基于反射率和亮度温度的锡林郭勒积雪深度估算

doi:10.12118/j.issn.1000-6060.2020.06.18

摘要/Abstract

摘要： 积雪是我国西北干旱半干旱区重要的水资源，也是影响全球气候变化的重要因子之一。目前光学影像反射率和雷达亮温数据是积雪遥感领域的主要数据，本文首次结合两类遥感数据估算积雪深度，并比较偏最小二乘法和机器学习算法（人工神经网络、支持向量机和随机森林算法）在积雪深度估算方面的表现。以锡林郭勒盟 2012—2015 年积雪深度数据为例，基于反射率和亮度温度相结合的积雪深度估算精度优于单个数据源，且随机森林算法表现最好，均方根误差为 2.93 cm，满足实际应用的需求。研究结果对我国西北地区水资源分布、生态环境评估等研究具有重要意义。

关键词: 积雪深度, MODIS, MWRI, 机器学习

Abstract: Snow is an important water resource in arid and semi-arid regions of northwest China, which is also one of the important factors that affect the global climate change. Snow depth is the main parameter of the water resources and its estimation using a remote sensing technique has a vital effect on the assessment of resources and disease of northern China. The reflectance and brightness temperature data were the main data for snow remote sensing. Taking the snow depth data of 2012—2015 for Xilin Gol League, Inner Mongolia, China as an example, The MODIS/Terra satellite reflectance data and FY-3B brightness temperature data for the estimation of snow depth are combined for the first time in this study. The estimation accuracy for partial least squares and machine learning (artificial neural network, support vector machine and random forest) are then compared. When only reflectivity data is used, best accuracy is achieved by the random forest algorithm (RMSE=3.65 cm), followed by the neural network and support vector machine. The stochastic forest algorithm still performs best (RMSE=3.12 cm) when only the brightness and temperature data are used. However, no obvious difference is found with the artificial neural network. Using the partial least squares method exhibits the worst performance among the methods used. For the four methods, the accuracy of estimating snow depth using brightness temperature data is higher than that using reflectivity data, which may be attributed to the fact that reflectivity data is easy to saturate. Comprehensive research shows that the snow depth estimation accuracy based on the combination of reflectance and brightness temperature was better than that with single data. The random forest algorithm obtained the best performance, which is able to meet the needs of practical application. However, the accuracy of the three machine learning algorithms is not much different, with RMSE of 2.93- 3.92 cm, which is far lower than that of the partial least squares method (RMSE=7.12 cm). Although partial least squares method can eliminate the collinearity of variables, its performance in snow depth estimation is poor, which may be attributed to its difficulty in adapting to complex snow scenes. Results of this paper have important significance for the study of water resources distribution and ecological environment assessment in northern China.

Key words: snow depth, MODIS, MWRI, machine learning

李慧融. 基于反射率和亮度温度的锡林郭勒积雪深度估算[J]. 干旱区地理, 2020, 43(6): 1567-1572.

LI Hui-rong. Estimation of snow depth based on reflectance and bright temperature in Xilin Gol League[J]. Arid Land Geography, 2020, 43(6): 1567-1572.

