Grid computing for Civil Protection An application about flash flood forecasting: first results and lesson learnt Ayral P.-A. 1 Thierion V. 1 , Angelini V. 2 , Mazzetti P. 2 , Nativi S. 2 , Sauvagnargues-Lesage S. 1 1 : Ecole des Mines d’Alès (EMA, France) ISGC Symposium 2 : Instituto di Metodologie per l’Analisi Taipei, 9-12 Marsh 2010 Ambientale (CNT-IMAA, Italia)
Outlines How Grid Technology enhance Civil Protection functioning such as Flash Flood crisis - Outlines management ? - CYCLOPS - Flash flood 1- Grid computing for Civil Protection ( CYCLOPS ) - G-ALHTAÏR - Lesson learnt 2- Flash flood crisis management context - Conclusion 3- G-AHTAÏR 4- Lesson learnt 5- Conclusion bigagli@imaa.cnr.it
Grid computing for Civil Protection Which Modalities - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR Bridging the gap between the two communities and requirements gathering : - Lesson learnt - Conclusion 1. Understand Civil Protection requirements in term of innovative technologies 2. Define grid technological improvements according to operational requirements 3. Development of a grid-enabled architecture for CP issues 2 Case studies about forest fires and flash floods bigagli@imaa.cnr.it
1/6 Flash flood crisis management context – The phenomenon - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
2/6 Flash flood crisis management context – The last event: September 2002 Rainfall situation 1. Rapid phenomena - Outlines ⇒ less than a half of day - CYCLOPS - Flash flood 2. Important accumulation ⇒ about 600 mm - G-ALHTAÏR Gard Region - Lesson learnt 3. Large extent - Conclusion ⇒ About 5000 km ² ⇒ 22 victimes in the Gard Region ⇒ 120 millions € of damages bigagli@imaa.cnr.it
2/6 Flash flood crisis management context – The last event: September 2002 Le Pont du Gard - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
2/6 Flash flood crisis management context – The last event: September 2002 Sommières - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
2/6 Flash flood crisis management context – The last event: September 2002 Le Pont de Ners - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
Flash flood crisis management 3/6 context – French Civil Protection for flood hazard management - Outlines - CYCLOPS Multi-scales - Flash flood Organization - G-ALHTAÏR - Lesson learnt - Conclusion Multi-partners organization Flash flood use-case developed in the framework of Cyclops project as grid adoption faisability study for Civil Protection whole functioning bigagli@imaa.cnr.it
Flash flood crisis management 4/6 context – Grand Delta Flood Warning Service (SPC-GD) X 170 stations and 20 watersheds - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion X 10 Zones (64 x 64 km) Mediterranean sea 100 km bigagli@imaa.cnr.it
5/6 Flash flood crisis management context – SPC-GD Information System - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
6/6 Flash flood crisis management context – SPC-GD Information System - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt Hydrological - Conclusion modelling and Radar rainfall data Geospatial Information System bigagli@imaa.cnr.it
1/6 G-ALHTAÏR Objectives - Outlines Porting hydrological modeling of SPC-GD - CYCLOPS (existing in SPC-GD) - Flash flood - G-ALHTAÏR Real time simulation - Lesson learnt - Conclusion Developing a flash flood forecasting module (innovative module) Forecast scenario (1h, 2h, 3h, 4h and 5h) bigagli@imaa.cnr.it
2/6 G-ALHTAÏR - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
3/6 G-ALHTAÏR - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
4/6 G-ALHTAÏR - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
5/6 G-ALHTAÏR - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion Potentially, forecaster can launch almost simultaneously more than 600 forecasting simulations bigagli@imaa.cnr.it
6/6 G-ALHTAÏR A Webmapping interface for the forecaster - Outlines - CYCLOPS - Flash flood - G-ALHTAÏR - Lesson learnt - Conclusion bigagli@imaa.cnr.it
1/2 Lesson learnt For forecaster and hydrologist - Outlines 1. Unlimited computational and storage resources - CYCLOPS - Flash flood 2. Ubiquitous functioning corresponding crisis management requirements (field investigation, rescues operations, - G-ALHTAÏR continuous decision making process, shared knowledge, …) - Lesson learnt - Conclusion 3. Operational solution to handle well-known limitations of short- term meteorological prediction (few hours) 4. Give an integrated framework to gather data, tools and skills bigagli@imaa.cnr.it
2/2 Lesson learnt Further research - Outlines 1. Dynamic workflows composition of simple OWS to handle - CYCLOPS more complex tasks - Flash flood 2. Gather the whole of crisis management actors around the - G-ALHTAÏR designed e-collaboration platform G-ALHTAIR - Lesson learnt - Conclusion 3. Design a grid use and data policies in order to guarantee fast responses of the operational services and consequently grid resources unpredictable access. bigagli@imaa.cnr.it
Conclusion Grid Technology Flash-Flood crisis management - Outlines - CYCLOPS - Flash flood Storage capacities Computational Power - G-ALHTAÏR 1. A Virtual Organization for all 1. Calibration models in post and - Lesson learnt Civil Protection actors real-time - Conclusion 2. Improve geospatial services to 2. Monitoring and forecasting a better reliability of the crisis large-scale hydrological management situation in crisis phase 3. Integrate the whole of data 3. Design a real-time and information for lessons hydrological modelling learnt process (operational platform for models inter- and scientific aspects) comparison bigagli@imaa.cnr.it
Thanks for your Attention bigagli@imaa.cnr.it
Grid computing for Civil Protection An application about flash flood forecasting: first results and lesson learnt Ayral P.-A. 1 Thierion V. 1 , Angelini V. 2 , Mazzetti P. 2 , Nativi S. 2 , Sauvagnargues-Lesage S. 1 1 : Ecole des Mines d’Alès (EMA, France) ISGC Symposium 2 : Instituto di Metodologie per l’Analisi Taipei, 9-12 Marsh 2010 Ambientale (CNT-IMAA, Italia)
RRM-GRID SPC-GD Local system User Interface Grid infrastructure bigagli@imaa.cnr.it
RRM-GRID 105 tasks simulating 1, 2, 3, 4 and 5 hours of forecast horizons with variations from -50% to 50% with 5% step. bigagli@imaa.cnr.it
Recommend
More recommend