the Project, or to some individual researchers. The most important outcome of this activity is that it will be very difficult that the Project results will not be used in the future planning the Plata River coast. 118

132 8 Outputs of the project Until now, the outputs of the Project are mainly about specific aspects treated by only one discipline. More integrated products are expected for the next months, 8.1 Published in peer-reviewed journals Inés Camilloni and Vicente Barros 2003: Extreme discharge events in the Parana River and their climate forcing. Journal of Hydrology, Discusses the major discharges and their climatic forcings of the greatest tributary of the Plata River Gustavo Escobar, Walter Vargas and Susana Bischoff 2004: Wind tides in the Rio de la Plata estuary: Meteorological conditions. Int. J. Climatol. 24, It is a statistical analysis of the storm surge conditions that produce floods in the Argentine coast of the Plata River Walter Dragani and Silvia Romero 2004: Impact of a possible local wind change on the wave climate in the inner Plata River. International Journal of Climatology 24, Presents the study of the waves in the Plata River under present wind conditions and in a future scenario that maintain the current trend on surface winds Vargas Walter, Escobar Gustavo, Bischoff Susana, Berman Ana Laura 2005: Las sudestadas, climatología y circulación asociada. Accepted in Geoacta It is again a statistical study of the storm surge tides. However in this case, the paper focused in the fields associated or not to heavy rains during the storm surge. It is in Spanish 8.2 Other outputs Vicente Barros, Inés Camilloni, and Angel Menendez 2003: Impact of Global Change on the Coastal Areas of the Rio de la Plata. AIACC Notes, Vol.2, 1, 9-11 Discuss the sensibility of the level of the Plata River to its natural forcings. In addition discuss the southwards trend of the surface pressure field during the last decades and its effects on the surface wind field. Claudia E. Natenzon 2003: Inundaciones catastróficas, vulnerabilidad social y adaptaciones en un caso argentino actual. Cambio climático, elevación del nivel medio del mar y sus implicancias. Climate Change Impacts and Integrated Assessment EMF Workshop IX. July 28 -August 7, Snowmass, Colorado. Discuss social implications of the recurrent floods in the Argentine coast of the Plata River.It is written in Spanish Mariano Ré and Ángel N. Menéndez Modelo Numérico del Río De La Plata y Su Frente Marítimo para la predicción de los efectos del Cambio Climático. Mecánica Computacional Vol. XXII. M. Rosales, V. Cortínez and D Bambill (Editors). Bahía Blanca, Argentina, November 2003 It describes the modelling of the Plata River, its calibration and validation. It is in Spanish Gustavo Escobar, Inés Camilloni and Vicente Barros 2003: Desplazamiento del Anticiclón Subtropical del Atlántico Sur y su relación con el cambio de vientos sobre el Estuario del Río de La Plata. Tenth Latin American and Iberic Congress of Meteorology. La Habana, May 2003, CD-Rom. 119

133 It shows the shift of the South Atlantic subtropical high towards the south in the last decades, it was presented in a Latin American congress and it is in Spanish. It was edited in CD. Mariano Ré, Martín Kind and Ángel N. Menéndez 2004: La elección del dominio de cálculo, el modelo matemático y la escala de resolución en la modelación numérica. XIV Congreso sobre Métodos Numéricos y sus Aplicaciones, ENIEF 2004, Bariloche, noviembre de The paper discusses mathematical aspects of the Plata estuary modelation. It was presented in a national congress and therefore, it is in Spanish Vicente Barros, Angel Menendez, Claudia Natenzon, Jorge Codignotto, Roberto Kokot and Susana Bischoff 2005: El cambio climático y la costa argentina del Rio de la Plata. Fundación Ciudad, Buenos Aires, February pp. It is booklet to disseminate the Project results to the general local public. Consequently, it is in Spanish. It has three sections. One devoted to global Climate Change, the second to the regional climate and hydrological trends and the third on the results of the Project. The edition was made by the City Foundation for a total number of 500 booklets. El Cambio Climático en el Río de la Plata. Editors Vicente Barros, Angel Menendez and Gustavo Nagy, CIMA, Buenos Aires, May pp It is a book with 19 chapters that contain a selection of 15 technical reports from the Project and the AIACC project Assessing Global Change Impacts, Vulnerability, and Adaptation Strategies for Estuarine Waters of the Rio de la Plata, preceded by 4 introductory chapters on climate change and regional climate and hydrological trends. It has 200 pages, not including figures, which are edited in an attached CD. It is in Spanish, and it is directed to the local technical public. The edition was made by CIMA for a total number of 500 books. 120

