Normalization of NEXRAD antenna gain for overlapping radars

The NOAA National Weather Service operates a network of NEXRAD Doppler meteorological radars to measure severe weather, and precipitation is inferred from the volume scattering from hydrometeors. The NEXRAD, WSR-88D radar, is a calibrated system that measures absolute volume reflectivity (dBZ). A geophysical algorithm is used to relate rainfall intensity to the measured dBZ. The NEXRAD system comprises overlapping radars to provide nearly complete coverage over the continental US. After installation and certification, absolute radar calibration for the NEXRAD is maintained by a loop-back calibration that samples the transmitter through a known attenuation into the receiver and processor; however there is no provision to recalibrate the antenna gain and radome losses. Without an external calibration any NEXRAD may have biases that can cause significant errors in the precipitation products. The problem of combining multiple radars to provide extended coverage is not new. This paper presents an improved method that uses low-elevation radar reflectivity scans of precipitation events taken asynchronously in time. It is applied to two NEXRAD radars located in Melbourne (KMLB) and Miami (KAMX), Florida. Reflectivity data analyzed for two precipitation events indicate consistent, time-invariant dBZ biases.