2018/TEL58/DSG/WKSP1/003 Flood Risk Reduction Strategies and the Early Warning Systems for Alerting the Population in Western Balkan Countries Submitted by: Alfa BK University Workshop on Earthquakes and Waterfloods Monitoring System with the Application of the Internet of Things Taipei, Chinese Taipei 1 October 2018
FLOOD RISK REDUCTION STRATEGIES AND THE EARLY WARNING SYSTEMS FOR ALERTING THE POPULATION IN WESTERN BALKAN COUNTRIES Emeritus prof. dr Larisa Jovanović ALFA BK University, Belgrade, Serbia ECOLOGICA Society, Belgrade, Serbia Abstract. The paper deals with flood risk reduction and methods of early flood warning. Flooding is one of most frequent and damaging type of natural disasters in Balkan countries. There is high probability that the impact of floods in Balkans will increase due to climate change. The watercourse arrangement includes the construction and maintenance of regulating facilities for watercourse regulation and the execution of works on maintaining the stability of the shore and riverbeds and providing water, ice and water throughput for the purpose of reducing the risk of flooding. Protection against harmful effects of water includes measures and works for protection against floods by external and internal waters, floods caused by ice barriers and floods of erosion and torrents. Since flood events are becoming more frequent and more intense, with the constant increase of exposure of people and goods due to intensive occupation of water-to-water areas, the need for increasing the level of protection of these areas is growing, so it is necessary to constantly improve and upgrade existing flood protection facilities and systems. Future research on floods should be focused on improving preparedness and response during a flood crisis, mainly warning and monitoring systems, also through the assessment of mitigation and prevention measures. Key words: flood, risk reduction, strategies, early warning, forecasting, modelling, Balkan countries. Introduction Balkan is one of the most dynamic and complex system in the world. It is also extremely fragile and sensitive to the effects of climate change. Climate change is gradually increasing the frequency and magnitude of extreme weather events and natural disasters in the region, which has led to higher levels of risk for population in Balkan countries. The paper deals with strategies of flood risk reduction and methods of early flood warning. The EU Flood Directive and management of flood risks in Western Balkan In order to avoid and reduce the adverse effects of floods, the European Parliament and Council adopted Directive 2007/60/EC on the assessment and management of flood risks. Given that the causes and consequences of flood vary across the countries and regions, the flood risk management plans for Western Balkan should consider the specific characteristics of areas they cover and provide for tailored solutions according to the needs and priorities of those areas, whilst ensuring relevant coordination within river basin districts and promoting the achievement of environmental objectives laid down in Community legislation. The Directive states that flood risk management plans should focus on prediction, prevention, protection and preparedness. The basis for priority setting and further technical, financial and political decisions regarding flood risk management is the monitoring of meteorological and hydrological data and development of flood risk maps showing possible adverse consequences associated with different flood scenarios, including information on potential sources of environmental pollution as a consequence of flooding. The legal framework of flood management in all countries of Western Balkan is aligned with the EU Flood Directive. Directive envisages Civil Protection for the purpose of assisting the population and improving the situation in crisis situations. 1