134 9 Policy Implications and Future Directions The increasing frequency of recurrent floods reaching progressively more land added to the pressure from different and competitive human pressures over the coastal areas of the Plata River requires a public regulation of these areas. The results of the Project and their dissemination between the stakeholders provide a technical basis for such regulations. The works of defence produces a sense of security, which in view of the increasing trend in the River level, and consequently of extreme storm surges, conceal a risk that will grow with time. As a consequence, defences seen as final control of floods could facilitate in the future the densification of the population in exposed areas increasing social and economic vulnerability. In each case, the convenience of such defences must be carefully weighted against non-structural measures. One of the measures can be the regulation of the coastal space, favouring the use of activities compatible with the recurrent flooding. The System of Alert for Severe Storms and Sudestadas needs to be improved. It is necessary to establish a hydro-meteorological network in the sub basins of the rivers and brooks that cross the metropolitan area of Buenos Aires, ending in the Plata River. The system should have operative hydrodynamic models of these sub basins as well of the Plata River. The system has to be organized as a part of a broader national meteorological and hydrological system of alert for severe storms, with instantaneous information on line to issue immediate alert warnings whenever needed. The Civil Defence Metropolitan Master Plan should be improved enhancing the participation of the civil society in the education, planning and information processes. The results of the present Project as well as their dissemination will be of help in this aspect. The lack of consideration of the implications of climate change in the planning and design of structural works that prevailed until now has to be abandoned. Since this attitude was basically originated in the lack of knowledge, it is very likely that will be rapidly modified after the dissemination of the Project results, namely the books referred in section Until recently, in the Argentine society, it prevailed the conception of natural disasters as exceptional events (punctual and static) opposite to the normality of the daily life of a society. The above-mentioned conception allowed the improvisation and the lack of plans and regulations to incorporate the exposure to floods within an integral management before, during and after the events. The catastrophic results produced severe critics to the authorities in charge that was always refuted by them on the basis of the extraordinary nature of the event. Recently, the increasing frequency of extreme precipitations and the toll of lives causes by the subsequent floods have started to change the view of the natural catastrophes as exceptional and unforeseen events, and as a result of that, some officials are even facing penal trials. The availability of sound technical studies well disseminated between public and private stakeholders will contribute to speed up this new view with respect to natural extreme events. This new view, in turn, will force the responsible officials to work more efficiently and coordinate better the tasks of different institutions involved. It is therefore important that projects like the present one will be undertaken to address the climate driven extreme events in other systems and regions of the country in the context of the climate change and of other changing factors. 121

