A Probabilistic Framework for Assessment and Interpretation of Quantitative Precipitation Forecasts from Storm-Scale Models

9714387 Foufoula美国天气研究计划（USWRP）是一项跨部门活动，旨在执行和实施必要的研究，以改善向全国提供的天气服务。根据该计划，国家科学基金会、国家海洋和大气管理局（NOAA）、国家航空航天局和海军研究办公室将联合评估和支持USWRP的高优先级研究。在过去的十年中，人们在研究和数据收集方面做出了重大努力，重点是改进数值天气预报。这些投资已经取得了一定程度的成功，通过对所涉及的物理过程的更深入的理解和改进的参数化，新的和更好的数据同化技术，以及大大改进的数值方案，数值天气预报的最新技术已经真正取得了进步。然而，尽管取得了这些进展，降水仍然是大气模式中最难定量预测的变量之一。此外，与数值预测相关的固有不确定性往往使预测的解释变得困难，并限制了其实际效用。主要研究人员提出了一种跨学科的方法来预测风暴尺度的降水和解释数值预报的问题。该方法将涉及利用先进的中尺度气象模式和先进的统计技术。研究人员将处理以下三个目标，重点是使用风暴尺度模型，利用雷达和其他精细尺度观测初始化，以产生定量降水预报和相关的集合，用于水文建模和洪水预测等应用。制定和研究一套预报性能的多尺度统计措施，有助于发现和纠正微物理参数化的缺陷，并获得适合应用尺度的预报评估，例如流域平均降雨量。制定一个概率框架，在明确承认预报不确定性和尺度依赖性的同时，评估当前数值天气预报模式的预报改进和可预测性限制。开发一个概率框架，以评估确定性预测对基于风险的决策的实际效用。这项研究的成功完成将有可能为改进定量降水预报和更好地利用/解释数值预报打开大门。＊＊＊

9714387 Foufoula The U.S. Weather Research Program (USWRP) is an interagency activity designed to perform and implement the research necessary to improve the delivery of weather services to the nation. Under this Program, the National Science Foundation, the National Oceanic and Atmospheric Administration (NOAA), the National Aeronautics and Space Administration and the Office of Naval Research are jointly evaluating and supporting research of high priority to the USWRP. Over the past decade, a significant efforts have been directed toward research and data collection initiatives that focus on improving numerical weather prediction. These investments have been somewhat successful and state-of-the-art in numerical weather prediction has truly advanced through a deeper understanding and improved parameterizations of the physical processes involved, new and better data assimilation techniques, and greatly improved numerical schemes. Despite this progress, however, precipitation still remains one of the most difficult variables to predict quantitatively in atmospheric models. Also, the inherent uncertainty associated with numerical predictions often makes the interpretation of forecasts difficult and limits their practical utility. The Principal Investigators have proposed an interdisciplinary approach to the problem of predicting precipitation on the storm-scale and interpreting the numerical forecasts. The approach will involve utilization of an advanced mesoscale meteorological model coupled with advanced statistical techniques. The Investigators will address the following three objectives, with emphasis on the use of storm-scale models, initialized using radar and other fine-scale observations, to generate quantitative precipitation forecasts and associated ensembles to be used in applications such as hydrologic modeling and flood prediction. Formulate and study a suite of multiscale statistical measures of forecast performance which can help to detect and correct deficiencies in microphysical parameterizations and obtain forecast assessments at application-appropriate scales, e.g., basin-average rainfall. Develop a probabilistic framework under which forecast improvements and the limits of predictability of current numerical weather prediction (NWP) models can be assessed while explicitly acknowledging forecast uncertainty and dependence on scale. Develop a probabilistic framework for assessing the practical utility of deterministic forecasts for risk-based decision making. Successful completion of this research will potentially open the doors to improved quantitative precipitation forecasting and better utilization/interpretation of numerical forecasts. ***