Precipitation observation using commercial microwave communication links

Rain rate observation over complex terrain is still afflicted with large uncertainties introduced by system inherent drawbacks of radar and gauge measurements. We use a new method for near surface rain rate estimation exploiting attenuation data from commercial microwave backhaul links. Our test region is the pre-alpine and alpine region of Southern Germany where we record received signal level (RSL) data with minute resolution directly at the communication towers using small data loggers. To dynamically set the RSL baseline from which the attenuation is calculated, a wet/dry classification based on spectral time series analysis is employed. This algorithm is applied to continuous minute resolution RSL data from July 2010 till October 2010 for five microwave backhaul links. The derived rain rates are compared to weather radar and gauge data yielding good correlations of up to R² = 0.84.