Operational mitigation of ground clutter using information from past and near-future radar scans

Weather radar scans are peppered with areas of incorrect or unreliable data, generally due to effects such as attenuation, echoes from non-meteorological targets, and the presence of ground clutter. This "bad" data will result in errors and inaccuracies in each individual scan, but also in the resulting hydrological products such as rainfall accumulations. In order to mitigate this propagation of error, these gaps must be filled using an algorithm to derive an approximation of the true state of the atmosphere.

This work introduces one such algorithm that corrects areas of radar ground clutter in near-real time by making use of unobstructed data at different heights and horizontal locations, but also at earlier and (if possible) later times. The resulting improvement of individual scans as well as cumulative hydrological products illustrates the benefits of a fully four-dimensional approach to filling voids in radar data.