A Novel Physical Consistency-Based Calibration Tool for Polarimetric Weather Radar

The calibration of radar reflectivity and differential reflectivity is essential for accurate quantitative precipitation estimation (QPE) using polarimetric weather radar. Mature calibration methods, which are based on a standard signal source, routine sun tracking, “bird bath” scanning, etc., have been widely used in the weather radar community. However, those methods are generally off-line calibration techniques and may interrupt the operational data collection. In addition, the calibration should be done frequently because the calibration term might vary with time. It is also noted that the calibration term might vary in radar sweeps with different elevation angles. Using conventional calibration techniques, frequent calibration becomes impractical. An automatic on-line calibration could overcome these limitations and would be desirable by the operational weather radars, especially within a regional or national network.

Recently, Enterprise Electronics Corporation (EEC) in Alabama, collaborated with the U.S. National Severe Storms Laboratory (NSSL) to develop a novel automatic calibration tool (ACAL), which facilitates on-line calibration for S-band and C-band polarimetric weather radars. The ACAL algorithm is based on the physical consistency among polarimetric radar measurements of precipitation, i.e., reflectivity, differential reflectivity, and differential phase. With the continuous execution of ACAL, the radar system generates the automatically-calibrated reflectivity product in real-time.

The current article introduces the theoretical basis of the ACAL algorithm and the ACAL module used in EEC's radar software system—EDGE. The ACAL performance is demonstrated by the case studies with EEC's C-band polarimetric weather radars with various transmitter configurations. The potential uncertainty of calibration is also discussed. Results show that ACAL has a robust calibration performance based on radar measurements of different precipitation events. The variance of calibration term is relatively small as compared to the typical uncertainty of radar measurements.