Improving 3-D reflectivity radar mosaics using an inverse method

Mosaicking volumetric reflectivity observations from multiple radars over a common domain is supposed to improve the quality of Quantitative Precipitation Estimates by reducing the effects of path attenuation by intense precipitation, beam blockage or the vertical profile of reflectivity. So far, composites of radar observations are commonly carried out through simple criteria (by picking the closest observation or the maximum value...) or using quality indices that need a priori definition of quality descriptors. This work proposes a methodology to retrieve the 3-D Cartesian reflectivity field most compatible with the observations from the different radars of the network. The methodology follows the concept of an inverse method based on the minimization of a cost function that penalizes discrepancies between simulated and actual observations for each radar of the network. The simulations are obtained with a model that simulates the radar sampling of the atmosphere, simulating the effects of distance (beam broadening and increasing of the sampling height) and path attenuation by intense precipitation. The model considers the specific characteristics of the radars (location, scanning strategy, frequency, beam width, pulse length...). The methodology has been applied to the network of C-band radars in the vicinity of Barcelona, Spain. The retrievals have been obtained from the volumetric reflectivity observations of two radars. The evaluation of the retrievals has been performed with the observations of a network of rain gauges. Also, observations from an independent radar have been used for verification at different heights. Furthermore, the results are compared with those obtained with conventional mosaicking techniques.