This article presents the evaluation of several polarimetric quantitative precipitation estimation (Pol-QPE) algorithm candidates for operational implementation in the Météo-France polarimetric weather radars (nine C-band and one S-band out of a total of 24 radars in 2011). Since the performance of the algorithms depends on both the quality of the polarimetric variables and the applied polarimetric processing, the article first describes the current polarimetric processing chain and then provides a comprehensive analysis of the quality of the polarimetric variables.
Three types of QPE algorithms have been studied at C-band in ideal conditions (short ranges, no partial beam blocking, no vertical profile of reflectivity effects):
(i) algorithms based on simple Z-R relationships with and without attenuation correction using the differential phase Φdp,
(ii) an algorithm based on reflectivity (Zh) and differential reflectivity (Zdr), and
(iii) algorithms based on specific differential phase (Kdp).
Evaluation against rain-gauges, performed hourly, confirms the superiority of polarimetric algorithms as repeatedly reported in the literature. It also shows that the benefits brought by algorithms based on power variables, Zh and Zdr, are critically dependent upon the capability to correctly calibrate them. Long-term monitoring shows that, with the current radar systems and operational calibration procedures, a stability of ±0.4 dB is achieved on Zdr. In that context, QPE algorithms based on Kdp, which are considered to be immune to partial beam blocking, calibration, attenuation, etc., are preferred for operational use until improvements are made to the calibration of Zdr. To this end, a hybrid algorithm based on attenuation-corrected Zh for low rain rates and Kdp for higher rain rates has been designed and tested Copyright © 2012 Royal Meteorological Society