一种基于FSS评分定义的降水预报跳跃指数及其应用

A precipitation forecast jumpiness index based on FSS and its application

中文关键词:  CMA-BJ系统，降水预报，预报不一致性，定量评估

英文关键词:CMA-BJ model, Precipitation forecast, Forecast inconsistency, Quantitative assessment

中文摘要:

      基于CMA-BJ系统提供的2021年8月9日两套降水预报结果（业务数值预报模式结果和同化试验）和2022年6月4日、2023年9月17日的业务数值预报模式结果，结合主观分析，利用四种客观评价指数（不确定度、均方根误差、跳跃指数和基于FSS评分定义的预报跳跃指数）对该系统降水预报不一致性特征进行了定量评估。三次降水过程的分析结果显示：基于FSS评分定义的预报跳跃指数不仅可以识别出2021年8月9日和2022年6月4日业务结果中降水量预报明显减小的3个预报时次，而且对于降水过程预报相对稳定的个例（2021年8月9日同化试验和2023年9月17日业务结果），随着预报时次逐渐临近最新预报，该指数整体呈现波动上升或者数值较大、波动较小的特征，表明15个连续降水预报特征逐渐与最新预报趋于一致或者大体相似，与主观分析结果相对吻合。由于预报跳跃指数关注选定窗区内的降水概率，不受降水量值的影响，因此能够更加直观地反映连续预报时次降水落区预报的空间分布特征，对于分析降水过程整体演变规律具有独特优势。其他三种指数对于降水预报不一致性问题的表征存在不足，例如不确定度和均方根误差显著受到降水预报量值的影响，同时不确定度不能反映预报不一致的时间特征，跳跃指数随预报时次逐时滚动变化较大，根据其筛选标准，确定的跳跃时次较多，与主观分析结果存在明显偏差。

英文摘要:

      Based on two sets of precipitation forecast results of CMA-BJ model (operational and assimilated forecast) on 9 August 2021 and operational results on 4 June 2022 and 17 September 2023, combined with subjective analysis, four objective evaluation indexes (uncertainty, root-mean-square error, inconsistency index and forecast jumpiness index based on FSS) were used to quantitatively evaluate the inconsistency characteristics of precipitation forecast. The analysis of three precipitation processes showed that the forecast jumpiness index based on FSS could not only identify the three forecast moments in which the precipitation forecast decreased significantly in the operational results on 9 August 2021 and 4 June 2022, but also distinguished them from the cases in which precipitation forecast was relatively stable based on the assimilated forecast on 9 August 2021 and the operational results on 17 September 2023. As the forecast moment gradually approached the latest forecast, the jumpiness index became rising or maintained a large value and small fluctuation on the whole, indicating that the 15 continuous precipitation forecast results gradually converged or were roughly similar to the latest forecast, which was relatively consistent with the results of subjective analysis. Since the forecast jumpiness index focused on the precipitation probability in the selected window area, it was not affected by the value, so it could more directly reflect the spatial distribution characteristics of the precipitation forecast in the continuous forecast results of the model, and had a unique advantage for analyzing the overall evolution characteristics of the precipitation process. The other three indexes were insufficient to characterize the inconsistency of precipitation forecast. For example, the uncertainty and the root- mean-square error were significantly affected by the value of precipitation forecast, and the uncertainty were not able to reflect temporal characteristics of the forecast inconsistency. The inconsistency index changed greatly as the forecast time rolled from time to time. According to the criteria of the inconsistency index, many jumpiness cases were selected, which deviated from the subjective analysis.