GPM与TRMM卫星降雨数据在福建省的适用性对比分析
|
摘要
降雨是引起土壤侵蚀的外部动力因素,卫星降雨产品能够有效地克服常规降雨观测数据在空间上的不连续性,具有明显空间分布优势。为了探究卫星降雨产品在福建地区的适应性,利用福建省22个全国气象站点的观测数据,分别对全球降雨测量GPM和热带降雨测量TRMM两种卫星降水数据在日、旬、月、季度几个不同时间尺度上进行了对比分析。结果表明:随着时间尺度增加,GPM和TRMM与地面实测数据的相关性都呈现出依次递增的现象,日均均方根误差RMSE~*逐渐降低,特别是由日到旬尺度变化时RMSE~*出现陡降,而旬和月的差异不大;卫星产品与站点实测数据拟合直线的斜率逐渐趋向于1。在日尺度上,相关性都小于0.5,RMSE~*均大于13 mm/d;在旬尺度上,相关性介于0.7～0.8,RMSE~*均小于5 mm/d;在月尺度上,两者与观测数据的相关性都大于0.9,GPM数据整体上精度优于TRMM数据,但是在福建东南沿海TRMM精度要高于GPM。在季度尺度上两种卫星产品的精度与季节总降雨量呈负相关,季节差异较大;夏季多雨时r值和K值低而RMSE最高,冬季少雨时与地面实测数据一致性高,r和K值最高而RMSE最低,春秋两季也有较好的精度;在不同的季节GPM各项指标均优于TRMM。综合不同时间尺度的结果,总体上,两种卫星降雨数据都具有较高的精度,其中GPM卫星降水数据在福建省的适用性较高,数据的空间分辨率更高,数据精度也略高,更适合于区域应用。
Rainfall is the external dynamic factor that causes soil erosion. Satellite rainfall products can effectively overcome the discontinuity of conventional rainfall observation data in space and have obvious spatial distribution advantages. In order to explore the adaptability of satellite rainfall products in Fujian Province, using the observational data of 22 meteorological stations in Fujian Province, a comparative analysis on the global rainfall measurement GPM and tropical rainfall measurement TRMM at different time scales of day, ten days, months and seasons was carried out. The results show that with the increase of time scale, the correlation of GPM and TRMM with the measured data on the ground also increases, and the daily root mean square error(RMSE~*) decreases gradually, especially the RMSE~* drops steeply from day to ten-day scale, however, the difference between RMSE~* of ten-day scale and that of the month scale was not obvious; the slope of the fitting line between satellite products and the measured data of the station gradually tends to be 1; the correlation is lower than 0.5, and the correlation of RMSE~* is greater than 13 mm/d at the daily scale, and the correlation is between 0.7～0.8, and RMSE~* is less than 5 mm/d at the ten-day scale; with respect to the monthly scale, the correlation is between data of GPM and TRMM and the observed data is greater than 0.9, the accuracy of GPM data is better than that of TRMM data, but the accuracy of TRMM data is higher than that of GPM data in the southeast coast of Fujian Province; with respect to the quarterly scale, the accuracy of the two satellite products is negatively correlated with the total seasonal rainfall, and there is a great seasonal difference; When a lot of rain occurs in summer, the r and K values are low and RMSE is the highest; when little rain occurs in winter, it is more consistent with the measured data on the ground, r and K values are the highest and RMSE is the lowest, precision is also high in spring and autumn; the precision of GPM is better than that of TRMM in different seasons. According to the results of different time scales, the two kinds of satellite rainfall data both have high accuracy, while the GPM satellite precipitation data has higher spatial resolution and higher data precision in Fujian Province and more suitability for regional applications.
