基于随机森林的遥感干旱监测模型的构建

doi:10.3724/SP.J.1047.2017.00125

摘要/Abstract

摘要：

利用遥感数据进行大面积旱情监测是现有干旱监测的重要方法之一,然而传统的遥感干旱监测方法主要侧重于对土壤湿度或植被状况等单一干旱响应因子进行监测,对综合多因子的干旱监测研究较为有限。随机森林是一种机器学习方法,具有学习过程快速、运算速度快、稳定性好、预测精度高的优点,近年来被应用于生态环境等多个领域。本文利用2001-2010年4-9月的MODIS数据提取的植被状态指数（VCI）、温度状态指数（TCI）和土地覆盖类型（LC）,TRMM降水资料计算的TRMM-Z指数及SRTM-DEM、土壤有效含水量（AWC）等多个遥感及土壤资料提取的干旱因子为自变量,以气象站点的综合气象干旱指数（CI）为因变量,利用随机森林模型构建遥感干旱监测模型,并以河南省为研究区进行了评价和分析。该模型在2009-2010年的监测值和实测CI值的具有显著的相关性,并且二者干旱等级的一致率为81%。在2001-2010年4-9月间,模型监测值与气象站点的标准降水蒸散发指数（SPEI）总体干旱等级一致率为74.9%,较为一致,其中9月的模型结果与SPEI的干旱等级一致率最高,达到82.4%,空评估率和漏评估率最低;与10 cm土壤相对湿度的相关系数在0.475-0.639之间,达到极显著水平。河南省2011年4-6月干旱事件同样验证了本文构建的模型旱情监测结果,说明本模型能较好地就应用于监测区域旱情监测。

关键词: 干旱, 遥感, 随机森林

Abstract:

The drought detection of a large area by using the remote sensing data has been an important method in drought monitoring. However, the conventional remote sensing methods mainly focus on some single drought response factors, such as the soil moisture or vegetation status, and the drought monitoring study that integrated with multiple factors is relatively limited. In order to explore the relationships among multiple drought factors, a random forest algorithm was applied. Random forest is a machine learning method, which has many advantages such as being accurate, handy, fast and stable, and it has been used in many fields in recent years. In this paper, a remote sensing drought model was developed using the random forest algorithm and the multi-source remote sensing data, including MODIS, TRMM and SRTM-DEM. Vegetation Condition Index (VCI), Temperature Condition Index (TCI), Land Cover types (LC), TRMM-Z, DEM and Available Water Capacity (AWC), which were extracted from the remote sensing data and other soil data, were used as the independent variables, and the comprehensive meteorological drought index (CI) was used as the dependent variable. The training and testing experiments were carried out in Henan Province from April to September annually between 2001 and 2010. The results showed that the model and CI had highly significant correlation and their concordance rate reached 81% with respect to the drought classes from 2009 to 2010. In the study case′s period, the overall concordance rate was 74.9% between the model results and the Standardized Precipitation Evapotranspiration Index (SPEI) of the meteorological stations, from which the concordance was found to be the highest and the vacancy and miss rate was the lowest in September. The correlation between the model results and the soil moisture within 10 cm depth was highly significant, and their correlation coefficient varied between 0.475 and 0.639, which indicated that this model could effectively detect the agriculture drought. In addition, the drought event of Henan Province from April to June in 2011 was simulated by the proposed model, and the results could reflect the actual drought situation and its spatial variation. Therefore, this method could be well applied to monitor regional drought events.

