Uncertainty Quantification of Precipitation Measurement with Weather Radar

Extreme rainfall events can cause flash floods and are responsible for socioeconomic damage worldwide. In Campinas, southeastern Brazil, countless events take place throughout the year. In order to monitor and predict these events, with the support of Fapesp's SOS-Chuva project, a mobile rainfall radar was installed in the region. With the purpose to identify the accuracy of this data, the radar data were compared with rain gauge data. Through this study, it is noted that, at some points, the difference between the rain gauges measurements and the radar data is significant, which may hinder the calibration and performance of a rainfall-runoff hydrological model. To improve the rainfall measurement considering both data source, this study proposes to combine both information and generate rainfall probabilistic maps, derived from geostatistical methods, thus making possible to quantify the uncertainty of these data.