Data Assimilation in a high-resolution, sub-mesoscale regional model of Hawaii

We utilize the incremental strong constraint four-dimensional variational data assimilation (IS4DVAR) method to study the circulation and dynamics around the Hawaiian Islands. In this system we assimilate observations from satellite radiometry and altimetry, as well as in situ data from Argo floats, surface drifters, autonomous gliders, and shipboard CTD (conductivity, temperature, depth) and ADCP (acoustic Doppler current profiler) instruments. We present results from a 2-year spin-up of the IS4DVAR model performed in preparation for the operational real-time system. Model-observation difference is reduced by over 70%; however, large reductions in misfit are only seen in sea surface temperature (SST). This is due to the vast difference in the relative number of observations of the various products, with satellite SST accounting for over 99% of all observations assimilated. Results are presented from experiments that focus on increasing the contribution of non-satellite SST observations.

(Dax Matthews and Brian Powell, School of Ocean and Earth Science and Technology, University of Hawai'i at Manoa)