Abstract：A spatial downscaling framework based on machine learning algorithms integrated with station data correction is proposed by combining enhanced vegetation index (EVI), precipitable water vapor (PWV), Land Surface Temperature (LST), Slope, Aspect, Altitude, LON, and LAT. Then, the robustness differences among three machine learning algorithms, namely Random Forest (RF), Feedforward Neural Network (FNN), and Gradient Boosting Decision Tree (DBGT), were evaluated for spatially downscaling the GPM IMERG over the Yangtze River basin. The time-lag effect of different seasonal vegetation on the downscaling results of precipitation from 2001~2019 was taken into account to explore how the correction effect of the downscaling framework on GPM is related to the precipitation. The results show that the three downscaling approaches can improve the accuracy of the data to different degrees while obtaining the GPM IMERG data with 1km spatial resolution. Specifically, the simulations of GBDT can express the detailed features of precipitation more obviously in terms of the spatial heterogeneity with the best simulation accuracy and stronger model robustness at annual and seasonal time scales in each study area (annual R2=0.748~0.958, seasonal R2=0.518~0.909). The hysteresis of vegetation in spring and winter were most sensitive to the response of GPM downscaling results, with the best hysteresis of 1and 2months in order. Yet the time lag effect was not significant in summer and autumn, with no lag period essentially. Compared to the original data, the increased R2of the downscaled data was positively correlated with precipitation at R2 of 0.630~0.844.
张寒博, 杨骥, 荆文龙, 邓应彬. 多种特征因子结合GBDT的降水数据降尺度方法研究[J]. 中国环境科学, 2023, 43(4): 1867-1882.
ZHANG Han-bo, YANG Ji, JING Wen-long, DENG Ying-bin. Downscaling method of precipitation data based on GBDT combined with multiple eigenfactors. CHINA ENVIRONMENTAL SCIENCECE, 2023, 43(4): 1867-1882.
阎广建,赵天杰,穆西晗,等.滦河流域碳、水循环和能量平衡遥感综合试验总体设计[J]. 遥感学报, 2021,25(4):856-870. Yan G J, Zhao T J, Mu X H, et al. Comprehensive remote sensing experiment of carbon cycle, water cycle and energy balance in Luan River Basin[J]. National Remote Sensing Bulletin, 2021,25(4):856-870.
[2]
蒋云钟,冶运涛,赵红莉,等.智慧水利解析[J]. 水利学报, 2021, 52(11):1355-1368. Jiang Y Z, Ye Y T, Zhao H L, et al. Analysis of smart water conservancy[J]. Journal of Hydraulic Engineering, 2021,52(11):1355-1368.
[3]
范田亿,张 翔,黄 兵,等.湘江流域TRMM卫星降水产品降尺度研究与应用[J]. 自然资源遥感, 2021,33(4):209-218. Fan T Y, Zhang X, Huang B, et al. Downscaling of TRMM precipitation products and its application in Xiangjiang River basin[J]. Remote Sensing for Natural Resources, 2021,33(4):209-218.
[4]
Zhang Z X, Tian J X, Huang Y H, et al. Hydrologic evaluation of TRMM and GPM IMERG satellite-based precipitation in a Humid Basin of China[J]. Remote Sensing, 2019,11(4):431.
[5]
Chen C, Chen Q W, Duan Z, et al. Multiscale comparative evaluation of the GPM IMERG v5and TRMM 3B42v7 precipitation products from 2015 to 2017 over a climate transition area of China[J]. Remote Sensing, 2018,10:944.
[6]
杜 懿,王大洋,张 智,等.GPM IMERG降水产品在珠江流域的适用性分析[J]. 水文, 2021,41(6):1-6. Du Y, Wang D Y, Zhang Z, et al. Applicability of GPM IMERG satellite precipitation product in Pearl River Basin[J]. Journal of China Hydrology, 2021,41(6):1-6.
[7]
Chen J, Wang Z, Wu X, et al. Accuracy evaluation of GPM multi-satellite precipitation products in the hydrological application over alpine and gorge regions with sparse rain gauge network[J]. Hydrology Research, 2019,50:1710-1729.
