Topographic beam blocking is a problem at many radars across the United States. Large azimuthal sections of some radars are blocked entirely in mountainous areas. However, many radars are affected by partial beam blocking, in which some percentage of the radar beam energy is blocked by the terrain. There, enough energy makes it past the terrain to allow useful, albeit biased, energy returns to the radar.

Even in areas where less than 50% of the beam is blocked, significant reflectivity errors (up to 3 dB) occur. At moderate reflectivities, short-term rain rates are not seriously affected. However, using the typical National Weather Service reflectivity-to-rainfall (Z-R) relationship, at reflectivities near 50 dBZ, a 2 dB error can cause rainfall underestimates around 17 mm hr^-1, and a 3 dB error can cause rainfall underestimates greater than 24 mm hr^-1 (almost 1 in. hr^-1). Rainfall errors double at reflectivities above 55 dBZ. Such errors in a typical flash flooding event, in which heavy rain may last more than one hour, could cause radar operators to miss rainfall accumulations which trigger flash flooding.

The Blocking of Radar by Elevation and Azimuth Model (BREAM) is designed to correct radar reflectivities in areas affected by partial beam blockage, allowing more uniform estimates of rainfall over an area for a given Z-R relationship, and more accurate estimates of rainfall in general. The BREAM model is rather simple, assuming a uniform radar beam with a circular cross-section, standard propagation, and topographic information from the GLOBE Digital Elevation Model (DEM). Since actual blocking and associated reflectivity errors are only calculated once for a given radar, computation times for real-time reflectivity corrections would be manageable. This could allow implementation of BREAM into local National Weather Service offices, as part of the Advanced Weather Information Processing System (AWIPS), or as a stand-alone program on any computer.

The process of using BREAM to determine beam blocking and subsequent reflectivity loss, by azimuth and range, will be demonstrated. The blocking patterns computed by BREAM for example radars will be shown and compared to actual blocking patterns at those radars. Case studies of corrected reflectivity and radar rainfall estimates will be presented for heavy rain events, and the results will be compared to actual rain-gauge and one-minute ASOS observations. Finally, an algorithm in BREAM which computes, by azimuth, the minimum elevation angle at which no beam blocking occurs for a given radar will be presented. In cases of partial beam blocking, it will be shown that increasing radar elevation angle by only 0.5 degrees can eliminate beam blocking altogether at up to 90% of azimuthal directions experiencing blocking.