Coordination Funds

高质量的近实时定量降水估计（QPE）及其未来数小时的预测（QPN）在气象、水文、农业、建筑、供水和下水道系统管理等领域有着重要的应用。特别是对于中小尺度流域洪水和城市强降水的及时预报，高分辨率、高质量的QPE/QPN的价值不容小觑。极化气象雷达由于其覆盖面积和高分辨率观测，为QPE/QPN提供了无可争议的核心信息，可以估计降水强度、水成物类型和风。尽管在这类气象雷达上投入了大量资金，但100多年来，QPE仍然主要基于雨量计测量，没有一个实际的洪水预报系统敢于使用雷达观测来进行QPE。RealPEP将QPE/QPN提升到一个阶段，当用于中小流域的洪水预测时，它可以验证优于雨量计观测。为了实现这一目标，最先进的雷达偏振测量将与来自商业微波链路网络的衰减估计相结合，以改善QPN，来自卫星的对流开始和演变信息以及来自地面网络的闪电计数将被用于改善QPN。随着预报范围的增加，基于观测的临近预报的预测能力迅速退化，并被基于数据同化的数值天气预报（NWP）所优于，然而，由于模式集成和自旋启动需要提前时间，NWP在最初几个小时内失败。因此，RealPEP将基于观测的QPN与NWP相结合，以实现无缝预测，从而提供从观测时间到未来几天的最佳预测。尽管最近在利用分布式、基于物理的模型模拟地表和次地表水文方面取得了进展，但用于业务洪水预测的水文成分仍然是概念性的，需要校准，并且无法客观地消化集水区状态的观测信息。RealPEP将证明，结合先进的QPE/QPN物理水文模型，结合流域状态观测的同化，在小到中尺度流域将优于传统的洪水预报

High-quality near-real time Quantitative Precipitation Estimation (QPE) and its prediction for the next hours (Quantitative Precipitation Nowcasting, QPN) is of high importance for many applications in meteorology, hydrology, agriculture, construction, water and sewer system management. Especially for the prediction of floods in small to meso-scale catchments and of intense precipitation over cities timely, the value of high-resolution, and high-quality QPE/QPN cannot be overrated. Polarimetric weather radars provide the undisputed core information for QPE/QPN due to their area-covering and high-resolution observations, which allow estimating precipitation intensity, hydrometeor types, and wind. Despite extensive investments in such weather radars, QPE is still based primarily on rain gauge measurements since more than 100 years and no operational flood forecasting system actually dares to employ radar observations for QPE. RealPEP will advance QPE/QPN to a stage, that it verifiably outperforms rain gauge observations when employed for flood predictions in small to medium-sized catchments. To this goal state-of-the¿art radar polarimetry will be sided with attenuation estimates from commercial microwave link networks for QPE improvement, and information on convection initiation and evolution from satellites and lightning counts from surface networks will be exploited to improve QPN. With increasing forecast horizons the predictive power of observation-based nowcasting quickly deteriorates and is outperformed by Numerical Weather Prediction (NWP) based on data assimilation, which fails, however, for the first hours due to the lead time required for model integration and spin-up. Thus, RealPEP will merge observation-based QPN with NWP towards seamless prediction in order to provide optimal forecasts from the time of observation to days ahead. Despite recent advances in simulating surface and sub-surface hydrology with distributed, physicsbased models, hydrologic components for operational flood prediction are still conceptual, need calibration, and are unable to objectively digest observational information on the state of the catchments. RealPEP will prove that in combination with advanced QPE/QPN physics-based hydrological models sided with assimilation of catchment state observations will outperform traditional flood forecasting in small to meso-scale catchments