History of significant floods in the Western Balkans countries The torrential rains and catastrophic floods that raged in the spring 2014 through parts of Bosnia and Herzegovina, Serbia, and Croatia were unprecedented in the historical record of the region, going back 120 years. In just a few days, some areas received an amount of rain equivalent to one third of their annual total. But extreme weather events like this one are something communities may have to contend with more and more because of global warming impacts. Figure 1. Catastrophic floods in Balkan countries (B&H), May 2014 Table 1. Major flood events in the WB, 2010-1015 Date Affected areas, municipalities Extent of damage Albania Jan. 2010 Nov-Dec. 2010 Nov. 2014 Feb. 2015 Bosnia and Herzegovina Dec. 2010 Shkodra, Lezhë and Durrës. Drin and Mati River, Deltas Ulza and Shkopeti reservoirs Tirana, Lezhë, Shkodër and Fier, Vlora and Fier, Berat, Elbasan and Gjirokaster; Rivers Vjosa, Devoll, Osu, Seman Drina River catchment, Municipalities of Bosanska Krupa, Domaljevac-Šamac, Orašje, Tuzla, Maglaj, Goražde, Foča- Ustikolina, Pale-Prača, Ravno, Čitluk, Čapljina, Stolac, Mostar, Trnovo, Ilidža, Novi Grad, Tomislavgrad, Drvar, Trebinje, Bileća, Nevesinje, Foča, Novo Goražde, Bratunac, Zvornik, Bijeljina 2 10,000 hectares flooded, over 5,000 people evacuated, 2,200 houses damaged 15,000 people evacuated, 6,000 km 2 land flooded, 4,800 houses flooded 11,000 people evacuated, 3 people died, 7500 houses damaged Flood impact rating 42,000 people affected 20,000 people affected, 5,000 houses flooded, 6,000 people evacuated
May 2014 Aug. 2014 Kosovo March 2013 Macedonia Feb. 2013 Jan-Feb. 2015 Sava tributaries: Una, Sana, Vrbas, Vrbanja, Bosna and Drina and Sava River at Raca Northern and Western Bosnia and Herzegovina. All areas along the Sava, Sava tributaries: Una, Vrbas, Stira; Banja Luka, Gracanica, Tuzla, Foča, Visegrad, Zvornik, Zepce, Lukavac, Zenica Municipalities of Klina/ Klina, Skenderaj/Srbica, Peja/Pec, Istog/ Istok, Kamenice/Kamenica, Gjakova/Djakovica, and Mitrovice/Mitrovica. Rivers Drini I Bardhe, Klina, Bistrica and Lushta River Kojnarka, Kumanovo, Shtip, Sveti Nikole, Strumica, Valandovo, Ohrid, Probishtip and Kochani Eastern region: River Crna - Region of Bitola, Municipalities of Mogila, Novaci and Bitola Nearly 15% of GDP lost, 13,200 km 2 flooded, over 1 million people in 46 municipalities affected, 25 lives lost Some 200 homes evacuated Flash floods several towns flooded. Water supply shortages Approximately 6,000 people affected Over 100,000 people affected extreme moderate Feb. 2015 Southern and central parts of the country 100,000 people affected March 2015 Municipalities of Kavadarci, Prilep and Kumanovo, Northern and central parts of the Country Agricultural lands affected Montenegro Dec. 2010 Whole of Montenegro to various extents Rivers Lim, Tara, Moraca, Drina tributaries and Bojana; Lakes Skadar, Piva and in Niksic 21 municipalities affected, 1.49% of GDP equalling to MEUR 43 lost moderate July 2014 Central and South-Montenegro Landslides, roads blocked moderate Serbia Eastern and central parts of Serbia: Feb. 2010 Zajecar, Aleksinac, Pozega and Knjazevac, 1,306 households damaged, Negotin, Svrljig, Boljevac; Merosina, more than 3,150 people affected Doljevac, Koceljeva, Ub, Lajkovac, Ljig, Vladimirci, Zitoradja, Priboj and Prijepolje June 2010 Kolubara 135 households affected, over 2000 ha flooded Approximately 3,500 people affected, 181 families evacuated EUR 1,525 million lost equal to about 3% of the GDP, 9,100 km 2 and 38 municipalities/cities affected, 1.6 million people affected, 51 lives lost moderate Feb. 2013 Pčinja District of southern Serbia Bujanovac, Preševo, Trgovište Western, South-western, Central and Eastern Serbia: Sava, Tamnava, Kolubara, May 2014 Jadar, Zapadna Morava, Velika Morava, extreme Mlava and Pek at Beli Brod on the tributary river Kolubara - Obrenovac July 2014 Central Serbia, municipalities of Kostolac and Pozarevac, Topola Power plant shot down moderate Aug. 2014 Western Serbia, River Stira, Cities of Loznica, Banja Koviljaca 100 homes flooded moderate Sep. 2014 Eastern Serbia municipalities of Kladovo, Approximately 7,000 people Majdanpek and Negotin affected Source:Various: Danube River Basin District: flood events in 2010 (ICPDR flood report 2010), FloodList, ReliefWeb, International Federation of Red Cross and Red Crescent Societies, http://www.telegraf.rs/vesti/1231030-posle-uzasa-u-istocnoj-srbiji-steta-od-poplava-5-miliona-evra 3
