Operational Adjustment of Z-R Relation Coefficients for Radar Rainfall Accuracy Improvement by Dual-Wave Attenuation Measurements

We present an operational method for correcting of Z-R relations coefficients A and b to its actual value by the dual-wave attenuation measurements.

In Moscow region the radar rainfall measurements are carrying based on two-wave (X- and S-band) digital weather radar MRL-5. For rainfall rate calculations we use simplest Z-R Marshall-Palmer relation Z=200*R^1.6. Our regular analysis of radar (S-band) and gauge accumulated rainfall data shows satisfactory agreement with diurnal precipitation root-mean-square difference of 30-40% on average, except for some situations when the errors may exceeds factor 2 and significant attenuation effect on X-band is observed. Probably the deviation of drop size distribution characteristics from Marshall-Palmer values is the main reason of such excesses.

In contrast with the attempts to specify DSD parameters using precipitation type separation of events we stablished the relationship between measured attenuation and averaged A and b coefficients in Z-R relation in this algorithm. We use a least squares regression procedure to identify the interesting parameters. From a posteriori rainfall analysis it was defined that minimal scattering and maximal correlation radar and gauge-measured daily precipitation amounts correspond to b=1,5-1,6 value and therefore at first step we adjust A coefficient only. The limitation and details of the method concerning the account for antenna diagram mismatching on X- and S-channels, polarization effects and so on, are discussed.

This method was applied to some heavy precipitation events and demonstrated essential reduction of errors in radar rainfall measurements.