A NAVGEM Data Assimilation Study Using Comparative ASCAT Experiments

The NAVy Global Environment Model (NAVGEM) assimilates ASCAT Ocean Surface Wind Vector (OSWV) retrievals and SSMIS Ocean Surface Wind Speed (OSWS) retrievals.  In a previous study, we examined the differences between assimilating ASCAT as ocean surface wind vectors versus ocean surface wind speed (OSWS) as measured by forecast skill and Forecast Sensitivity Observation Impact (FSOI).  From a quick glance, the FSOI from ASCAT wind vector assimilation was expected to be twice as large for OSWS assimilation purely because there two pieces of information (speed and direction vs. speed) and twice as many observations.  However, we found that the FSOI from the ASCAT OSWS was significantly less than 50% of the ASCAT OSWV.  The results from that study suggest that the NAVGEM 4DVAR data assimilation of ocean surface winds needs to be revisited to adequately account for differences in sensor quality and resolution, as well as assess 2DVar pre-processing of ocean surface wind speed to generate wind vectors.  In this study, we revisit the NAVGEM 4DVAR assimilation of ocean surface winds as part of our preparation for the assimilation of CYGNSS ocean surface wind speed retrievals.  The NASA CYGNSS mission consists of a constellation of eight GPS-R microsatellites to measure ocean surface wind speed in the tropics; therefore, methods that convert CYGNSS wind speed to wind vectors will be explored to maximize NAVGEM forecast improvement.