Estimating radar wind observation error and NCEP WRF background wind error covariances from radar radial-velocity innovations

By using the non-isotropic form of error covariance function derived for radial-velocity fields on conical surfaces of low-elevation radar scans, a statistical method is developed based on least-square fitting to estimate Doppler radar radial-velocity observation error covariance and background vector wind error covariance from time series of radar radial-velocity innovation (observation minus independent background) fields. The method is applied to radar radial-velocity innovation data collected from six radars (KINX, KLZK, KSGF, KSRX, KTLX, KVNX) from 11:00 to 23:50 UTC on 05/21/2005 under a clear but windy weather condition. The background wind fields were provided by NCEP WRF 3-hour forecasts (on a 321x161x61 E-grid with a sigma vertical coordinate and 8 km horizontal resolution) over the central US. The results show that radar radial-velocity observation errors are correlated between neighboring range gates and between neighboring beams. The radial-velocity observation error correlations can be quite consistently estimated along with the observation error variances and background wind biases and covariances by the method. The detailed results will be presented at the conference.