135 10 References Albarracín, J. 1987: Análisis de la Hidrodinámica del Río de La Plata a partir de una Simulación Numérica, Tesis de grado para la Licenciatura en Oceanografía, ITBA). Director: A. N. Menéndez, Informe LHA-INCYTH Anschutz G. 2000:. Comparison between SAR-ERS and Waverider Buoy meaurements in the outer Rio de la Plata estuary. In CoastalWave Meeting, September, Barcelona, Spain. Balay MA. 1961: El Río de la Plata entre la atmóosfera y el mar. Publicación H-621. Servicio de Hidrografía Naval, Armada Argentina, Buenos Aires Center; Barros, V., E. Castañeda and M. Doyle 2000: Recent precipitation trends in Southern South America to the East of the Andes: an indication of a mode of climatic variability. Chapter of Southern Hemisphere Paleo and Neoclimates Concepts, Methods, Problems. Ed P Smolka and W. Wolkheimer. Springer. Barros, V., L Chamorro, G. Coronel and J. Báez: The major discharge events in the Paraguay River; Magnitudes, source regions and climate forcings. J Hydrometeorology 2004 Vol 5, Bejarán, R. and V. Barros 1998: Sobre el aumento de la temperatura en los meses de otoño en Argentina Subtropical, Meteorológica, 23, Camilloni.2004: Representación del Clima del sur de América del Sur por los modelos climáticos globales) Fundación Di Tella. Proyecto ARG/03/ pp. Camilloni, I and V. Barros 2000: The Paraná River Response to El Niño and events. J. of Hidrometeorology. 1, Camilloni, I and V. Barros 2003: Extreme discharge events in the Paraná River and their climate forcing.: J. of Hydrology, 278, Campos, J. D., C. A. Lentini, J. L. Miller and A. R. Piola 1999: Interanual variability of the sea surface temperature in the South Brazilian Bight, Geophysical Research Letters, 26 (14), CARP 1989: Estudio para la Evaluación de la Contaminación en el Río de la Plata, SHN y SOHMA. Cartwright, D. E. and R. D. Ray 1990: Oceanic tides from Geosat altimetry, Journal of Geophysical Research, 95, Codignotto, J.O and Aguirre, M.L. 1993: Coastal evolution, changes in sea level and molluscan fauna in northeastern Argentina during the Late Quaternary. Marine Geology, (110): CERC 1984: Wave and Water Level Prediction. In Shore Protection Manual, Ch. 3, vol. 1. US Government Printing Office WashingtonDC; CERC 2002: Meteorology and Wave Climate. In Coastal Engineering Manual, Ch. 2, vol. 2. US Army Coastal Engineering Research D Onofrio E., Fiore M. and S. Romero 1999: Return Periods of extreme water levels estimated for some vulnerable areas of Buenos Aires. Continental Shelf Research, 4: Dangavs, N. V. 1983: Geología del complejo lagunar Salada Grande de General Lavalle y General Madariaga, provincia de Buenos Aires. Revista de la Asociación Geológica Argentina, 38(2): Defant A., 1961: Physical Oceanography. Volume II. Pergamon Press. 598pp. Dragani, W. 2002: SHN, comunicación personal. Dragani, W. C. and Mazio, C.A., Refracción de olas. Depto. Oceanografía, Servicio de Hidrografía Naval, Buenos Aires, Informe Técnico nro. 61/91, pp. 48. Fidalgo, F., F. De Francesco and R. Pascual 1975: Geología superficial de la Llanura Bonaerense. VI Congreso geológico argentino. Relatorio. Bahía Blanca. pag Foremann M. G., 1977: Manual for tidal heights analysis and prediction. Pac. Mar. Sci. Rep , 97 pp. Foremann, M. G. 1978: Manual for tidal heights analysis and prediction. Pac. Mar. Sci. Rep. 78 6, 70 pp. Framiñán, M. B., M. P. Etala, E. M. Acha, R. A. Guerrero, C. A. Lasta and O. Brown 1999: Physical characteristics and proceses of the Río de la Plata estuary in G. M. Perillo, M. C. Piccolo, M. Pino (Eds), Estuaries of South America, their geomorfology and dynamics, Springer-Verlag, Berlin, pp Furlong 1963: Cartografía Histórica Argentina. Mapas, Planos y Diseños que se conservan en el Archivo General de la Nación, Tomos I, II y III, Buenos Aires. Furlong, G. 1936: Cartografía jesuítica del Río de la Plata. En: Publicaciones del Instituto de Investigaciones Históricas N 71 de la Facultad de Filosofía y Letras de la Universidad de Buenos Aires, Casa Jacobo Peuser, Buenos Aires. Godin G., The analisys of tides.liverpool University Press, 264pp. 122

