13th Conference on Mesoscale Processes

5.2

A Prototype future hurricane prediction system: Realtime cloud-resolving ensemble data assimilation and forecasting during the 2008 Atlantic season

Yonghui Weng, Texas A&M University, College Station, TX; and F. Zhang, J. Gamache, and F. D. Marks

Hurricanes are among the costliest and deadliest nature disasters whose forecast accuracy is ever in demand but remains highly uncertain. This study explores a prototype future hurricane prediction system that performs cloud-resolving ensemble data assimilation and forecasting in massively parallelized high-performance computing facilities with assimilation of high-resolution airborne radar observations during the 2008 Atlantic hurricane season. It represents the first time that airborne Doppler radar observations are assimilated into hurricane prediction models, and the first time that the cloud-resolving ensemble analyses and forecasts for hurricanes in realtime. Also unprecedented are the realtime coordination, parallelization, and on-demand usage of more than 30,000 computer cluster cores simultaneously. Moreover, besides providing flow-dependent analysis and forecast uncertainty, this prototype system outperforms not only other deterministic model forecasts but also the official consensus forecast issued by the National Hurricane Center for the limited number of cases tested in realtime. Plans are in place to run this prototype system in realtime assimilating high-resolution in situ and remotely sensed observations for the 2009 Atlantic hurricane season under the auspice of the NOAA's HFIP project.

Session 5, Transferring research results to operations
Monday, 17 August 2009, 5:00 PM-6:00 PM, The Canyons

Previous paper  Next paper

Browse or search entire meeting

AMS Home Page