Strategy of Flood Risk Reduction in Serbia In order to improve water management policy and support further harmonization with requirements of the relevant legal acquis of the EU it is necessary to make a number of steps in the legislative framework of the Balkan countries. In Republic of Serbia the Law on Waters was amended in 2016, in which transposition provisions of the Water Directives have been transposed in it. In 2017, the Government of the Republic of Serbia adopted the Strategy for Water Management in RS up to 2034. On the basis of the Strategy, sector reforms will be implemented in order to achieve the necessary measures in water management. Structural and nonstructural measures Structural measures (construction of flood defenses such as riverbed regulation and the rehabilitation or construction of dikes) were the most prevalent measures in Serbia and B&H in the previous period. Structural measures identified as non-regret in the Action Plan implemented through the EU Instrument for pre-accession assistance 2014-2020 (IPA 2) and World Bank programs. Out of 39 areas identified as under high flood risk, implementation of the above-mentioned programs is planned for 15. Non-structural measures are proposed for the remaining 24 areas in 19 municipalities. Non-Structural Measures encompass activities that can be categorized as non-physical (planning and designing structural measures, preparedness measures, environmental measures, government and legislative measures, financial measures) or physical (emergency response measures). Measures for Reducing the Flood Risk to the Housing Sector The existing system of protection in the Republic of Serbia relies primarily on passive measures. Thus, in the waters of the first order, there were about 2,500 km of embankments, linear structures at a length of about 1,275 km, 58 dams. In spite of such important infrastructure facilities, the concept of the fight against floods has proved to be unsustainable in recent times, especially after the flood events recorded over the past ten years. The reasons for this should be sought in the fact that for the establishment and maintenance of this concept, very large investments are needed, which even much more developed and economically more powerful societies cannot provide. Flood monitoring and mapping Most countries have developed some sort of flood risk maps for some areas, although it is highly different how these are made. Many countries are referring to the EU Flood Directive (EU, 2007). Flood maps are still missing for many rivers, or are only based on previous floods (Macedonia). The flood risk maps will to some degree be used as a planning tool for future flood protection. The EU Flood Program for B&H recognizes the importance of investing in future risk informed decision making and thus initiated the development of a Flood and Landslide Risk Assessment for the Housing Sector in B&H. The total population living in the areas exposed to very significant risk of flooding (category 4 hazard maps) is 283,777, while the total population living in areas at very significant risk of landslides (category 4 hazard maps). 4
Figure 2. a) Relative flood risk assessment for the housing sector per municipality b) Relative landslide risk assessment for the housing sector per municipality Figure 3. Relative combined multi-hazard risk per municipality Meteorological and hydrological networks The Weather Watch Department of the Republic Hydro-meteorological Service of Serbia, as part of the World Weather Watch, monitors weather 24 hours a day through its sections: Upper-air Observations Section, makes upper-air and ground measurements of meteorological parameters, process data and gives reports. Radio sounding is the basic method for upper-air measurements and observations and is made by balloon-borne radiosondes. The meteorological network density is very low for Bosnia and Hercegovina and for Montenegro, also. Serbian meteorological network has 458 stations, only 28 of these are automatic. The manual stations observe 3 times per day (53 stations) or once per day (390 stations). 5
