Improvements to NAVDAS 3DVAR Analysis for COAMPS

The NRL Atmospheric Variational Data Assimilation System (NAVDAS) is a three dimensional variational data assimilation system, which is currently used by the Fleet Numerical Meteorology and Oceanography Center (FNMOC) in its operational mesoscale numerical weather prediction system (COAMPS). In the past year, improvements have been made, expanding the number and type satellite wind observations and humidity retrievals assimilated by COAMPS, and changing the supper-obbing strategy used. Assimilation of AMSU/A radiances is being tested for future operational implementation. Feature-track winds from MODIS and AVHRR have greatly increased the number of observations used in polar areas. Surface marine winds from the ASCAT scatterometer and the WINDSAT passive microwave radiometer have increased the coverage of wind observations used over the oceans. For the satellite winds, different averaging and thinning strategies were tested. Best results overall were achieved by averaging the satellite wind observations in one degree horizontal boxes, to produce the super-obs. In the tropics, better definition of the flow environment in the vicinity of tropical cyclones lead to improved track forecasts in certain cases. Total precipitable water (TPPW) retrievals from SSM/I, SSMIS and WINDSAT are also super-obbed and used to create profiles of retrieved pseudo-relative humidity (ratio of mixing ratio of observation and saturation mixing ratio of background). At locations of the super-obbed TPPW observations, background humidity profiles are adjusted to match the vertically integrated humidity to the TPPW observation. To dry a background humidity profile, the mixing ratios are reduced in proportion to their magnitude; while to moisten a profile, mixing ratios are increased in proportion to their saturation capacity, given by the difference of the saturated and unsaturated values of the background mixing ratio. Noticeable improvements are found to the analyzed humidity in the region of tropical cyclones. Initial testing of the assimilation of AMSU/A radiances have shown promise for replacing the use of AMSU/A temperature retrievals in the COAMPS application of NAVDAS. Bias corrections generated by the global NOGAPS model are used to correct the radiance bias for use with COAMPS. Testing on the satellite upgrades is being conducted for several different regional areas, and the impact of the improvements will be presented.