参考文献 27

[1]	SALOMONSON V V, APPEL I. Estimating fractional snow cover from MODIS using the normalized difference snow index[J]. Re⁃ mote Sensing of Environment, 2003, 89(3): 351-360.
[2]	刘艳. 基于 MODIS 数据的积雪深度反演研究[D]. 武汉: 武汉大学, 2005. [LIU Yan. The study of snow depth derivation from MO⁃ DIS data[D]. Wuhan: Wuhan University, 2005. ]
[3]	姜康, 包刚, 乌兰图雅, 等. 基于 MODIS 数据的蒙古高原积雪时空变化研究 [J]. 干旱区地理, 2019, 42(4): 782- 789. [JIANG Kang, BAO Gang, WULANTUYA, et al. Spatiotemporal changes of snow cover in Mongolian Plateau based on MODIS data[J]. Ar⁃ id Land Geography, 2019, 42(4): 782-789. ]
[4]	周远刚, 赵锐锋, 张丽华, 等. 博格达峰地区冰川和积雪变化遥感监测及影响因素分析 [J]. 干旱区地理, 2019, 42(6): 1395- 1403. [ZHOU Yuangang, ZHAO Ruifeng, ZHANG Lihua, et al. Remote sensing monitoring of the change of glacier and snow cov⁃ er and its influencing factors in Mount Bogda[J]. Arid Land Ge⁃ ography, 2019, 42(6): 1395-1403. ]
[5]	李震, 邵雨阳, 周建民, 等. 基于 MEMLS 模型的积雪深度反演方法[J]. 遥感学报, 2013, 17(3): 657-670. [LI Zhen, SHAO Yuyang, ZHOU Jianmin, et al. 2013. Snow- depth retrieval algorithms on MEMLS [J]. Journal of Remote Sensing, 17( 3): 657－670. ]
[6]	陈乾, 陈添宇. 祁连山区季节性积雪资源的气候分析[J]. 地理研究, 1991, 10(1): 24- 38. [CHEN Qian, CHEN Tianyu. Climatical analysis of seasonal snow resources in Qilian MT[J]. Geographical Research, 17(3): 657－670. ]
[7]	周咏梅, 贾生海, 刘萍. 利用 NOAA-AVHRR 资料估算积雪参量 [J]. 气象科学, 2001, 21(1): 117-121. [ZHOU Yongmei, JIA Shen⁃ hai, LIU Ping. The method of snowcover parameters estimation us⁃ ing NOAA- AVHRR date[J]. Scientia Meteorologica Sinica, 2001, 21(1): 117-121. ]
[8]	孙知文, 于鹏珊, 夏浪, 等. 被动微波遥感积雪参数反演方法进展[J]. 国土资源遥感, 2015, 27(1): 9-15. [SUN Zhiwen, YU Peng⁃ shan, XIA Lang. Progress in study of snow parameter inversion by passive microwave remote sensing[J]. Remote Sensing For Land & Resources, 2015, 27(1): 9-15. ]
[9]	CHANG A T C, FOSTER J L, HALL D K. Nimbus- 7 SMMR de⁃ rived global snow cover parameters[J]. Annals of Glaciology, 1987, (9): 39-44.
[10]	于惠, 张学通, 王玮, 等. 基于 AMSR-E 数据的青海省雪深遥感监测模型及其精度评价[J]. 干旱区研究, 2011, 28(2): 255-261. [YU Hui, ZHANG Xuetong, WANG Wei, et al. Monitoring model and accuracy evaluation of snow depth in Qinghai Province based on AMSR-E data[J]. Arid Zone Research, 2011, 28(2): 255-261. ]
[11]	卢新玉. 基于被动微波遥感的北疆地区积雪深度反演[D]. 乌鲁木齐: 新疆师范大学, 2011. [LU Xinyu. Snow depth retrieval based on the passive microwave remote sensing of northern Xinji⁃ ang[D]. Urumqi: Xinjiang Normal University, 2011. ]
[12]	卢新玉, 王秀琴, 谢国辉, 等. 基于被动微波遥感的积雪深度反演[J]. 干旱区资源与环境, 2012, 12: 108-112. [LU Xinyu, WANG Xiuqin, XIE Guohui, et al. Snow depth retrieval based on the pas⁃ sive microwave remote sensing[J]. Journal of Arid Land Resources and Environment, 2012, 12: 108-112. ]
[13]	蒋玲梅, 王培, 张立新, 等. FY3B-MWRI 中国区域雪深反演算法改进 [J]. 中国科学: 地球科学, 2014, 44(3): 531- 547. [JIANG Lingmei, WANG Pei, ZHANG Lixin, et al. Improvement of snow depth retrieval for FY3B-MWRI in China[J]. Scientia Sinica（Ter⁃ rae）, 2014, 44 (3): 531-547. ]
[14]	王功雪, 蒋玲梅, 武胜利, 等. FY-3B 与 FY-3C/MWRI 交叉定标及雪深算法应用 [J]. 遥感技术与应用, 2017, 32(1): 49- 56. [WANG Gongxue, JIANG Lingmei, WU Shengli, et al. Intercali⁃ brating FY-3B and FY-3C/MWRI for synergistic implementing to snow depth retrieval algorithm[J]. Remote Sensing Technology and Application, 2017, 32(1): 49-56. ]
