Effects of beam broadening and earth curvature in radar data assimilation

The operational Doppler NEXRAD (WSR-88D) radar network is an essential observing system for initializing non-hydrostatic, storm-resolving (i.e., horizontal grid spacing of order 1 km) numerical weather prediction (NWP) models. To assimilate the radar data from NEXRAD radars into an NWP model, it is necessary to consider the effect of beam broadening and to use suitable ray path equations to obtain the physical location of each radar measurement to have accurate forward operators to convert model winds to radial velocity in any data assimilation schemes. Currently, most studies in the radar data assimilation use a very simple straight line ray path equations to model the forward operator that projects the 3D wind fields from NWP model to the radial direction and do not consider the effect of ray broadening (e.g., Gao et al. 1998; Weygandt et al. 2002a,b; Shapiro et al. 2003). This may cause significant errors in radar data assimilation, especially for strong convective weather when radars are far away from connective storms.

This study aims at estimating errors of wind analysis caused by using simplified ray path equations and simplified beam pattern in a 3D variational data assimilation system (3DVAR) with an idealized tonadic thunderstorm case. Our preliminary results show that the error caused by this kind of simplification varies with range, and can be rather large when the surface range between the center of the analyzed storm and the radar location is large. In the other hand, neglecting earth curvature in the radar observation operator also results in considerable error in the 3DVAR analysis within this particular case.