Additionally, Serbia is covered by15 Doppler radars. Radar Meteorology Section, tracks down clouds and cloud systems in the area of 300 km around Belgrade. This is very precise method for time-spatial estimation of precipitation and extreme weather events. Weather Forecast Section, incorporates work of all other Sections of the Weather Watch Department by making operational forecasts. Numerical Weather Prediction Section, makes objective analysis, initialization of meteorological fields and preparation of the boundary conditions for synoptic and mesoscale numerical weather models (ETA model), using collected and processed data. Table 2 gives overview of meteorological network for Balkan countries. Table 2. Overview of the meteorological network Country Manual Automatic Data transfer Radar(s) Albania 120 30 Immediately - monthly 1 Bosnia and Herzegovina 24 21 Daily 0 Macedonia 103 33 Immediate - monthly 2 (1) Montenegro 29 15 Immediate - monthly 0 Serbia 458 28 Immediate 15 (1) These radars are old and/or there are problems with maintenance. Figure 4. The number of meteorological stations per 1000 km 2 http://publications.jrc.ec.europa.eu/repository/bitstream/jrc108843/online_version.pdf 6
Figure 5. Monitoring of meteorological data Hydrological network Table 3 gives an overview of the hydrological networks for each country. Figure 6 shows the number of hydrological stations in each country per 1000 km 2. Compared to the area, the highest numbers of stations are found in Israel, with high density also in Albania, and Serbia. Serbian hydrological network consists of 180 manual stations and 95 automatic stations. The automatic stations are collocated with manual stations; hence the total can be regarded to be 180 stations. Automatic stations measure every hour (10 minutes in some cases) whereas observations at the manual stations are taken once per day. Observations are available on the website of RHMSS and in data base. Table 3. Overview of the hydrological network Country Manual Automatic Hydrological data transfer Albania 85 20 2 hours - monthly Bosnia and Herzegovina 15 118 Daily Macedonia 65 10 Immediate - daily Montenegro 3 24 Daily - monthly Serbia 180 95 Immediate 7
Figure 6. The number of hydrological stations per 1000 km 2 Forecasting, modelling and satellite monitoring Most countries with own forecasting service are running WRF, including Bosnia and Herzegovina, Macedonia, Montenegro. Serbia have ongoing projects for implementing a domestic weather forecasting model. Table 4. Weather forecasting in different countries Member ECMWF (1) National NWP (2) ECMWF COSMO/ ICON GFS/ NCEP DWD WRF Aladin/ Arpege/ Arome Albania I (3) I I X Bosnia and Herzegovina X X (4) X X X Macedonia C X X (4) X Montenegro C X X X X Serbia X X X X X X (1) X means member, C means cooperating state (2) Which countries have their own numerical weather prediction (3) I - the country receives images (charts) from the large scale forecasters (4) X - the country receives numerical forecasts from the large scale forecasters Several models are used for forecasting for different rivers. For the large ones, it is mainly a combination of statistical methods and river routing. For small and medium catchments, the HBV model is applied. For catchments without models, the forecasts depend on expert judgement by experiences hydrologists. Several weather forecast models are used, including global models (ECMWF, DWD, NMMB) and limited area models (WRF Europe, WRF Balkan, NMMB Europe, NMMB Balkan and ETA). Sava Super Model - Sava SM (Croatia). The Sava Super Model (Sava SM) is a 1D / 2D model, created on the MIKE11-DHI software platform, covering the Sava River Basin on the territory of the Republic of Croatia, from the border with the Republic of Slovenia to the border with the Republic of Serbia.. http://meteo.hr/infrastruktura.php?section=prognosticki_modeli¶m=sava_sm 8
