BIAS CORRECTION OF CMROPH SATELLITE PRECIPITATION PRODUCTS OVER GUIZHOU
-
摘要: CMORPH卫星反演降水产品具有全天候、全球覆盖的特点,其时空分布相对均匀、独立,但是CMORPH本质上是通过间接手段反演得到,其降水精度无法与地面观测降水精度相比,并且存在一定的系统误差。结合地面自动站降水资料采用概率密度匹配法对贵州地区CMORPH卫星反演降水产品进行系统误差订正,该方法将每个格点的卫星降水累积概率分布曲线和地面降水概率密度分布曲线匹配,获取降水误差订正值;其中误差订正效果受降水累积概率分布拟合曲线的影响,而考虑到降水累积概率分布是非正态分布,因此选用Gamma分布拟合降水累积概率分布曲线。通过对2018年5月三次降水过程的订正结果分析得到如下结论:(1) 逐时的CMORPH卫星反演降水产品存在明显的非独立系统误差,误差范围随降水量级的变化而变化,存在低值高估的特点;(2) 在小时尺度下地面降水的累积概率密度呈指数衰减分布,而CMORPH的降水累积概率密度分布更加复杂,其在中雨、大雨区间内的降水概率较高;(3) 通过概率密度匹配法订正后的CMORPH与订正前相比降水空间结构更加贴近地面降水,强降水中心的量级和范围明显减小,平均绝对误差和均方根误差均减小,其中偏差订正值在0.114~0.468 mm/h,均方根误差订正在0.24~1.49 mm/h之间。经概率密度匹配法订正后的CMORPH卫星反演降水产品精度明显提升,更加接近于实际降水。
-
关键词:
- CMORPH卫星反演降水产品 /
- 地面降水 /
- 概率密度匹配法 /
- 误差订正 /
- Gamma分布函数
Abstract: The CMORPH satellite-retrieved precipitation products feature all-weather and global coverage, and their spatio-temporal distribution is relatively uniform and independent. However, CMORPH products are obtained using indirect methods, their accuracy is lower than that of ground observations, and there is systematic bias. Using the precipitation data from ground automatic stations, this paper adopts the probability density matching method to correct the system bias of the CMORPH satellite precipitation products over Guizhou. The satellite precipitation cumulative probability distribution curve of each grid point is matched with the surface precipitation probability density distribution curve to correct precipitation bias, and the process is influenced by the precipitation cumulative probability distribution fitting curve. Given that the precipitation cumulative probability distribution is non-normal, Gamma distribution is selected to fit the precipitation cumulative probability distribution curve. Through the analysis of the correction results of three precipitation processes in May 2018, following conclusions are obtained: (1) There is obvious and non-independent systematic bias in the hourly CMORPH satellite-retrieved precipitation products. The bias varies with the level of precipitation, and sometimes precipitation is overestimated. (2) On the hourly scale, the cumulative probability density of surface precipitation shows an exponential decay distribution, while the cumulative probability density distribution of CMORPH products is more complicated, and the probability of precipitation in the interval of moderate rain and heavy rain is higher. (3) The CMORPH products corrected by using the probability density matching method are closer to ground precipitation than those before the correction. The magnitude and range of heavy precipitation center are significantly reduced, and the average absolute error and root mean square error are reduced. The deviation correction value is between 0.114~0.468 mm/h, and the root means square error is between 0.24~1.49 mm / h. The accuracy of the CMORPH satellite precipitation products after correction by using the probability density matching method has been significantly improved and is closer to actual precipitation. -
表 1 订正效果评估
评估参数 ARE RMSE 订正前 订正后 订正前 订正后 2018年5月22日10:00 1.272 0.804 3.814 2.324 2018年5月22日11:00 0.955 0.645 3.119 2.080 2018年5月22日12:00 0.708 0.594 2.360 2.120 2018年5月22日13:00 0.735 0.611 2.726 2.196 表 2 2018年5月13日、5月29日两次降水过程订正效果评估