Key words: drought, remote sensing, random forest

沈润平, 郭佳, 张婧娴, 李洛晞. 基于随机森林的遥感干旱监测模型的构建[J]. 地球信息科学学报, 2017, 19(1): 125-133.DOI:10.3724/SP.J.1047.2017.00125

SHEN Runping,GUO Jia,ZHANG Jingxian,LI Luoxi. Construction of a Drought Monitoring Model Using the Random Forest Based Remote Sensing[J]. Journal of Geo-information Science, 2017, 19(1): 125-133.DOI:10.3724/SP.J.1047.2017.00125

图/表 7

图1

表1

表2

表3

图2

图3

图4

参考文献 29

[1]	唐侥,孙睿.基于气象和遥感数据的河南省干旱特征分析[J].自然资源学报,2013,28(4):646-655. doi: 10.11849/zrzyxb.2013.04.011
	[ Tang R, Sun R.Drought characteristics in Henan province with meteorological and remote sensing data[J]. Journal of Natural Resources, 2013,28(4):646-655. ] doi: 10.11849/zrzyxb.2013.04.011
[2]	唐巍,覃志豪,秦晓敏.农业干旱遥感监测业务化运行方法研究[J].遥感信息,2007(2):37-41. doi: 10.3969/j.issn.1000-3177.2007.02.009
	[ Tang W, Tan Z H, Qing X M.Operational method for crop drought monitoring[J]. Remote Sensing Information, 2007,2:37-41. ] doi: 10.3969/j.issn.1000-3177.2007.02.009
[3]	孙丽,王飞,李保国,等.基于多源数据的武陵山区干旱监测研究[J].农业机械学报,2014,45(1):246-252. doi: 10.6041/j.issn.1000-1298.2014.01.038
	[ Sun L, Wang F, Li B G, et al.Study on drought monitoring of Wuling Mountain area based on multi-source data[J].Transactions of the Chinese Society of Agricultural Machinery, 2014,45(1):246-252.] doi: 10.6041/j.issn.1000-1298.2014.01.038
[4]	Rhee J, Im J, Carbone G J.Monitoring agricultural drought for arid and humid regions using multi-sensor remote sensing data[J]. Remote Sensing of Environment, 2010,114(12):2875-2887. doi: 10.1016/j.rse.2010.07.005
[5]	Brown J F, Wardlow B D, Tadesse T, et al.The Vegetation Drought Response Index (VegDRI): A new integrated approach for monitoring drought stress in vegetation[J]. Giscience & Remote Sensing, 2008,45(1):16-46. doi: 10.2747/1548-1603.45.1.16
[6]	Wu J, Zhou L, Liu M, et al.Establishing and assessing the Integrated Surface Drought Index (ISDI) for agricultural drought monitoring in mid-eastern China[J]. International Journal of Applied Earth Observation & Geoinformation, 2013,23(8):397-410. doi: 10.1016/j.jag.2012.11.003
[7]	Du L, Tian Q, Yu T, et al.A comprehensive drought monitoring method integrating MODIS and TRMM dataInternational[J]. Journal of Applied Earth Observation & Geoinformation, 2013,23(8):245-253. doi: 10.1016/j.jag.2012.09.010
[8]	姚云军,秦其明,赵少华,等.基于ΔTs-Albedo光谱信息的土壤水分监测新指数研究[J].光谱学与光谱分析,2011,31(6):1557-1561. doi: 10.3964/j.issn.1000-0593(2011)06-1557-05
	[ Yao Y J, Qi Q M, Zhao S H, et al.New index for soil moisture monitoring based on ΔTs-Albedo spectral information[J]. Spectroscopy and Spectral Analysis, 2011,31(6):1557-1561. ] doi: 10.3964/j.issn.1000-0593(2011)06-1557-05
[9]	Yao Y, Liang S, Qin Q, et al.Monitoring drought over the conterminous United States using MODIS and NCEP reanalysis-2 data[J]. Journal of Applied Meteorology and Climatology, 2010,49:1665-1680. doi: 10.1175/2010JAMC2328.1
