World Meteorological Organization AMDAR Panel. Aircraft Data for NWP

Transcription

1 World Meteorological Organization AMDAR Panel Aircraft Data for NWP By J. J. Stickland, Technical Coordinator 8th WORKSHOP ON METEOROLOGICAL OPERATIONAL SYSTEMS ECMWF November 2001

2 Historical Background First Aircraft to Satellite Data Relay (ASDAR)system developed in mid 70s First Aircraft Meteorological Data Relay (AMDAR)systems developed in mid 80s using the Aircraft Communications And Reporting System (ACARS) WMO ASDAR program commences in 1991 By mid 90s, 3 major operational AMDAR centres had been developed - Oceania, Nth America, Europe WMO AMDAR Panel formed in 1998 Southern Africa project commences operations in 2000 The AMDAR Panel adopts AMDAR as the generic name for all automated aircraft meteorological observing systems

3 10 Countries Producing AMDAR Observations Oceania Nth America Europe Africa Australia US Holland South Africa New Zealand UK Namibia France Sweden Germany Note: 5 European providers are supported by 14 EUMETNET countries 16 Airlines Qantas United KLM SAA Ansett? American BA Air Namibia Air New Zealand Nth West Air France Delta SAS Fed Ex Lufthansa UPS

4 3 Countries Testing New AMDAR Programs Canada (Air Nova, extending to other regionals and Air Canada) Hong Kong (Cathay Pacific) Saudi Arabia (Saudia) 12+ Countries Planning or Developing New AMDAR Programs Spain (Iberia) Finland (Finnair) Iceland (Iceland Air) Austria (Lauda) Italy (Lauda) Poland (LOT) Morocco (RAM) Japan (JAL) China (Several) Chile (Lan Chile) Russian Federation Southern Africa Region (under SADC) including Mauritius

5 Kenya Countries Interested in Developing AMDAR Programs Egypt (Egypt Air) Arab Emirates (Emirates, Gulf Air) Greece Ukraine Using National Airlines Kazakstan Singapore Malaysia Korea Brazil 15 countries in central and west Africa under ASECNA Targeted Observations: The use of AMDAR equipped foreign aircraft operating into remote airports to provide observation profiles and over-flight data. Data are provided by the host airline/nmhs on the GTS.

6 Australia US Europe Countries Providing Targeted Observations Countries Receiving or Contemplating Targeted Programs Canada (E-AMDAR) Eastern Europe (E-AMDAR) South Africa (E-AMDAR, Australia) Oman (E-AMDAR) Africa (E-AMDAR) Middle East (E-AMDAR) ASECNA group (E-AMDAR) Hong Kong (Australia, US) Caribbean, Central America & Gulf of Mexico (US, E-AMDAR) SW Pacific and SE Asia (RAV)

7 E-AMDAR Targeting into Middle East, Eastern Europe & Independent States Blue - Routine program Red - Targeted Program

8 Network Optimisation: The automated control of collecting AMDAR data - through selection by phase of flight at appropriate temporal and spatial density to minimise data redundancy to improve cost effectiveness. Techniques: onboard software control (lat/long boxes, airport lists) ground based system through uplinking using scheduled flight and/or real time aircraft operations information Applications: Regional Global potential

9 Optimised E-AMDAR Data

10 E-AMDAR Optimisation Reduction of 37% in Lufthansa observations following automated optimisation

11 24 Hour Global Coverage Courtesy: NOAA Forecast Systems Laboratory, Boulder, Colorado

12 Typical Number of Daily Observations: US 80,000 (not all on GTS) Australasia 3,000 Europe 25,000 Africa 1000 WMO ASDAR 1,000 Approx. Total 111,000 Observations Used Operationally by ECMWF 100 No. Obs (x1k)

13 E-AMDAR Temperature Quality 18% O-B temperature difference distribution QEvC 2001Q02 16% 14% 12% LVR/LVW ASC/DES percentage 10% 8% 6% 4% 2% 0% temperature difference (K) Frequency distribution of the mean temperature difference (OBS Background) KNMI QEV Report - April - June 2001

14 E-AMDAR Wind Speed Quality 40% O-B wind speed difference distribution QEvC 2001Q02 35% 30% percentage 25% 20% 15% LVR/LVW ASC/DES 10% 5% 0% wind speed difference (m/s) Frequency distribution of the mean wind speed difference (OBS Background) KNMI QEV Report - April - June 2001

15 E-AMDAR Wind Speed Quality 30% O-B abs wind direction difference distribution QEvC 2001Q02 25% LVR/LVW ASC/DES 20% percentage 15% 10% 5% 0% wind direction difference (deg) Frequency distribution of the mean of the absolute wind direction difference ( OBS Background ) KNMI QEV Report - April - June 2001

16

17 E-AMDAR Data Timeliness timeliness (cumulative) FM42 QEvC2001Q2: May % percentage received records 75% 50% 25% 0% delay (minutes). Cumulative frequency distribution for all FM42 encoded EU-Amdar observations during the period May 2001 as a function of the interval between observation and time of reception. (median value: 7,2 minute).