Figure 7. Hydrological state and prognosis: The scope of the model with indicated prognostic points The model of Sava was completed in 2016 as part of the project "Continue implementation of pilot project and capacity building for flood forecasting", jointly by SHMS, Croatian waters and consulting houses PRONING-DHI, with the inter-state support of the Republic of Slovenia (ARSO), Bosnia and Herzegovina (AVP Sarajevo, FHMS Sarajevo), Republic of Srpska (HMS Bijeljina). The Sava Super Model is supported by an operational system for forecasting water supply and flow on the Sava River Basin from the border with the Republic of Slovenia to the border with the Republic of Serbia. The feature of the hydrologic prognostic model for the Sava River Basin is the forecast of water supply and flow up to 120 hours in advance, based on 145 prognostic points, i.e. hydrological stations, of which 124 in Croatia and 21 in B&H. The timing of the simulation run is every 1 hour, with the results being published on the protected websites of SHMS, Croatia. The Sava model contains the retention systems of the Central Sava basin as well as the preliminary hydrological models of the basins of Una and the right tributaries of the Sava from Bosnia and Herzegovina. The model included: 159 basins, 517 km of Sava river stream, 3082 km of all tributaries, 3060 cross-sections, 23 overflows, 15 control objects, 13 retention, 17139 count points. 9
Figure 8. Prognosed water levels for 5 days in advance in the area of middle Sava basin (on June 15, 2018, simulation at 12:00 UTC + 1) International cooperation Flooding is quite often a transboundary issue, and all countries have the cooperation with the neighboring countries. All of the assessed countries that are covered by the current extent of the European Flood Awareness System (EFAS), are now also partners of EFAS. EFAS is a partnerbased service under the umbrella of Copernicus Emergency Management Service, monitoring and forecasting flood events across Europe. This system offers probabilistic 1-10-day flood forecasts for all rivers above 2 000 km 2 in addition to flash flood forecasts and a range of related products: forecast precipitation, remotely sensed and estimated soil moisture content and snow depth. Serbia also has bilateral agreements with Romania and Hungary, and is partner of the Sava Flood Forecasting and Warning System (Sava FFWS). There are some international projects aiming at better flood forecasting across country borders. Two of these are on Balkan, for the Sava basin and for the Drina basin. One example is the project Climate change adaptation in the Western Balkans, managed by GIZ and funded by the German Federal Ministry for Economic Cooperation and Development 10
(BMZ). The project focuses on the Drin river, implementing a flood forecasting model for the basin - in Macedonia, Albania, Montenegro and Kosovo. There are also other EU driven projects that bring countries together: 1. IPA FLOODS has been developed in order to support the approximation to the EU Floods directive in Western Balkans countries. It is a Program for Prevention, Preparedness and Response to Floods (http://ipafloods.ipacivilprotection.eu/). 2. IPA DRAM In the same framework, the Program for Disaster Risk Assessment and Mapping (IPA DRAM) contributes to enhance the capabilities of the partner countries to strengthen disaster risk management (http://www.ipadram.eu/about-the-programme/) Early Warning System Framework for Balkan cities Early Warning Systems (EWS) can play a crucial role in mitigating flood risk by detecting conditions and predicting the onset of a catastrophe before the event occurs, and by providing real time information during an event. EWSs thus fulfil multiple roles as general information systems, decision support systems and alarm systems for multiple stakeholders including government, companies, NGOs and the general population. Figure 9. Model of EWS (http://www.urbanflood.eu/pages/aboutus.html) Meteoalarm EU (www.meteoalarm.eu) provides the most relevant information needed to prepare for extreme weather, expected to occur somewhere over Europe (today or tomorrow). This information is presented consistently to ensure coherent interpretation as widely as possible throughout Europe. The participating countries in Meteoalarm EU are: Austria, Belgium, Bulgaria, Switzerland, Cyprus, Czech Republic, Germany, Denmark, Estonia, Spain, Finland, France, Greece, Hungary, Ireland, Iceland, Italy, Luxemburg, Latvia, Republic of Macedonia, Malta, Montenegro, Netherlands, Norway, Poland, Portugal, Romania, Serbia, Croatia, Sweden, Slovenia, Slovakia, United Kingdom. 11
