Canadian Radar Quality Control in Multi-Radar-Multi-Sensor System

Environment Canada deployed an operational weather radar network ~1999–2003 and has been providing real-time base-level data from ~30 C-band single-polarized radars. The data is integrated with the United States NEXRAD data in the Multi-Radar Multi-Sensor (MRMS) system for various severe weather and precipitation product generations over the CONUS and southern Canada. A key process before the integration of the data is the quality control (QC) that removes non-meteorological echoes in the reflectivity observations. This paper presents a QC scheme for the Canadian radar's full volume scan (“CONVOL”) data, which only contains reflectivity data and not velocity and spectrum width observations. The QC involves two-steps: 1) to remove persistent ground clutter via statistical clutter maps; 2) to remove anomalous propagation (AP) clutter via an image processing technique based on intensity, texture, and vertical variations of reflectivity. The QC scheme was tested on a large number of cases and performed very well in removing persistent ground clutter and a majority of AP clutter. AP clutter mixed with precipitation, especially at far ranges, remains a challenge for the QC due to the limitation of available observations currently.