Rainfall is the external dynamic factor that causes soil erosion. Satellite rainfall products can effectively overcome the discontinuity of conventional rainfall observation data in space and have obvious spatial distribution advantages. In order to explore the adaptability of satellite rainfall products in Fujian Province, using the observational data of 22 meteorological stations in Fujian Province, a comparative analysis on the global rainfall measurement GPM and tropical rainfall measurement TRMM at different time scales of day, ten days, months and seasons was carried out. The results show that with the increase of time scale, the correlation of GPM and TRMM with the measured data on the ground also increases, and the daily root mean square error(RMSE~*) decreases gradually, especially the RMSE~* drops steeply from day to ten-day scale, however, the difference between RMSE~* of ten-day scale and that of the month scale was not obvious; the slope of the fitting line between satellite products and the measured data of the station gradually tends to be 1; the correlation is lower than 0.5, and the correlation of RMSE~* is greater than 13 mm/d at the daily scale, and the correlation is between 0.7～0.8, and RMSE~* is less than 5 mm/d at the ten-day scale; with respect to the monthly scale, the correlation is between data of GPM and TRMM and the observed data is greater than 0.9, the accuracy of GPM data is better than that of TRMM data, but the accuracy of TRMM data is higher than that of GPM data in the southeast coast of Fujian Province; with respect to the quarterly scale, the accuracy of the two satellite products is negatively correlated with the total seasonal rainfall, and there is a great seasonal difference; When a lot of rain occurs in summer, the r and K values are low and RMSE is the highest; when little rain occurs in winter, it is more consistent with the measured data on the ground, r and K values are the highest and RMSE is the lowest, precision is also high in spring and autumn; the precision of GPM is better than that of TRMM in different seasons. According to the results of different time scales, the two kinds of satellite rainfall data both have high accuracy, while the GPM satellite precipitation data has higher spatial resolution and higher data precision in Fujian Province and more suitability for regional applications.
引文
[1] Mou T,Ab I,Zheng D,et al.Evaluation of six high-resolution satellite and ground-based precipitation products over Malaysia[J].Remote Sensing,2015,7(2):1504-1528.
[2] Panagos P,Ballabio C,Borrelli P,et al.Spatio-temporal analysis of rainfall erosivity and erosivity density in Greece[J].Catena,2015,137:161-172.
[3] 蔡研聪,金昌杰,王安志,等.中高纬度地区TRMM卫星降雨数据的精度评价[J].应用生态学报,2014,25(11):3296-3306.
[4] 嵇涛,杨华,刘睿,等.TRMM卫星降水数据在川渝地区的适用性分析[J].地理科学进展,2014,33(10):1375-1386.
[5] Robbins J C.A probabilistic approach for assessing landslide-triggering event rainfall in Papua New Guinea,using TRMM satellite precipitation estimates[J].Journal of Hydrology,2016,541(Part A):296-309.
[6] 李德俊,李跃清,柳草,等.利用TRMM卫星资料对“07.7”川南特大暴雨的诊断研究[J].暴雨灾害,2009,28(3):235-240.
[7] Castro E,Campozano L.Validation of satellite precipitation (TRMM 3B43) in Ecuadorian,coastal plains,Andean highlands and Amazonian rainforest[J].Isprs-International Archives of the Photogrammetry,Remote Sensing and Spatial Information Sciences,2016,DOI:10.5194/isprsarchives-XLI-B8-305-2016.
[8] Zhang S,Wang D,Qin Z,et al.Assessment of the GPM and TRMM precipitation products using the rain gauge network over the Tibetan Plateau[J].Journal of Meteorological Research,2018,32(2):324-336.
[9] Mou T,Zheng D.Assessment of GPM and TRMM precipitation products over Singapore[J].Remote Sensing,2017,9(7):720-736.
[10] 金晓龙,邵华,张弛,等.GPM卫星降水数据在天山山区的适用性分析[J].自然资源学报,2016,31(12):2074-2085.
[11] 王振平,汪小钦,曾舒娇,等.福建省水土流失现状分析[J].水土保持通报,2017,37(5):169-175.
[12] 黄婕,王跃峰,高路,等.1960—2011年福建省不同等级降水时空变化特征[J].中国水土保持科学,2015,13(2):17-23.
[13] 李蒙,秦天玲,刘少华,等.怒江上游TRMM 3B42V7降水产品资料时空验证及降水特征分析[J].高原气象,2017,36(4):950-959.
[14] 李麒崙,张万昌,易路,等.GPM与TRMM降水数据在中国大陆的精度评估与对比[J].水科学进展,2018,29(3):303-313.
[15] 张月圆,李运刚,季漩,等.红河流域TRMM卫星降水数据精度评价[J].水资源与水工程学报,2017,28(2):1-8.
[16] Tang Guoqiang,Ma Yingzhao,Long Di,et al.Evaluation of GPM Day-1 IMERG and TMPA Version-7 legacy products over Mainland China at multiple spatiotemporal scales[J].Journal of Hydrology,2016,533:152-167.
