Probabilistic Quantitative Precipitation Estimates with Ground-Based Radar Networks

The uncertainty in radar quantitative precipitation estimation (QPE) is needed at fine spatiotemporal scales (e.g. 1-km/5-min) for applications such as hydrological modeling, storm prediction, and flash flood monitoring. These applications require more than just one deterministic “best estimate” to adequately cope with the intermittent, highly skewed distribution that characterizes precipitation. We propose to advance the use of uncertainty as an integral part of radar QPE across the conterminous US with the NOAA/NSSL Multi-Radar/Multi-Sensor. Probability distributions of precipitation rates are computed instead of deterministic values using models quantifying the relation between radar reflectivity and the corresponding “true” precipitation. This approach conditions probabilistic quantitative precipitation estimates (PQPE) on factors such as the precipitation rate and typology. It preserves the fine space/time sampling properties of the sensor and integrates sources of error in QPE and the impacts of algorithms assumptions. Precipitation probability maps compare favorably to the deterministic QPE. The PQPE approach is shown to mitigate systematic biases from deterministic retrievals, quantify uncertainty, and advance the monitoring of precipitation extremes. It provides the basis for precipitation probability maps for early warning and mitigation of hydrometeorological hazards, and hydrological modeling.