Assimilation of radial velocity and reflectivity data from coastal WSR‐88D radars using an ensemble Kalman filter for the analysis and forecast of landfalling hurricaneIke(2008)

Abstract

AbstractEnsemble Kalman filter (EnKF) assimilation and forecasting experiments are performed for the case of HurricaneIke(2008), the third most destructive hurricane hitting the USA. Data from two coastal WSR‐88D radars are carefully quality controlled before assimilation. In the control assimilation experiment, reflectivity (Z) and radial velocity (Vr) data from two radars are assimilated at 10 min intervals over a 2 h period shortly beforeIkemade landfall. A 32‐member forecast ensemble is initialized by introducing both mesoscale and convective‐scale perturbations to the initial National Centers for Environmental Prediction (NCEP) operational global forecast system (GFS) analysis background, and the ensemble spread during the analysis cycles is maintained using multiplicative covariance inflation and posterior additive perturbations. The radar data assimilation results in much improved vortex intensity and structure analysis over the corresponding GFS analysis. Compared with the forecast starting from the GFS analysis, the forecast intensity, track and structure ofIkeover a 12 h period are much improved in both deterministic and ensemble forecasts. Assimilation of eitherVrorZleads to improvement in the forecasts, withVrdata exhibiting much greater impacts thanZdata. With the 2 h assimilation window, 30 min assimilation intervals produced results similar to 10 min intervals, while 60 min intervals were found to be too long. The ensemble forecasts starting from the EnKF analyses are found to be mostly better than the corresponding deterministic forecast, especially after ensemble post‐processing, such as probability matching for precipitation. Precipitation equitable threat scores were calculated and compared. Copyright © 2012 Royal Meteorological Society