[10]	Breiman L.Random forests[J]. Machine Learning, 2001,45(1):5-32. doi: 10.1023/A:1010933404324
[11]	江佳乐,刘湘南,刘美玲,等.基于随机森林的香港海域海表盐度遥感反演模型[J].海洋通报,2014,33(3):333-341. doi: 10.11840/j.issn.1001-6392.2014.03.013
	[ Jiang J L, Liu X N, Liu M L, et al.Remote sensing retrieval model of sea surface salinity in Hong Kong waters based on the random forest[J]. Marine Science Bulletin, 2014,33(3):331-341. ] doi: 10.11840/j.issn.1001-6392.2014.03.013
[12]	范钦栋,梁国付,马顺兴.河南省土地利用下的生态系统服务变化分析[J].河南农业大学学报,2014,48(2):224-228.
	[ Fan Q D, Liang G F, Ma S X.Analysis on ecosystem service change based on land use in Henan Province[J]. Journal of Henan Agricultural University, 2014,48(2):224-228. ]
[13]	杜灵通,田庆久,黄彦,等.基于TRMM数据的山东省干旱监测及其可靠性检验[J].农业工程学报,2012,28(2):121-126. doi: 10.3969/j.issn.1002-6819.2012.02.022
	[ Du L T, Tian Q J, Huang Y, et al.Drought monitoring based on TRMM data and its reliability validation in Shandong province[J]. Transactions of the Chinese Society of Agricultural Engineering, 2012,28(2):121-126. ] doi: 10.3969/j.issn.1002-6819.2012.02.022
[14]	李景刚,李纪人,黄诗峰,等.基于TRMM数据和区域综合Z指数的洞庭湖流域近10年旱涝特征分析[J].资源科学,2010,32(6):1103-1110.
	[ Li J G, Li J R, Huang S F, et al.Characteristics of the recent 10-year flood/drought over the Dongting Lake Basin based on TRMM precipitation data and regional integrated Z-index[J]. Resources Science, 2010,32(6):1103-1110. ]
[15]	王峥,Nguyen,Tinh,马孝意,等.基于SRTM_DEM的泾河流域特征信息提取研究[J].中国农村水利水电,2011(11):32-36.
	[ Wang Z, Nguyen, Tinh, Ma X Y, et al. Research on extraction of hydrological information in the Jinghe River Basin based on SRTM_DEM[J]. China Rural Water and Hydropower, 2011,11:32-36. ]
[16]	周丹,张勃,罗静,等.基于SPEI的华北地区近50年干旱发生强度的特征及成因分析[J].自然灾害学报,2014,23(4):192-202.
	[ Zhou D, Zhang B, Luo J, et al.SPEI-based intensity characteristics and cause analysis of drought in north China during recent 50 years[J]. Journal of Natural Disasters, 2014,23(4):192-202. ]
[17]	Vicente-Serrano S M, Beguería S, López-Moreno J I. A multiscalar drought index sensitive to global warming: the standardized precipitation evapotranspiration index[J]. Journal of Climate, 2010,23(7):1696-1718.
[18]	GBT 20481-2006, 气象干旱等级[S].2006:40-41.
	[ GBT 20481-2006, Meteorological drought grade[S]. 2006:40-41. ]
[19]	史本林,朱新玉,胡云川,等.基于SPEI指数的近53年河南省干旱时空变化特征[J].地理研究,2015,34(8):1547-1558. doi: 10.11821/dlyj201508012
	[ Shi B L, Zhu X Y, Hu Y C, et al.Spatial and temporal variations of drought in Henan province over a 53-year period based on standardized precipitation evapotranspiration index[J]. Geographical Research, 2015,34(8):1547-1558. ] doi: 10.11821/dlyj201508012
[20]	杜灵通. 基于多源空间信息的干旱监测模型构建及其应用研究[D].南京:南京大学,2013.
	[ Du L T.Drought monitoring model based on multi-source spatial information and its application[D]. Nanjing: Nanjing University, 2013. ]
