Impact of Integrated Satellite Observations on Forecasts of Tropical Cyclones: A Case Study

Satellite measurements including Atmospheric Motion Vectors (AMVs), surface winds, temperature and water vapor profiles, and Total Precipitable Water (TPW) data provide extensive information about tropical cyclones and their environment. However, it has been challenging to use these data effectively to improve analyses and forecasts of tropical cyclones. One of the challenges is that the forecast errors are heavily flow-dependent with significant multivariate covariances; this problem is particularly severe for water vapor at mesoscales in the tropics. This challenge can be addressed by using advanced ensemble data assimilation techniques that use short-range ensemble forecasts to estimate the forecast error covariance.

In this study, hourly and rapid scan AMVs, AIRS Infrared single field of view temperature and water vapor soundings, and AMSR Microwave TPW data are assimilated for tropical cyclone track and intensity analysis and prediction using the NCAR mesoscale WRF Data Assimilation Research Testbed (DART) system. Track errors in analyses and forecasts of tropical cyclones Sinlaku and Ike (both 2008) are greatly reduced when these data are assimilated. Analyses and forecasts of the intensification of the cyclones are also significantly improved.