张野,胡庆芳,黄勇,王银堂,李伶杰,云兆得.基准降水对GPM数据精度的影响——以南京“20170610”极端降水事件为例[J].水利学报,2023,54(3):369-380 |
基准降水对GPM数据精度的影响——以南京“20170610”极端降水事件为例 |
The influence of baseline precipitation on the accuracy of GPM data evaluation: A case study of “20170610” extremt precipitation event in Nanjing |
投稿时间:2022-05-23 |
DOI:10.13243/j.cnki.slxb.20220405 |
中文关键词: 卫星反演降水 GPM 地表基准降水 精度验证 站点密度 |
英文关键词: satellite-retrieved precipitation GPM surface reference precipitation accuracy verification gauge density |
基金项目:国家重点研发计划课题(2021YFC3000104,2019YFC0408903);国家自然科学基金面上项目(51479118,52009081) |
|
摘要点击次数: 893 |
全文下载次数: 430 |
中文摘要: |
现有文献多采用两种途径生成评价GPM等卫星遥感反演降水数据精度的基准资料:站点基准降水和插值基准降水,然而,关于基准资料生成方法对卫星降水数据精度的影响以及应优先采纳哪种基准资料尚缺乏深入研究。因此,本文以南京"20170610"极端降水事件为例,利用地面稠密雨量站的观测资料,设计了不同地表雨量站点密度下的随机抽取试验,解析了不同地表基准降水条件下IMERG Early、IMERG Late和GSMaP NOW三种GPM数据的精度特征。结果发现,基准降水对GPM数据精度的影响与雨量要素、数据类型、指标类型、站点密度有关。进一步以高超密度站点插值降水作为"真实"降水,量化了不同站点密度条件下两种基准降水的误差,发现对于24 h累积雨量和逐小时降水过程,存在一定的雨量站点平均密度临界值。若雨量站点密度高于该临界值,以插值降水为基准评价GPM数据的精度更合理,否则采用插值降水和站点降水为基准均可。本文研究深化了对GPM 降水数据误差特性的认识,可为合理选取评估卫星降水数据的地表基准数据生成方法提供参考。 |
英文摘要: |
In the existing literature,there are two major methods to obtain reference precipitation for evaluating accuracy of precipitation retrievals from satellite such as GPM:gauges reference precipitation and interpolates reference precipitation.However,existing studies are limited due to the lack of understanding about the effects of the two benchmarking methods in satellite precipitation estimates (SPEs) evaluation and the strategy and priority on how to use them.This work takes the June 10,2017 extreme rainfall event in Nanjing as an example.We implement a Randow extraction study by randomly sampling from a dense gauge network to derive reference precipitation under different gauge density scenarios.The accuracy characteristics of three GPM data (IMERG Early,IMERG Late and GSMaP Now) are analyzed and compared using the two reference precipitation estimation methods.The results show that the influence of different reference precipitation estimation methods depends on analysis scale,GPM data type,evaluation metric,and gauges density.Taking the interpolated reference from the highest density gauges as ground "truth",the evaluation errors based on the two reference precipitations are further quantified under different gauges densities.A certain gauge-density threshold is identified for the evaluation of 24-hour cumulative rainfall and hourly rainfall,respectively.If rain gauge density is higher than that threshold,it is more reasonable to use the interpolation-based reference estimation method;otherwise,either interpolation- or station-based reference estimation methods can be used without significant difference.This study deepens the understanding of error characteristics of GPM precipitation product,and provides a guidance for reasonably determining the ground reference for evaluation of SPEs. |
查看全文
查看/发表评论 下载PDF阅读器 |
关闭 |
|
|
|