[21]	Gupta S C, Larson W E.Estimating soil water retention characteristics from particle size distribution, organic matter percent, and bulk density[J]. Water Resources Research, 1979,15(6):1633-1635. doi: 10.1029/WR015i006p01633
[22]	杜灵通,田庆久,王磊,等.基于多源遥感数据的综合干旱监测模型构建[J].农业工程学报, 2014,30(9):126-132. ] doi: 10.3969/j.issn.1002-6819.2014.09.016
	[ Du L T, Tian Q J, Wang L, et al. A synthesized drought monitoring model based on multi-source remote sensing data[J]. Transactions of the Chinese Society of Agricultural Engineering, 2014,30(9):126-132. ] doi: 10.3969/j.issn.1002-6819.2014.09.016
[23]	于敏,王春丽.不同卫星遥感干旱指数在黑龙江的对比应用[J].应用气象学报,2011,22(2):221-231. doi: 10.3969/j.issn.1001-7313.2011.02.011
	[ Yu M, Wang C L.Satellite remote sensing drought monitoring methods based on different biophysical indicators[J]. Journal of Applied Meteorological Science, 2011,22(2):221-231. ] doi: 10.3969/j.issn.1001-7313.2011.02.011
[24]	计淇才. 遥感干旱指数的时空格局适应性研究[D].郑州:郑州大学,2011.
	[ Ji Q C.The adaptation research on spatiotemporal patterns of remote sensing drought index[D]. Zhengzhou: Zhengzhou University, 2013. ]
[25]	夏文韬, 王莺, 冯琦胜,等. 甘南地区MODIS土地覆盖产品精度评价[J].草业科学, 2010,27(9):11-18.
	[ Xia W T, Wang Y, Feng Q S, et al.Accuracy assessment of MODIS land cover product of Gannan prefecture[J]. Pratacultural Science, 2010,27(9):11-18. ]
[26]	谢五三,田红,王胜,等.基于CI指数的淮河流域干旱时空特征研究[J].气象,2013,39(9):1171-1175. doi: 10.7519/j.issn.1000-0526.2013.09.012
	[ Xie W S, Tian H, Wang S, et al.Study on spatial-temporal characteristics of drought in Huaihe River Basin based on CI index[J]. Meteorological Monthly, 2013,39(9):1171-1175. ] doi: 10.7519/j.issn.1000-0526.2013.09.012
[27]	甄亿位,郝敏,陆宝宏,等. 基于随机森林的中长期降水量预测模型研究[J].水电能源科学,2015,33(6):6-10.
	[ Zhen Y W, Hao M, Lu B H, et al.Research of medium and long term precipitation forecasting model based on random forest[J]. Water Resources and Power, 2015,33(6):6-10. ]
[28]	张岳军,郝智文,王雁,等.基于SPEI和SPI指数的太原多尺度干旱特征与气候指数的关系[J].生态环境学报,2014,23(9):1418-1424.
	[ Zhang Y J, Hao Z W, Wang Y, et al.Multiscale characteristics of drought based on SPEI and SPI in association with climate index in Taiyuan[J]. Ecology and Environmental Sciences, 2014,23(9):1418-1424. ]
[29]	庄少伟. 基于标准化降水蒸发指数的中国区域干旱化特征分析[D].兰州:兰州大学,2013.
	[ Zhuang S W.The drought characteristics in the regions of China based on the standardized precipitation evapotranspiration index[D]. Lanzhou: Lanzhou University, 2013. ]

等级	类型	综合气象干旱指数值	标准化蒸散发指数值
1	无旱	-0.6<CI	-0.5<SPEI
2	轻旱	-1.2<CI≤-0.6	-1<SPEI≤-0.5
3	中旱	-1.8<CI≤-1.2	-1.5<SPEI≤-1
4	重旱	-2.4<CI≤-1.8	-2<SPEI≤-1.5
5	特旱	CI≤-2.4	SPEI≤-2

	平均相对误差绝对值/%	均方根误差	相关系数
训练集（2001-2008年）	2.853	0.377	0.955^**
测试集（2009-2010年）	2.573	0.771	0.695^**

	4月	5月	6月	7月	8月	9月
一致率/%	65.9	64.7	74.4	81.8	79.4	82.4
空评估率/%	21.6	21.2	12.2	12.9	9.09	8.24
漏评估率/%	34.5	29.6	17.6	13.4	11.42	8.16