Feasibility of Using Mountain Return for the Correction of Ground-Based X-Band Weather Radar Data

The Marseilles Hydrometeorological Experiment has been designed to improve rainfall measurement techniques over the space scales and timescales useful for urban hydrology applications. Among other sensors, an X-band light configuration weather radar system was set up near the city. The main objective of the present paper is to test the feasibility of using mountain return to correct rain attenuation effects, an application of the well-known “surface reference technique” developed for spaceborne radar configurations. The radar siting and scan procedures were defined so as to obtain both (i) rain reflectivity measurements free of ground detection over a large domain and (ii) strong mountain return for the path-integrated attenuation (PIA) estimation in a reduced azimuthal sector. The so-called PIA constraint equation is proposed to relate the measured rain reflectivity profile in one direction to the corresponding mountain PIA. A study of this equation for the 23 September 1993 rain event provides valuable information concerning (i) the parameterization of the radar data processing and (ii) the comparison of various models proposed for the estimation of the mountain and blind-range PIAs. Three attenuation correction algorithms, already proposed in the literature, are then reviewed. Two of them make direct use of the mountain PIA, while the other is a modified version of the well-known Hitschfeld and Bordan algorithm. A consistency test between the various sources of rainfall data available for the considered rain event is presented. First, disdrometer drop size distribution measurements are modeled and used to derive the (Z, k, R) relations required for radar data processing. Then, a validation procedure is used to test the effectiveness of the attenuation correction algorithms by comparison with rain gauge measurements. The performance characteristics of the three algorithms are basically equivalent—a result to be considered with respect to the range of PIAs observed during the rain event (0–20 dB)—and the overall consistency of the various sources of rainfall data used is found to be satisfactory. Further improvements could be obtained from (i) a better understanding of the blind-range attenuation effects and (ii) a higher radar data sampling rate.