18 Cost per Observation: Communications Costs (Fully automated observation containing position, time, temperature, wind speed and direction, humidity, turbulence and additional supporting data) In the range <1 US cent to 11 US cents Median value is 4 US cents Profile Cost: (Profile from surface to 200 hpa or cruise levels) At approx. 30 obs per profile and 4c per ob.= $1.20 GPS Radiosonde Sounding Cost: (Sounding from surface to 10 hpa) Approx. $300 including staff time. Cost benefit ratio: Approx 300:1

19 Annual Cost of Targeted Observations Program Program Annual cost (USD) Radiosonde One profile per day ,000 Two profiles per day ,000 4 profiles per day ,000 Every 3 hours 3500 Limitations Maximum height limited to cruise level of aircraft Humidity currently not available Limited to locations covered by comms providers (ARINC, SITA etc.) Many aircraft not equipped

20 Data Exchange GTS Codes: Text: WMO FM 42-XI - Oceania, Africa, Europe except Germany BUFR: WMO FM94 - US, Germany Secure FSL Web Site: Very powerful real time interactive graphical display of coverage and profiles, plus data in text format is available to all NMHS and bona fide meteorological non-commercial research groups at: NOAA FSL: To obtain log-in name and password, contact: Bill Moninger at - moninger@fsl.noaa.gov

21 Consists of 4 levels: Quality Monitoring and Control Onboard software (systems cross checks, gross error limits, outlyer rejection); Ground based data acquisition at the NMHS before data are distributed for operational use (format checks, gross error limits, outlyer rejection, rates of change, use of lists of blacklisted and poor performing aircraft to inhibit data exchange); National/regional monitoring centres providing daily statistical reports (referenced against NWP models) and detailed performance enquiries to national focal points; Global monitoring centres providing monthly statistical and blacklisted aircraft information to regional or national focal points.

22 Onboard Software Standards Two global standards are currently in operational use: 1. ACARS Aircraft AMDAR (AAA) suitable for Teledyne ACMS avionics hardware originally developed by KNMI for KLM and recently upgraded by E-AMDAR and the Panel implemented in various forms by 9 airlines very flexible, provides excellent targeting and optimisation controls. 2. ARINC 620 Meteorological Report the avionics industry standard suitable for limited models of avionics hardware from all major manufacturers implemented in various forms on 6 airlines has limited operational flexibility and controls for targeting and optimisation specification is being upgraded to AAA standard by the Panel in conjunction with the Airlines Electronic Committee (AEEC) Datalink Systems Sub-committee to make ONE UNIVERSAL STANDARD SPECIFICATION FOR ONBOARD SOFTWARE

23 1. Turbulence New Observations Operational reporting of turbulence in AMDAR commenced in 1985 on Australian aircraft (Derived Equivalent Vertical Gust Velocity (DEVG)), and subsequently on aircraft from New Zealand, UK, Netherlands and France. It forms part of the AAA specification. It is also reported by some US aircraft as delta g AIREPS use ICAO turbulence categories More recently, the US has developed and is trialling Eddy Dissipation Rate (EDR). This has been adopted by ICAO for automated met reports. Considerable verification and evaluation work is needed on EDR Although work is proceeding at NCAR to use EDR in models, no NMHS has firm plans to do so although some of the bigger centres are considering it.

24 2. Humidity/Water Vapour Development work in the US continues on a much needed Water Vapour Sensing System (WVSS) Operational trials on WVSS-1 based on Vaisala thin film Humicap sensors are nearing completion. This technology is no longer considered suitable. Development of WVSS-2 based on laser diode technology is nearing completion. This sensor is considered more appropriate for aircraft operations. First certification on an operational B757 is scheduled for early Sensor is integrated into new TAT probe along with the TAT sensor Installation cost is of the order of US$20k Airlines must carry spares and need to upgrade installation and maintenance procedures and manuals. Work also continues in the UK on the development of a laser diode sensor but the project is several years behind the US.

25 3. Icing Various technologies are being evaluated No operational systems have been implemented Additional Systems ICAO Automated Dependence Surveillance (ADS) Met Reporting Troposphere Aircraft Meteorological Data Reporting - new system of automated aircraft reporting being developed by NASA, NOAA and FAA for smaller regional and general aviation. Combined sensor, electronics and communications packages are expected to cost around $10k

26 Australian Data Volumes at Major Airports Pre- and Post- Ansett s Demise

27 Future Directions Encourage and assist more countries, particularly those in data sparse areas to implement AMDAR programs of targeted observations and if possible their own operational systems on local aircraft More lower flying regional aircraft will become operational (15k - 20k ft) New sensors will become operational over the next few years Encourage the development of regional programs to minimise duplication of effort and to increase cost effectiveness Continue development of regional and global data monitoring and control systems Continue publicising AMDAR to all countries so that they may be aware of the availability of data and its operational usefulness END