Data assimilation (or state estimation) for the hurricane is exceptionally
challenging because of the range of important spatial scales. The
interaction between the larger-scale flow and the hurricane is perhaps the
most crucial. Errors in the larger scales rapidly lead to hurricane track
errors and make even initializing the vortex in the correct location
difficult. Significant displacements of the vortex also lead to
non-Gaussian effects which call into question many assimilation techniques.
I will review progress in the use of ensemble-based, Monte-Carlo techniques
for hurricane data assimilation, particularly as applied to assimilation of
observations of hurricane position.
Audio (MP3 File, Podcast Ready)
Copyright. All Rights Reserved.