评估参数 ARE RMSE 订正前 订正后 订正前 订正后 2018年5月13日06:00 0.402 0.371 1.583 1.178 2018年5月13日07:00 0.456 0.363 2.051 1.197 2018年5月13日08:00 0.439 0.374 2.435 1.377 2018年5月13日09:00 0.435 0.371 2.441 1.388 2018年5月29日16:00 1.564 1.262 3.593 2.743 2018年5月29日17:00 1.536 1.202 3.664 2.800 2018年5月29日18:00 1.492 1.164 3.726 2.761 2018年5月29日19:00 1.376 1.091 3.872 2.834 -
[1] MORRISSEY M L, MALIEKAL J A, GREENE J S, et al. The uncertainty of simple spatial averages using rain gauge networks[J]. Water Resour Res, 1995, 31(8): 2 011-2 017. [2] VERDIN A, RAJAGOPALAN B, KLEIBER W, et al. Bayesian kriging approach for blending satellite and ground precipitation observations[J]. Water Resour Res, 2015, 51(2): 908-921. [3] 潘旸, 谷军霞, 宇婧婧, 等. 中国区域高分辨率多源降水观测产品的融合方法试验[J]. 气象学报, 2018, 76(5): 755-766. [4] YU J J, LI X F, LEWIS E, et al. UKGrsHP: a UK high-resolution gauge-radar-satellite merged hourly precipitation analysis dataset[J]. Climate Dyn, 2020, 54: 2 919-2 940. [5] EUGENIA K. Atmospheric modeling, data assimilation and predictability[M]. Cambridge: Cambridge University Press, 2005. [6] 师春香, 刘玉洁. 国外部分卫星产品质量评价和质量控制方法[J]. 应用气象学报, 2004, 15(S1): 142-151. [7] SHEN Y, XIONG A Y, WANG Y, et al Performance of high-resolution satellite precipitation products over China[J]. J Geophy Res, 2010, 115, D02114, doi:10.1029/2009JD012097. [8] 宇婧婧, 沈艳, 潘旸, 等. 概率密度匹配法对中国区域卫星降水资料的改进[J]. 应用气象学报, 2013, 24(5): 544-553. [9] WANG W Q, XIE P P. A Multiplatform-Merged (MPM) SST analysis[J]. J Climate, 2007, 20(9): 1 662-1 679. [10] BOUSHAKI F I, HSU K, SOROOSHIAN S, et al. Bias adjustment of satellite precipitation estimation using ground-based measurement: a case study evaluation over the southwestern United States[J]. J Hydrometeor, 2009, 10(5): 1 231-1 242. [11] 胡庆芳. 基于多源信息的降水空间估计及其水文应用研究[D]. 北京: 清华大学, 2013. [12] 江志红, 丁裕国, 宋桂英. 黄淮流域夏半年旱涝概率时空分布的研究[J]. 自然灾害学报, 1998, 7(1): 96-106. [13] 张华龙, 肖柳斯, 陈生, 等. 基于GPM卫星的广东汛期降水日变化特征与评估[J]. 热带气象学报, 2020, 36(3): 335-346. [14] TURK F J, EBERT E E, OH H J, et al. Validation of an Operational Global Precipitation Analysis at Short Time Scales[C]//12th Conf on Satellite Meteorology and Oceanography, 2003. [15] HUFFMAN G J, ADLER R F, STOCKER E F, et al. Analysis of TRMM 3-hourly multi-satellite precipitation estimates computed in both real and post-real time[C]//12th Conference on Satellite Meteorology and Oceanography, 2003. [16] 周迪, 陈静, 陈朝平, 等. 暴雨集合预报-观测概率匹配订正法在四川盆地的应用研究[J]. 暴雨灾害, 2015, 34(2): 97-104. [17] WOOLHISER D A. Modeling daily precipitation-process and problems[M]// Statistics in the Environmental & Earth Sciences, London: Walden AT, Guttorp P Halsted Press, 1992: 71-89. [18] JOYCE R J, JANOWIAK J E, ARKIN P A, et al. CMORPH: A method that produces global precipitation estimates from passivemicrowave and infrared data at high spatial and temporal resolution[J]. J Hydrometeor, 2004, 5(3): 487-503. [19] 冯锦明, 赵天保, 张英娟. 基于台站降水资料对不同空间内插方法的比较[J]. 气候与环境研究, 2004, 9(2): 261-277. [20] 宣腾. 基于克里金法的地质勘探位置分析[D]. 哈尔滨: 哈尔滨工业大学, 2016. [21] 彭芳, 吴古会, 杜小玲. 贵州省汛期短时降水时空特征分析[J]. 气象, 2012, 38(3): 307-313. [22] 林爱兰, 谷德军, 彭冬冬, 等. 体现大尺度特征的区域持续性强降水过程定义指标[J]. 热带气象学报, 2020, 36(3): 289-298. [23] 黄桢, 李双林, 张超. 1991、1998和2016年三个大水年长江中下游夏季降水季节内特征的对比[J]. 热带气象学报, 2020, 36(1): 13-24.