[17] 余赛英.福建省降水特性分析[J].黑龙江水专学报,2002,29(2):24-26.
[18] 杨晓梅,刘李,史婷婷,等.基于TRMM数据的福建省降水时空格局BME插值分析[J].地球信息科学学报,2014,16(3):470-481.
[19] 唐国强,万玮,曾子悦,等.全球降水测量(GPM)计划及其最新进展综述[J].遥感技术与应用,2015,30(4):607-615.
[20] Tan M L,Santo H.Comparison of GPM IMERG,TMPA3B42 and PERSIANN-CDR satellite precipitation products over Malaysia[J].Atmospheric Research,2018,202:63-76.
[2] Panagos P,Ballabio C,Borrelli P,et al.Spatio-temporal analysis of rainfall erosivity and erosivity density in Greece[J].Catena,2015,137:161-172.
[3] 蔡研聪,金昌杰,王安志,等.中高纬度地区TRMM卫星降雨数据的精度评价[J].应用生态学报,2014,25(11):3296-3306.
[4] 嵇涛,杨华,刘睿,等.TRMM卫星降水数据在川渝地区的适用性分析[J].地理科学进展,2014,33(10):1375-1386.
[5] Robbins J C.A probabilistic approach for assessing landslide-triggering event rainfall in Papua New Guinea,using TRMM satellite precipitation estimates[J].Journal of Hydrology,2016,541(Part A):296-309.
[6] 李德俊,李跃清,柳草,等.利用TRMM卫星资料对“07.7”川南特大暴雨的诊断研究[J].暴雨灾害,2009,28(3):235-240.
[7] Castro E,Campozano L.Validation of satellite precipitation (TRMM 3B43) in Ecuadorian,coastal plains,Andean highlands and Amazonian rainforest[J].Isprs-International Archives of the Photogrammetry,Remote Sensing and Spatial Information Sciences,2016,DOI:10.5194/isprsarchives-XLI-B8-305-2016.
[8] Zhang S,Wang D,Qin Z,et al.Assessment of the GPM and TRMM precipitation products using the rain gauge network over the Tibetan Plateau[J].Journal of Meteorological Research,2018,32(2):324-336.
[9] Mou T,Zheng D.Assessment of GPM and TRMM precipitation products over Singapore[J].Remote Sensing,2017,9(7):720-736.
[10] 金晓龙,邵华,张弛,等.GPM卫星降水数据在天山山区的适用性分析[J].自然资源学报,2016,31(12):2074-2085.
[11] 王振平,汪小钦,曾舒娇,等.福建省水土流失现状分析[J].水土保持通报,2017,37(5):169-175.
[12] 黄婕,王跃峰,高路,等.1960—2011年福建省不同等级降水时空变化特征[J].中国水土保持科学,2015,13(2):17-23.
[13] 李蒙,秦天玲,刘少华,等.怒江上游TRMM 3B42V7降水产品资料时空验证及降水特征分析[J].高原气象,2017,36(4):950-959.
[14] 李麒崙,张万昌,易路,等.GPM与TRMM降水数据在中国大陆的精度评估与对比[J].水科学进展,2018,29(3):303-313.
[15] 张月圆,李运刚,季漩,等.红河流域TRMM卫星降水数据精度评价[J].水资源与水工程学报,2017,28(2):1-8.
[16] Tang Guoqiang,Ma Yingzhao,Long Di,et al.Evaluation of GPM Day-1 IMERG and TMPA Version-7 legacy products over Mainland China at multiple spatiotemporal scales[J].Journal of Hydrology,2016,533:152-167.
[17] 余赛英.福建省降水特性分析[J].黑龙江水专学报,2002,29(2):24-26.
[18] 杨晓梅,刘李,史婷婷,等.基于TRMM数据的福建省降水时空格局BME插值分析[J].地球信息科学学报,2014,16(3):470-481.
[19] 唐国强,万玮,曾子悦,等.全球降水测量(GPM)计划及其最新进展综述[J].遥感技术与应用,2015,30(4):607-615.
[20] Tan M L,Santo H.Comparison of GPM IMERG,TMPA3B42 and PERSIANN-CDR satellite precipitation products over Malaysia[J].Atmospheric Research,2018,202:63-76.