基于卫星降雨的辽宁省气象干旱实时监测研究

卫星实时降雨产品的出现为开展大范围干旱监测提供了可能。然而,卫星遥感降雨通常存在误差大、数据序列短等问题,限制了卫星降雨自身在干旱监测诊断方面的应用。本文通过误差矫正,将TMPA-RT卫星降雨数据无缝"拼接"具有长期历史序列(1961—2016年)的地面观测产品(CN05.1),以弥补卫星遥感数据序列短的不足。借助上述"拼接"的实时降雨序列,通过估算不同时间尺度下的标准化降雨指数(SPI)来实现干旱的实时监测诊断。以干旱频发的辽宁省为研究区,对比评估了矫正前后TMPA-RT产品在干旱识别方面的表现。结果表明,在辽宁省近半数地区,TMPA-RT原始数据与CN05.1地面观测两者存在明显的降雨误差(10 mm/月);经过误差矫正,TMPA-RT数据能够重现同期地面观测的降雨量及年际变化。基于3个典型干旱月的评估结果表明,基于TMPA-RT矫正数据与CN05.1"拼接"得到的序列能够重现观测的干/湿空间分布,准确监测辽西地区的干旱状况;而原始的TMPA-RT数据难以提供准确可靠的干旱信息。将上述框架实际应用于2011—2016年干旱监测,评估发现该框架不仅能合理地重现辽宁省干旱情况,还能提供不同等级干旱的面积分布信息。

Satellite-based real-time meteorological drought monitoring in Liaoning Province

Satellite remote sensing mokes it possible to conduct the large-scale drought monitoring. However, satellite-based precipitation product always suffers from considerable retrieve errors and short data length,hampering its effective application in drought diagnosis and detection. To solve this issue,this paper attempts to correct the TRMM satellite precipitation (TMPA-RT) relative to the multi-decadal ground observations (CN05.1) through a quantile-mapping (QM) bias correction approach, and thus derives a set of real-time precipitation data conjunction with a long-term consistent climatology. Using the satellite-gauge integrated real-time data record,this paper enables an operational drought monitoring through the real-time estimation of standardized precipitation index (SPI) in the drought-prone Liaoning Province,and its performance in drought diagnosis is assessed against with that implied from the observation data in the same period. Results show that an obvious precipitation bias (10mm/month) is found in the raw (uncorrected) TMPA-RT data across a large portion of Liaoning region. With the QM bias correction,the TMPA-RT data is found to well reproduce the observations,in terms of daily precipitation intensity and the interannual variability of monthly precipitation. Further evaluation analyses in three typical drought months show that the integrated satellite-gauge data product,after bias correction,can successfully reproduce the observational SPI patterns and thus provides accurate drought monitoring information in the western Liaoning. The real-time drought monitoring application during 2011-2016 shows that this approach is not only able to reasonably capture the drought conditions, but also can provide detailed area information affected by varying drought intensity. This study provides an additional framework for the existing drought monitoring approaches,which will greatly benefit the regional drought management and risk mitigation.