2018年夏季我国极端降水及滑坡泥石流灾害预测

基于动力降尺度预测系统，中国科学院大气物理研究所竺可桢-南森国际研究中心对2018年夏季我国极端降水日数及滑坡泥石流灾害的发生风险进行了超前4个月的实时预测试验。与实测结果相比，该系统对2018年夏季我国极端降水日数空间分布的预测与实况基本相符，但大部分地区存在明显低估；滑坡泥石流的预测结果与目前统计的由于降水引发的滑坡泥石流灾害事件的分布基本吻合。此次预测试验表明，中国科学院大气物理研究所竺可桢-南森国际研究中心发展的动力降尺度预测系统对我国夏季极端降水和滑坡泥石流灾害具有一定的预测能力，具有实时预测价值。

Real-time prediction for 2018 JJA extreme precipitation and landslides

In this study,based on a dynamical downscaling prediction system,four-month leading real-time prediction experiments for JJA extreme precipitation and landslides were performed in February 2018 by the Nansen-Zhu International Research Centre,Institute of Atmospheric Physics,Chinese Academy of Sciences.The study results show that the spatial pattern of predicted summer extreme precipitation days is quite similar to the observations,yet with a significant underestimation over most regions of mainland China.The distribution of the predicted landslides is basically consistent with the reported landslide events caused by precipitation.The results show that the prediction system has a certain degree of predictability for summer extreme precipitations and landslide disasters,and possesses prospects for real-time prediction.