[8]
王书涛,康 旺,孔德明,等.基于回归克里金的高分-1影像融合[J]. 激光与光电子学进展, 2022,59(8):523-531. Wang S T, Kang W, Kong D M, et al. GF-1image fusion based on Regression Kriging[J]. Laser & Optoelectronics Progress, 2022,59(8):523-531.
[9]
张寒博,韦梦思,覃金兰,等.基于不同植被指数的TRMM数据降尺度及误差校正研究[J]. 水土保持通报, 2021,41(4):214-223. Zhang H B, Wei M S, Qin J L, et al. Downscaling and error correction of TRMM data based on different vegetation indices[J]. Bulletin of Soil and Water Conservation, 2021,41(4):214-223.
[10]
Jing W L, Zhang P Y, Jiang H, et al. Reconstructing satellite-based monthly precipitation over Northeast China using machine learning algorithms[J]. Remote Sensing, 2017,9(8):781.
[11]
Zhang X X, Liu G D, Wang H T, et al. Application of a hybrid interpolation method based on support vector machine in the precipitation spatial interpolation of basins[J]. Water, 2017,9(10):760.
[12]
Chen F H, Wu S H, Cui P, et al. Progress and prospects of applied research on physical geography and the living environment in China over the past 70years (1949-2019)[J]. Journal of Geographical Sciences, 2021,31(1):3-45.
[13]
Sharifi E, Saghafian B, Steinacker R. Downscaling satellite precipitation estimates with multiple linear regression, artificial neural networks, and spline interpolation techniques[J]. Journal of Geophysical Research:Atmospheres, 2019,124(2):789-805.
[14]
Zhang W, Huai B, Sun W, et al. The spatial downscaling of TRMM precipitation data for the middle part of the Chinese Tianshan Mountains[J]. Arabian Journal of Geosciences, 2019,12(18):573.
[15]
Chen S, Zhang L, She D, et al. Spatial Downscaling of Tropical Rainfall Measuring Mission (TRMM) annual and monthly precipitation data over the middle and lower reaches of the Yangtze River Basin, China[J]. Water, 2019,11(3):568.
[16]
Lopez P L, Immerzeel W W, Sandoval E A R, et al. Spatial downscaling of satellite-based precipitation and its impact on discharge simulations in the Magdalena River Basin in Colombia[J]. Frontiers in Earth Science, 2018,6:68.
[17]
Chen C, Chen Q, Qin B, et al. Comparison of different methods for spatial downscaling of GPM IMERG V06B satellite precipitation product over a typical arid to semi-arid area[J]. Frontiers in Earth Science, 2020,8(525):1-16.
[18]
Jia S F, Zhu W B, Lu A F, et al. A statistical spatial downscaling algorithm of TRMM precipitation based on NDVI and DEM in the Qaidam Basin of China[J]. Remote Sensing of Environment, 2011, 115(12):3069-3079.
[19]
黄 涛,徐力刚,范宏翔,等.长江流域干旱时空变化特征及演变趋势[J]. 环境科学研究, 2018,31(10):1677-1684. Huang T, Xu L G, Fan H X, et al. Temporal and spatial variation characteristics and the evolution trends of droughts in the Yangtze River Basin[J]. Research of Environmental Sciences, 2018,31(10):1677-1684.
[20]
张寒博,徐 勇,窦世卿,等.基于GWR模型的长江流域TRMM数据降尺度[J]. 水土保持研究, 2021,28(3):149-155. Zhang H B, Xu Y, Dou S Q, et al. TRMM downscaling data of Yangtze based on GWR model[J]. Research of Soil and Water Conservation, 2021,28(3):149-155.
[21]
曾岁康,雍 斌.全球降水计划IMERG和GSMaP反演降水在四川地区的精度评估[J]. 地理学报, 2019,74(7):1305-1318. Zeng S K, Yong B, Yang Z K. Evaluation of the GPM-based IMERG and GSMaP precipitation estimates over the Sichuan region[J]. Acta Geographica Sinica, 2019,74(7):1305-1318.
