Combinations of CML wet/dry corrections based on MSG SEVIRI data in Germany

Data from commercial microwave links (CML) operated by mobile network providers is attenuated by precipitation. The line integral of the precipitation intensity can be calculated from this attenuation and is used as a complement to common precipitation measurements. A major issue in CML signal processing is the distinction between dry and wet time steps. In addition to time series-based methods, precipitation patterns based on weather radar or satellite are also used.

One aim of our work is to evaluate the suitability of MSG SEVIRI data as a wet/dry indicator. The advantage of SEVIRI is the almost global coverage with a resolution of about 4x6 km in the middle latitudes, every 15 minutes. We apply two products from NWCSAF that derive the probability of precipitation from a combination of SEVIRI channels. Our study is based on 3901 CMLs in Germany for the summer of 2021. We evaluate the performance differences at various precipitation intensities as well as between daytime and nighttime since fewer SEVIRI channels are available at night. Another aim of our work is to investigate the possibilities of improvements by a combination of time series and SEVIRI-based wet/dry detections.

Our results show at least equivalent performance of SEVIRI compared to common time-series-based methods. The Matthews Correlation Coefficient (MCC) values for combinations are even better. This is especially the case for light precipitation. In addition, there are also no deteriorations for individual CMLs compared to common methods.