1. SIGMA: Diagnosis and Nowcasting of In-flight Icing – Improving Aircrew Awareness Through FLYSAFE Christine Le Bot Agathe Drouin Christian Pagé

2. Outline  Quick overview of the FLYSAFE PROJECT  Development of icing diagnostic product  Current SIGMA Icing Diagnostic  Algorithm description  Evaluation performed and to be performed  New SIGMA algorithm  Overview  New input data  Conclusions and further developments

3. FLYSAFE: an Integrated Project  Full title: Airborne Integrated Systems for Safety Improvement, Flight Hazard Protection and All Weather Operations  Integrated Project of the Sixth Framework Programme of the European Commission, budget of 53 M€, 29 M€ EC funded  Aimed at designing a Next Generation Integrated Surveillance System (NG ISS):  A generation further than the emerging integrated safety systems  Started on 1 February 2005  Duration: 4 years  36 Partners / 14 countries

4. FLYSAFE Partners

5.  FLYSAFE will address these three types of threats:  Traffic collision  Ground collision  Adverse Weather Conditions  Threats information  Aircrew and Air Traffic Controllers FLYSAFE: addressing Threats

6. FLYSAFE: Meteorological aspects  To improve severe hazards identifications and forecasts  Main meteorological hazards are addressed in FLYSAFE  Clear air turbulence  Wake vortices  Thunderstorms  In-flight icing

7. SIGMA : Context  In-flight icing induces  Hazards for aircraft safety  Significant impacts on accidents and delays A tool to detect icing areas was needed  SIGMA  Icing diagnostic tool  Developed for forecasters  SIGMA Current developments  Adaptation for direct use by aircrew (in the scope of the FLYSAFE project)  Use of data from new meteorological instruments

8. Decision Tree SIGMA Icing Diagnostic System : DATA FUSION Satellite Numerical model forecast Cloud Top Temperature Icing Index Meteosat-8 IR channel 15min / 4km 2D composite 15min / 1km Meteo-France NWP model Arpege 3h / 25 km Aladin 3h / 11km Radar Radar Reflectivity Icing risk Optional Vertical Motion threshold

9. SIGMA : Numerical Index

10. SIGMA : Infra Red Image

11. SIGMA : Radar reflectivity

12. SIGMA : Types

13. SIGMA Output Light (light icing) Moderate (moderate icing) High (moderate to severe icing) Icing risk 3D output + maximum in the vertical Covers France and neighbouring countries 1 km horizontal resolution Output refreshed each 15 minutes

14. SIGMA Output: 2D Objects Light (light icing) Moderate (moderate to heavy icing) High (severe icing) Icing risk

15. Validation – 1 Evaluation performed against :  Pilot reports from French Air Force Army  Observed Radiosoundings analysed by forecasters  Routine Pilot Reports (PIREPS) over USA  Flight tests from AIRS2 campaign in Canada Difficult to perform a real-time verification of icing conditions  No automated observation system  No PIREPS database over Europe  TAMDARs ? Plans for new evaluations  Collect of pilot reports from a civil pilot school « private database » started on 1st Nov 2007  FLYSAFE Flight test campaign : 80 flight hours  feb08  Aug 08

16. First Conclusions… First evaluations performed have shown that data fusion is an appropriate methodology to improve the identification of icing areas. Data fusion enables to extract the most relevant information from each kind of data. But… improvements can be done

17. Evolution… to the New Algorithm In-flight ICING Meteosat-8 & 9 Icing Cloud Product Cloud Top Pressure Cloud Types 4 km resolution AROME Model High resolution Cloud identification Microphysics 2.5 km resolution 3D composite Precipitation occurrence Microphysics (polar.) Melting layer altitude 1 km resolution Satellite Numerical modelGround Weather Radar Meteosat-7 Cloud-Top IR Temperature 6 km resolution ARPEGE Model French Icing Index 25 km resolution 2D composite Precipitation occurrence 1 km resolution OLD NEW

18. Evolution… to the New Algorithm Advanced multi-channels Meteosat geostationary satellite  Integrated in NWP model  PRODUCTS:  Cloud Types  Non-cloudy areas  No icing risk areas  Convective clouds (evolution)  Icing Clouds (NCAR algorithm -evolution)  Identify cloud top phase  Cloud Top Pressure & Temperature  Identify top of cloud layer  Eliminate too warm clouds

19. Evolution… to the New Algorithm New high-resolution numerical weather prediction model AROME Météo-France new high-resolution model Operational in 2008 2.5 km horizontal resolution over France Forecast output : 1h time step Explicit cloud microphysics Prognosis of hydrometeors  Better spatial and temporal resolution matching with satellite and radar data  Better representation of mesoscale features  Radar and satellite data help to correct for defaults in model convective assessment at mesoscale

20. Evolution… to the New Algorithm Melting Layer Identification 0°C altitude Melting Layer Identified  Mixed Layer No Melting Layer  convection, supercooled rain Ground Based Weather Radar

21. Evolution… to the New Algorithm Polarimetric Information  Precipitation Typing  Microphysics Ground Based Weather Radar

22. The New Algorithm Meteosat-8 & 9 Icing Cloud Product Cloud Top Pressure Cloud Types 4 km resolution In-flight ICING Satellite Numerical modelGround Weather Radar AROME Model High resolution Cloud identification Microphysics 2.5 km resolution 3D composite Precipitation occurrence Microphysics (polar.) Melting layer altitude 1 km resolution SIGMA ICING DIAGNOSTIC over Terminal Manoeuvring Area

23. Conclusions  The « DATA FUSION » methodology gives good results  Icing diagnostics have been adapted to be sent directly to aircrew and Air Traffic Controllers (FLYSAFE)  Development of a new algorithm was possible owing to the availability of better spatial and temporal resolutions of the new generation products  More evaluation data will be needed to correctly assess the new upcoming algorithm version –Private Pilot reports database –Two flight tests campaigns within the FLYSAFE project  Further developments : from diagnostic to forecast

24. Questions ?