[22]
祝新明,宋小宁,冷 佩,等.多尺度地理加权回归的地表温度降尺度研究[J]. 遥感学报, 2021,25(8):1749-1766. Zhu X M, Song X N, Leng P, et al. Spatial downscaling of land surface temperature with the multi-scale geographically weighted regression National[J]. Remote Sensing Bulletin, 2021,25(8):1749-1766.
[23]
Zhang D, Wen Z, Huang W, et al. Upscaling of surface soil moisture using a deep learning model with VIIRS RDR[J]. ISPRS International Journal of Geo-Information, 2017,6(5):1-20.
[24]
盛佳慧,饶 鹏.基于风云气象卫星的土壤湿度数据降尺度方法研究[J]. 红外与毫米波学报, 2021,40(1):74-88. Sheng J H, Rao P. The research on downscaling methods based on Fengyun meteorological satellite soil moisture data[J]. Journal of Infrared and Millimeter Waves, 2021,40(1):74-88.
[25]
Zhang D Y, Zhang W, Huang W, et al. Upscaling of surface soil moisture using a deep learning model with VIIRS RDR[J]. ISPRS International Journal of Geo-Information, 2017,6(5):130.
[26]
Friedman J H. Stochastic gradient boosting[J]. Computational Statistics & Data Analysis, 2002,38:367-378.
[27]
Johnson R, Zhang T. Learning nonlinear functions using regularized greedy forest[J]. Ieee T Pattern Anal, 2014,36(5):942-954.
[28]
谷宇峰,张道勇,鲍志东,等.利用梯度提升决策树(GBDT)预测渗透率——以姬塬油田西部长4+5段致密砂岩储层为例[J]. 地球物理学进展, 2021,36(2):585-594. Gu Y F, Zhang D Y, Bao Z D, et al. Permeability prediction using Gradient Boosting Decision Tree(GBDT):a case study of tight sandstone reservoirs of member of Chang 4+5 in western Jiyuan Oilfield[J]. Progress in Geophysics(in Chinese), 36(2):585-594.
[29]
Yan X, Chen H, Tian B, et al. A downscaling-merging scheme for improving daily spatial precipitation estimates based on random forest and cokriging[J]. Remote Sensing, 2021,13(11):2040.
[30]
杨晋云,刘晓辉,刘 蛟,等.不同偏差校正方法对青藏高原地区GPM的应用效果研究[J]. 山地学报, 2021,39(3):439-449. He K L, Liu X H, Liu J, et al. Performances of different deviation calibration methods on GPM in the Qinghai-Tibet Plateau, China[J]. Mountain Research, 2021,39(3):439-449.
[31]
丁明泽,雍 斌,杨泽康.全球降水观测计划多卫星联合反演降水产品的极端降水监测潜力研究[J]. 遥感学报, 2022,26(4):657-671. Ding M Z,Yong B, Yang Z K. Extreme precipitation monitoring capability of the multi-satellite jointly retrieval precipitation products of Global Precipitation Measurement(GPM)mission[J]. National Remote Sensing Bulletin, 2022,26(4):657-671.
[32]
Li X, Chen Y B, Deng X C, et al. Evaluation and hydrological utility of the GPM IMERG precipitation products over the Xinfengjiang River Reservoir Basin, China[J]. Remote Sensing, 2021,13(5):866.
[33]
Elnashar A, Zeng H W, Wu B F, et al. Downscaling TRMM monthly precipitation using Google Earth Engine and Google Cloud Computing[J]. Remote Sensing, 2020,12(23):3860.
[34]
Li Y G, Zhang Y Y, He D M, et al. Spatial downscaling of the tropical rainfall measuring mission precipitation using geographically weighted regression kriging over the Lancang River Basin, China[J]. Chinese Geographical Science, 2019,29(3):86-102.
[35]
Nemani R R. Climate-driven increases in global terrestrial net primary production from 1982 to 1999[J]. Science, 2003,300(5625):1560-1563.
[36]
Zhan C, Jian H, Shi H, et al. Spatial downscaling of GPM annual and monthly precipitation using regression-based algorithms in a mountainous area[J]. Advances in Meteorology, 2018,2018(2):1-13.
