An Examination of the Gage-Radar Relationship with respect to Frontal Boundaries

Accurate estimates of precipitation are essential in many real world applications. Since the advent of weather radar, many users utilize radar data to study and assess real-time precipitation conditions over their areas. Wilson and Brandes (1979) point out the utility of using rain gage data in addition to radar data to accurately measure areal precipitation. Many studies (Calheiros and Zawadzki 1987, Rosenfeld 1994, Crosson et al. 1996, to name a few) have subsequently adopted various ways of adjusting the radar measurements of precipitation based on the determination of a statistical relationship between the gage and radar's respective precipitation distributions. In all of these cases the gage-radar relationship was examined from a statistical perspective, and the adjustment was applied uniformly across the entire radar region.

This study attempts to demonstrate the utility in separating a radar scene into different zones of gage-radar pairs, and subsequently different precipitation processes, based on their proximity to significant frontal boundaries. Since real-time studies are limited to data with high temporal resolution, WSR-88D radar data and NWS and Oklahoma Mesonet meteorological data from the central Oklahoma region were used to examine this relationship. An objective frontal analysis technique was used to consistently assess the location of potential frontal boundaries in spatially interpolated equivalent potential temperature (Θ_e) fields. These boundaries were then used to segment the gage-radar pairs into pre-frontal and post-frontal zones. Overall, this method appears to be useful in separating a radar scene into similar precipitation processes based on proximity to frontal boundaries and has potential for use in real-time radar calibration.