环渤海地区卫星降水产品精度检验及校正

发布时间：2018-04-11 04:22

本文选题：卫星降水产品 + 精度评价　；参考：《鲁东大学》2016年硕士论文

【摘要】：降水是重要的水文气象变量,其在全球能量和水循环中起重要作用。同时,降水具有较高的时空变异性,是旱涝灾害形成的重要因素。因此,降水数据的准确性对洪水、干旱监测及预报、水资源开发利用等工作至关重要,是认识水文过程、揭示水文变化的重要基础。环渤海地区属于我国“三大经济圈”之一,是我国的政治、经济、文化中心。水资源匮乏已对环渤海地区的经济和社会发展产生严重影响。准确的降水数据是该区域干旱监测及水资源科学管理的重要基础。目前,降水数据获取的主要手段包括地面降水观测站、卫星遥感、模型模拟等。其中,地面降水观测站空间分布不均匀,致使站点降水资料具有明显的局限性;模型模拟需要基于大量物理数据,且仅能模拟大尺度的降水情况,弊端较为明显;卫星降水产品的时空分辨率较高,能有效弥补站点降水的不足,但其对降水的监测是瞬时值,监测精度仍表现出较多的不确定性。本文研究目的主要在于,通过对环渤海多卫星降水产品精度检验,建立高分辨率卫星降水产品在我国环渤海地区应用的评估与分析基础,全面理解多种卫星降水产品的数据性能,并为其改进提供信息支持;同时,选取综合精度较优的卫星降水产品,对其日尺度下估算精度进行校正,提高其科学与应用价值。精度检验结果表明,总体上,TRMM 3B42 V7的精度最高,PERSIANN精度最低,但CMORPH对日降水事件的探测成功率最高;3种卫星降水产品的月尺度估算精度优于日尺度,且夏季高于其他季节;其中,各卫星降水产品对日降水量的估算结果均不理想,对日降水事件探测率低和错报率较高是误差形成的重要因素。因此,本文选取环渤海地区综合精度较高的TRMM 3B42数据,基于小波多分辨率分析方法对其进行校正。在对比订正前的3B42产品误差特征的基础上,分析订正后3B42数据的降水估算精度。结果表明:订正后的3B42数据与站点降水数据之间的相关系数显著提高,MAE、RMSE则明显降低;订正后的3B42数据对日降水事件的探测成功率得到提高,错报率有所下降;小波多分辨率分析方法有效提高了3B42数据对各季节,特别是冬季日降水的估算精度;但,订正后的3B42数据对日降水量仍存在高估现象。
[Abstract]:Precipitation is an important hydrometeorological variable, which plays an important role in the global energy and water cycle.At the same time, precipitation has high temporal and spatial variability, which is an important factor in the formation of drought and flood disasters.Therefore, the accuracy of precipitation data is of great importance to flood, drought monitoring and forecasting, water resources development and utilization, and is an important basis for understanding hydrological processes and revealing hydrological changes.Bohai Rim is one of the three economic circles in China, which is the political, economic and cultural center of our country.Lack of water resources has had a serious impact on the economic and social development of Bohai Rim.Accurate precipitation data is an important basis for drought monitoring and scientific management of water resources in the region.At present, the main methods of precipitation data acquisition include ground precipitation observation station, satellite remote sensing, model simulation and so on.Among them, the spatial distribution of the ground precipitation observation station is uneven, which makes the precipitation data of the station have obvious limitations, the model simulation needs to be based on a large number of physical data, and can only simulate the large-scale precipitation situation, the disadvantages are obvious.The spatial and temporal resolution of satellite precipitation products is high, which can effectively make up for the shortage of precipitation at the station, but the monitoring of precipitation is instantaneous, and the monitoring accuracy is still more uncertain.The purpose of this paper is to establish the basis of evaluation and analysis of the application of high-resolution satellite precipitation products in China's Bohai Rim area by testing the precision of the multi-satellite precipitation products around the Bohai Sea.We comprehensively understand the data performance of various satellite precipitation products and provide information support for its improvement. At the same time, we select satellite precipitation products with better synthetic precision to correct their estimation accuracy on a daily scale, so as to improve their scientific and practical value.The precision test results show that the precision of TRMM 3B42 V7 is the highest and that of PERSIANN is the lowest on the whole, but the accuracy of the monthly scale estimation of the three satellite precipitation products by CMORPH is the highest, and the summer is higher than that of other seasons, and the accuracy of the monthly scale estimation of the three satellite precipitation products is higher than that of the daily precipitation events.The estimation results of daily precipitation by satellite precipitation products are not satisfactory. The low detection rate and high misreporting rate of daily precipitation events are important factors for the formation of errors.Therefore, the TRMM 3B42 data with high accuracy are selected and corrected based on wavelet multi-resolution analysis method.On the basis of comparing the error characteristics of 3B42 products before revision, the precipitation estimation accuracy of the revised 3B42 data is analyzed.The results show that the correlation coefficient between the revised 3B42 data and the precipitation data is significantly increased, and the success rate of detecting daily precipitation events is improved, and the misreporting rate is decreased.Wavelet Multiresolution Analysis method can effectively improve the accuracy of 3B42 data estimation of daily precipitation in different seasons, especially in winter, but the revised 3B42 data still overestimate the daily precipitation.
【学位授予单位】：鲁东大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P412.27

【参考文献】