Traduction auto.

frendeitrues

Evènements

Le jeudi 7 juin 2018 à 11h

LATMOS site Guyancourt
Amphithéâtre Gérard Mégie, Observatoire Versailles St Quentin
===============================================================

Probabilistic Quantitative Precipitation Estimates with Ground- and Space-based Remote Sensing

Pierre-Emmanuel Kirstetter

National Severe Storm Laboratory, University of Oklahoma (USA)

Progress in precipitation estimation is critical to advance weather and water budget studies and prediction of natural hazards caused by extreme rainfall events from local to global scale. An interdisciplinary challenge in remote sensing, meteorology and hydrology is the impact, representation, and use of uncertainty. Understanding of hydrometeorological processes and applications require more than just one deterministic "best estimate" to adequately cope with the intermittent, highly skewed distribution that characterizes precipitation. Yet the uncertainty structure of quantitative precipitation estimation (QPE) from ground-based radar networks like NEXRAD and satellite-based active and passive sensors of the Global Precipitation Measurement (GPM) mission is largely unknown at fine spatiotemporal scales near the sensor measurement scale (e.g. 1-km/5-min for ground-based radars, 5-km/instantaneous for space-based radars). We propose to advance the use of uncertainty as an integral part of QPE for ground-based and space-borne sensors. Probability distributions of precipitation rates are computed instead of deterministic values using models quantifying the relation between the sensor measurement and the corresponding "true" precipitation. This approach preserves the fine space/time sampling properties of the sensor and integrates sources of error in QPE. It provides a framework to diagnose uncertainty when instruments sample raining scenes or processes challenging the assumptions of the QPE algorithms. Precipitation probability maps compare favorably to deterministic QPE and improve precipitation estimation. Probabilistic QPE is shown to mitigate systematic biases from deterministic retrievals, quantify uncertainty, and advance the monitoring of precipitation extremes. It provides the basis radar and satellite precipitation ensembles needed for multisensor merging of precipitation, early warning and mitigation of hydrometeorological hazards, and hydrological modeling. Perspectives for improved understanding and parameterizations of precipitation processes, estimation of precipitation at multiple scales, hydrological prediction and risk monitoring will be presented.