Impact of different sources of precipitation data on urban rainfall-runoff predictions: A comparison of rain gauges, commercial microwave links and radar

In this study, we investigate whether commercial microwave links (CML) can bridge the gap between rain gauge and radar. We performed a rainfall-runoff monitoring experiment with rain gauges, CMLs and weather radar in a small urbanized catchment during several months. The results of comparing rainfall data and resulting sewer flows suggest that quantitative precipitation estimates from CMLs require adjustment to local rain gauges to produce satisfactory results in a similar fashion as weather radar. In our study, CML with baseline removal and wet-antenna correction, but without adjustment, lead to substantial bias and underestimate rainfall-runoff ratios by more than 48% (Radar: ca. 40%). In contrast, adjusted CML produced results very similar to those from a dense rain gauge network and underestimated rainfall-runoff by about 11%. Further work is needed to better understand influential factors on the data quality of QPEs from CML. In addition, incentives are needed to motivate both telecommunication companies to provide CML data and weather service companies to use them in their products, i.e. commercializing information from opportunistic sensors.