Rainfall detection by tomographic inversion of commercial radio link data: a pilot project in Italy

Within MOPRAM project, we investigate the use of a network of microwave radio links to a) improve the accuracy and reliability of precipitation fields detection and rainfall estimation and b) to enhance the knowledge of river basins answer, and in particular the prediction of river discharge, based on the availability of more accurate space-time precipitation measurements. A tomographic technique is applied for the first time to reconstruct rain fields from link data collected by the backhauling network of a major mobile phone company operating in Italy. The raw data are a network's by-product, available at almost no-cost, once the network software is properly updated. The tomographic method has to be fine-tuned on the network's topology, to be stable even in the presence of noisy measurements and perform consistently. The proposed methodology could guarantee more accurate information about the rainfall fields with respect to weather radars (lacking in terms of quantitative reliability) or rain gauges (lacking in terms of spatial resolution).