136 Griswold GM. 1963: Numerical calculation of wave refraction. Journal of Geophysical Research 68(6): Hamming R.A., Digital filters. Prentice - Hall, 223pp. Harris F., On the use of windows for harmonic analysis with discrete Fourier transform. Proceedings of the IEEE, 66-1, 51, 83pp. Intergovernmental Panel on Global Change (IPCC) 2001: IPCC WGI Third Assesment Report: The Scientific Basis, Chapter 2. Cambridge University Press. Jaime, P. y A. N. Menéndez 1999: Modelo hidrodinámico Río de la Plata 2000, Report LHA-INA , INA. Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Sha, G. White, J. Woollen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K. C. Mo, C. Ropelewski, J. Wang, A. Leetmaa, R. Reynolds, R. Jenne, y D. Joseph, 1996: The NCEP/NCAR 40-year Reanalysis Project. Bull. Amer. Meteor. Soc. 77, Kokot 1999: Cambio Climático y evolución costera en Argentina. Doctoral Thesis, FCEyN (UBA) 329 pp. Kokot, R. 1997: Littoral drift, Evolution and Management in Punta Médanos, Argentina. Journal of Coastal Research, 13(1): Maestrojuan, P., M. Marino, y G. de la Mota 2000: Enclaves urbanos atípicos en el Área Metropolitana de Buenos Aires: Su impacto socio- territorial. Ed. Oikos, Buenos Aires. Menendez 1990: Sistema HIDROBID II para simular corrientes en cuencos, Revista internacional de métodos numéricos para cálculo y diseño en ingeniería, Vol 6, 1. Menendez 2001: Description and modeling of the hydrosedimentologic mechanisms in the Rio de la Plata River, VII International Seminar on Recent Advances in Fluid Mechanics, Physics of Fluids and Associated Complex Systems, Buenos Aires. Menéndez, A. N. and R. Norscini 1982: Spectrum of Shallow Water Waves: An Analysis, Journal of the Hydraulics Division, ASCE, Vol. 108, No. HY1, January. Ministerio de Obras Públicas, Dirección General de Obras Hidráulicas, Bs. As., 1908: Memoria sobre el Río de la Plata, presentada al XI Congreso Internacional de Navegación. Molinari, G. N. 1986: Simulación numérica de la circulación en el Río de la Plata, Tesis de grado para la Licenciatura en Oceanografía, ITBA. Director: A. N. Menéndez, Informe LHA-INCYTH S Olalde, A. M. 1988: Simulación numérica de corrientes de deriva en el Río de la Plata, Tesis de grado para la Licenciatura en Oceanografía, ITBA, Director: A. N. Menéndez, Informe LHA-INCYTH Outes, F. 1930: Cartas y planos inéditos de los siglos XVII y XVIII y del primer decenio del XIX conservadas en el Archivo de la Dirección de Geodesia, Catastro y Mapa de la provincia de Buenos Aires, Talleres S.A. Casa Jacobo Peuser. Parker, G., C. Paterlini, P. Costa, R. Violante, S. Maarcolini and J. Cavalotto 1990: La sísmica de alta resolución en el estudio de la evolución costera del noreste bonaerense durante el Cuaternario. International Syumposium on Quaternary Shorelines: Evolution, Processes and Future Changes. Abstracts: 52. La Plata. Piola, A. R., E. J. Campos, O. O. Möller, M. Charo and C. Martinez 2000: Subtropical Shelf Front off eastern South America, Journal of Geophysical Research, 105(C3), Putnam, J. A. and J. W. Johnson, 1942, The dissipation of wave energy by bottom friction, Transactions of the American Geophysical Union, 30, 1, Ríos, D. 2002: Vulnerabilidad, urbanizaciones cerradas e inundaciones en el partido de Tigre durante el período Tesis de Licenciatura en Geografía Facultad de Filosofía y Letras, UBA, 191 pp. SHN, 1999a, Río de la Plata Medio y Superior, Carta Náutica H116, 4th ed., Servicio de Hidrografia Naval, Armada Argentina, Buenos Aires. SHN, 1999b, Río de la Plata Exterior, Carta Náutica H113, 2nd ed., Servicio de Hidrografia Naval, Armada Argentina, Buenos Aires. Simionato C, Vera C and F. Siegismund 2003: Surface wind variability on seasonal and interannual scales over Río de la Plata. Submitted to Journal of Coastal Research. Simionato, C., W. Dragani, M. N. Nuñez and M. Engel 2002: A set of 3-d nested models for tidal propagation from the Argentinean Continental Shelf to the Río de la Plata Estuary, Part I M2, submitted to Continental Shelf Research. SMN 1992: Estadisticas Climatologicas , Serie B, Nro. 37, Servicio Meteorológico Nacional, Fuerza Aérea Argentina, Buenos Aires. Spaletti, L. A., S. Matthews and D. Poiret 1987: Sedimentology of the Holocene littoral ridges of samborombón Bay (Central Buenos Aires province, Argentina). Quaternary of South America and Antarctic Peninsula, 5: Rotterdam. 123

137 Teruggi, M. E. 1949: Contribución a la Psamografía Argentina. Las arenas de la zona de Mar de Ajó. Notas Museo de la Plata. Tomo XIV. Geol. N 61. La Plata. Tricart, J. 1973: Geomorfología de la Pampa Deprimida. Colección Científica INTA, pp. Buenos Aires. Vincent, C. and A. Carrie, 1988: Evaluation of an energy-propagation wave refraction model. Continental Shelf Research, 8(3), Violante, R. A. 1988: Geología de la "planicie costera" entre Villa Gesell y Faro Querandí, provincia de Buenos Aires. Doct. Thesis, Fac. Cs. Nat. Mus. (UNLP). Watanabe, A., 1982: Numerical modelling of nearshore currents and beach deformation. Coastal Engineering in Japan, 25, Whitham, G. B. 1974: Linear and nonlinear waves, Interscience, New York. 124

138 For copies of final reports from the AIACC project and other information about the project, please contact: AIACC Project Office The International START Secretariat 2000 Florida Avenue, NW, Suite 200 Washington, DC USA Tel Fax Or visit the AIACC website at: