Influence of ocean surface wind observations on numerical forecasts of hurricanes using an ensemble Kalman filter

Ocean surface vector wind observations are a useful data source for numerical prediction of tropical cyclones. The effect of assimilating surface observations on the numerical prediction of hurricanes is examined in this study. First, a series of numerical experiments are conducted to examine the impact of ensemble Kalman filter assimilation of surface observations on the prediction of the landfalls of Hurricane Katrina (2005). It is found that the assimilation of both surface Mesonet observations over land and QuikSCAT ocean surface vectors can improve the prediction of Katrina's track and structure through modifying low-level thermal and dynamical fields such as wind, humidity, and temperature and through enhanced low-level convergence and vorticity. Obvious enhancements are also found in the forecasts of intensity, realistic convection, and surface winds. More importantly, surface data assimilation results in significant improvements in quantitative precipitation forecasts (QPF) during Katrina's landfalls.

The numerical results are further applied to study the air-sea and land-atmosphere interactions and their influence on Katrina's intensity and structure changes. In addition, the proposed observing system simulation experiments for NASA CYGNSS ocean surface wind observations and their relationship to current studies will be reported.