Figure 10. Community based flood early warning system CONCLUSION Whereas the meteorological conditions cannot be controlled, early warning, preparedness and response to floods are essential to reduce the damages of flooding. Meteorological and flood records will give an indication of flood risk and help in developing flood forecasts. Real-time monitoring, access to meteorological forecasts and a capacity to issue qualitative or quantitative forecasts will give more time for evacuation, and in many cases, also for installation of temporary flood protection measures along rivers. Developing comprehensive domestic monitoring tools and early warning systems for flood response at local and regional level requires not only an effective monitoring infrastructure for hydrometeorological variables, but also specific expertise, adequate IT and a solid administrative and legal framework. The capacity for flood monitoring and modelling is highly different in the different countries. Some countries have a large number of automatic meteorological and hydrological stations, others are still relying on manual stations. Some countries have good modelling tools for many watersheds, whereas other countries are relying on expert interpretation of weather forecasts. Developing, implementing and maintaining flood early warning and monitoring systems requires a wide range of highly specific skills from hydrology, meteorology and ICT. The main need in Balkan countries are more automatic meteorological and hydrological stations (in mountainous regions, specially). Also, more staff would be necessary to be able to implement and run rainfall-runoff models. There is a high need and demand for training and capacity building in all areas (monitoring, forecasting, modelling, etc.) by almost all countries of Western Balkan. 12
REFERENCES 1. Directive 2007/60/EC 2. H. Unnerstall, Legal Framework for Public Participation in Flood Risk Mapping - A Comparative Study of the Responses of Different European Member States to Some Requirements of the Floods Directive, Utrecht Law Review, Vol. 8, No. 3, 2012. 3. A. Keessen, H.F.M.W. van Rijswick, Adaptation to Climate Change in European Water Law and Policy, Utrecht Law Review, Vol. 8, No. 3, 2012. 4. L. Jovanović, D. Stojadinović, Managing of water resources in urban areas of Serbia. 4th European congress on regional geoscientific cartography and information systems, Bologna 2003, Proceedings, Vol. I, p. 216. 5. L. Jovanović, The study of floods and landslides in Belgrade. 14th Meeting of the Association of European Geological Societies (MAEGS 14), Torino 2005, Abstracts, pp. 93-94. 6. Strategy for Water Management in the Republic of Serbia up to 2034. 7. Law on Water, Official Gazette of RS, No. 30/2010, 93/2012 and 101/2016. 8. Law on Emergency Situations 9. EU Flood Recovery Program for Bosnia and Herzegovina 10. Danube River Basin District: flood events in 2010 (ICPDR flood report 2010), FloodList, ReliefWeb, International Federation of Red Cross and Red Crescent Societies, http://www.telegraf.rs/vesti/1231030-posle-uzasa-u-istocnoj-srbiji-steta-od-poplava-5- miliona-evra 11. http://www.hidmet.gov.rs/eng/meteorologija/index.php 12. http://www.urbanflood.eu/pages/aboutus.html 13. http://meteo.hr/infrastruktura.php?section=prognosticki_modeli¶m=sava_sm 14. www.meteoalarm.eu 15. https://unfccc.int/climate-action/momentum-for-change/activity-database/communitybased-flood-early-warning-system-india 16. http://ipafloods.ipacivilprotection.eu/ 17. http://www.ipadram.eu/about-the-programme/ 18. http://publications.jrc.ec.europa.eu/repository/bitstream/jrc108843/online_version.pdf 19. http://emergency.copernicus.eu/ 20. https://www.scientificamerican.com/article/unprecedented-flooding-in-balkans-caused-bylow-pressure-parked-over-se-europe/ 21. https://www.bostonglobe.com/...flooding-croatia.../story.html 22. www.xinhuanet.com/english/2018-03/20/c_137050988.htm 23. https://reliefweb.int/disaster/ff-2014-000059-srb 24. htps://www.usnews.com/news/world/articles/2018-07-01/romania-h 13