[15]	刘宝康, 冯蜀青, 杜玉娥, 等. 积雪被动微波遥感研究进展与前景展望[J]. 草业科学, 2009, 26(11): 37-43. [LIU Baokang, FENG Shuqing, DU Yu e, et al. Research progress and prospect of re⁃ mote sensing with passive microwave from snow[J]. Pratacultural Science, 2009, 26(11): 37-43. ]
[16]	TEDESCO M, MILLER J. Northern hemisphere snow-covered area mapping: Optical versus active and passive microwave data[J]. IEEE Geoscience & Remote Sensing Letters, 2007, 4(2): 221- 225.
[17]	邵东航, 李弘毅, 王建, 等. 基于多源遥感数据的积雪反照率反演研究[J]. 遥感技术与应用, 2017, 32(1): 71-77. [SHAO Dong⁃ hang, LI Hongyi, WANG Jian, et al. Retrieval of snow albedo based on multi-source remote sensing data[J]. Remote Sensing Tech⁃ nology and Application, 2017, 32(1): 71-77. ]
[18]	徐倩, 刘志辉, 房世峰. 融雪期积雪深度高光谱反演方法研究 [J]. 光谱学与光谱分析, 2013, 33(7): 1927-1931. [XU Qian, LIU Zhihui, FANG Shifeng. Retrieval method for estimating snow depth using hyperspectral data in snowmelt period[J]. Spectrosco⁃ py and Spectral Analysis, 2013, 33(7): 1927-1931. ]
[19]	王超, 王建明, 冯美臣, 等. 基于多变量统计分析的冬小麦长势高光谱估算研究 [J]. 光谱学与光谱分析, 2018, 38(5): 1520- 1525. [WANG Chao, WANG Jianming, FENG Meichen, et al. Hy⁃ perspectral estimation on growth status of winter wheat by using the multivariate statistical analysis[J]. Spectroscopy and Spectral Analysis, 2018, 38(5): 1520-1525. ]
[20]	SHA Z Y, XIE Y C, TAN X C, et al. Assessing the impacts of hu⁃ man activities and climate variations on grassland productivity by partial least squares structural equation modeling (PLS- SEM) [J]. Journal of Arid Land, 2017, 9(4): 473-488.
[21]	侯海艳, 侯金亮, 黄春林, 等. 基于人工神经网络和 AMSR2 多频微波亮温的北疆地区雪深反演[J]. 遥感技术与应用, 2018, 33 (2): 241- 251. [HOU Haiyan, HOU Jinliang, HUANG Chunlin, et al. Retrieve snow depth of north of Xinjiang region from ARMS2 data based on artificial neural network technology[J]. Remote Sens⁃ ing Technology and Application, 2018, 33(2): 241-251. ]
[22]	MOUNTRAKIS G, IM J, OGOLE C. Support vector machines in re⁃ mote sensing: A review[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2011, 66: 247-259.
[23]	ABDEL-RAHMAN E M, AHMED F B, ISMAIL R. Random forest regression and spectral band selection for estimating sugarcane leaf nitrogen concentration using EO-1 Hyperion hyperspectral da⁃ ta[J]. International Journal of Remote Sensing, 2013, 34(2): 712- 728.
[24]	于新洋, 赵庚星, 常春艳, 等. 随机森林遥感信息提取研究进展及应用展望 [J]. 遥感信息, 2019, 34(2): 8- 14. [YU Xinyang, ZHAO Gengxing, CHANG Chunyan, et al. Random forest classifi⁃ er in remote sensing information extraction: A review of applica⁃ tions and future development[J]. Remote Sensing Information, 2019, 34(2): 8-14. ]
[25]	包青岭, 丁建丽, 王敬哲, 等. 基于随机森林算法的土壤有机质含量高光谱检测 [J]. 干旱区地理, 2019, 42(6): 1404- 1414. [BAO Qingling, DING Jianli, WANG Jingzhe, et al. Hyperspectral detection of soil organic matter content based on random forest al⁃ gorithm[J]. Arid Land Geography, 2019, 42(6): 1404-1414. ]
[26]	李长春, 徐轩, 包安明, 等. 基于 FY3B-MWRI 数据新疆区域积雪深度反演[J]. 遥感技术与应用, 2018, 18(6): 1030-1036. [LI Changchun, XU Xuan, BAO An- ming, et al. The study on snow depth retrieval in Xinjiang region based on FY3B-MWRI data[J]. Re⁃ mote Sensing Technology and Application, 2018, 18(6): 1030-1036. ]
[27]	岳继博, 李长春, 齐修东, 等. 天山雪深反演算法验证与分析[J]. 遥感信息, 2016, 31(4): 56-61. [YUE Jibo, LI Changchun, QI Xi⁃ udong, et al. Validation and analysis of snow depth inversion algo⁃ rithm in Tianshan Mountain[J]. Remote Sensing Information, 2016, 31(4): 56-61. ]