基于Ts-NDVI特征空间的吉林省西部农业干旱研究
|
摘要
近年来，干旱的频繁发生始终困扰着人类社会、经济的可持续发展，干旱对农业的影响尤为明显。吉林省西部地区作为吉林省农牧业主产区，受其自然地理与气候条件的影响，干旱频发，给当地工农业生产和人民生活造成了巨大损失。相对于传统的农业干旱监测，遥感技术以其快速、高效、低成本等的特性，越来越多的被应用于农业干旱监测中。地表温度和植被指数与植被生理、生长联系密切，成为研究农业干旱监测的重要指标。但单独利用地表温度或植被指数进行干旱监测时，会出现植被在受水分胁迫时短期内仍能保持原有绿色的时间滞后问题，从而降低旱情监测的准确度和实用性。因此将两者耦合进行干旱监测的方法成为当前干旱监测研究的一个重要方向。
本文综合运用GIS、RS和GPS技术，利用2000-2008年5-9月MODIS地表温度数据和地表反射率数据，建立Ts-NDVI特征空间，反演得到吉林省西部植被生长期的温度植被干旱指数（TVDI），分析了TVDI与Ts、NDVI、表层土壤湿度和气象数据的关系以及研究区2000-2008年月际和年际旱情时空分异。研究表明：
1）研究区Ts-NDVI特征空间的干、湿边并非呈一条直线，且不同时期特征空间的干、湿边具有相似的形状；干、湿边的斜率随着NDVI值的变化而发生变化；
2）TVDI与Ts、NDVI、采样点表层土壤湿度以及气象站点降水量具有一定关系，从关系中可以发现TVDI作为旱情评价指标有一定的合理性。TVDI与Ts、NDVI和采样点表层土壤湿度的相关性分析显示，TVDI与Ts呈显著正相关，与NDVI呈负相关,与土壤湿度呈显著负相关；从TVDI与气象站点降水量数据变化趋势来看，随着TVDI值的增大，气象站点降水量呈减小趋势；
3）2000-2008年的月际变化分析表明，研究区北部的湿润区总体变化不大；研究区各年的5-6月以中旱和重旱为主，易导致春旱发生，其中白城、洮南、通榆三地较为严重；各年的7-8月全区中旱和重旱面积有所减小，正常和轻旱面积增加，松原、前郭、长岭三地旱情减轻尤为明显，是旱情较轻的时间；到了9月，研究区干旱情况大体可分为两部分，大安以西的地区（包括大安）仍然会以中旱和重旱为主，而大安以东的地区则以正常和轻旱为主，部分地区受秋旱影响；
4）2000-2008年的年际变化分析表明，2000-2008年研究区全区主要以轻旱和中旱为主。研究区内由东向西分别呈现出由轻旱向重旱的过渡性变化，其中研究区西南部通榆地区出现重旱年份最多。2000-2008年年间，2007年的中旱和重旱面积比最大，高达71.5%，旱情最为严重；2005年中旱和重旱面积比最小，两者和为38.7%，是旱情最轻的一年。
Drought is one of major natural disasters that happens frequently inworld-wide in recent years, they not only disturbed social and economicsustainable development seriously, but also affected agriculture especially.The western region of Jilin Province where is the major agriculture andlivestock production in Jilin province is affected by drought frequently,because of the natural geographic and climate. Compare with thetraditional agricultural drought monitoring, remote sensing is more andmore applied in agricultural drought monitoring for its fast, efficient, lowcost and other characteristics. Land surface temperature and vegetationindex has the close relationship with the vegetation physiology andgrowth, so they have become the important indexes of agriculturaldrought monitoring. But only one index for drought monitoring willcause a problem which reduces the accuracy and usefulness of thedrought monitoring because short-term vegetation subjected to waterstress can maintain the original green time. So the methods which use twoindexes become an important direction of current drought monitoring.
In this paper, using MODIS land surface temperature data andsurface reflectivity data between May and September from2000to2008with GIS, RS and GPS technology to retrieved TVDI in the westernregion of Jilin Province, and analyze the relationship between TVDI with Ts, NDVI, surface soil moisture and meteorological data as well as itsspatial variability inter-month, inter-annual change. The conclusions areas follows:
1) The study area of Ts-NDVI feature space dry and wet edge arenot a straight line at all, and there are similar shapes of dry or wet edge indifferent feature space; the slope of dry and wet edge changes with theNDVI values.
2) There are some relationships between TVDI with Ts, NDVI,samples of soil moisture and precipitation; we can conclude thattemperature vegetation drought index is reasonable to indicate theevaluation of drought. From the correlation of TVDI and Ts, NDVI andthe samples of soil moisture showed that TVDI and Ts has a significantpositive correlation, TVDI and NDVI has a negatively correlation, TVDIand soil moisture has a significant negative correlation. TVDI with theprecipitation showed that the data change trend has a contrary trend withthe increase of TVDI values.
3) From month change of the western region of Jilin Province from2000to2008, we can find that there was little change in the wet areas ofnorthern of region; the drought was middle dry and severe dry in thewestern region of Jilin Province at May and June and it was easy to causethe spring drought, especially in Baicheng, Taonan, Tongyu. With thedevelopment, the middle dry and severe dry reduced at July and August, and the normal and light dry increased at the same time, the drought ofSongyuan, Qianguo and Changling eased particularly; The droughtconditions can be roughly divided into two parts when entered toSeptember. One part was the west of Da’an (include Da’an), the type ofdrought was still middle dry and severe dry, another part was the eastregion of Da’an, the drought dominated normal and light drought.
4) The inter-annual variability from2000to2008analysis showedthat, the region was mainly in light drought and middle drought. And thedrought changed with middle drought to severe drought from east to west,especially the southwest Tongyu had the most years of the severe drought.There were the biggest areas of the middle drought and the severedrought in2007, two areas were accounted for71.5percent of the region,and the drought is the most serious in this year. The areas of middledrought and severe drought were the least areas in2005, they wereaccounted for38.7percent, and this year is the lightest drought than otheryears.
本文综合运用GIS、RS和GPS技术，利用2000-2008年5-9月MODIS地表温度数据和地表反射率数据，建立Ts-NDVI特征空间，反演得到吉林省西部植被生长期的温度植被干旱指数（TVDI），分析了TVDI与Ts、NDVI、表层土壤湿度和气象数据的关系以及研究区2000-2008年月际和年际旱情时空分异。研究表明：
1）研究区Ts-NDVI特征空间的干、湿边并非呈一条直线，且不同时期特征空间的干、湿边具有相似的形状；干、湿边的斜率随着NDVI值的变化而发生变化；
2）TVDI与Ts、NDVI、采样点表层土壤湿度以及气象站点降水量具有一定关系，从关系中可以发现TVDI作为旱情评价指标有一定的合理性。TVDI与Ts、NDVI和采样点表层土壤湿度的相关性分析显示，TVDI与Ts呈显著正相关，与NDVI呈负相关,与土壤湿度呈显著负相关；从TVDI与气象站点降水量数据变化趋势来看，随着TVDI值的增大，气象站点降水量呈减小趋势；
3）2000-2008年的月际变化分析表明，研究区北部的湿润区总体变化不大；研究区各年的5-6月以中旱和重旱为主，易导致春旱发生，其中白城、洮南、通榆三地较为严重；各年的7-8月全区中旱和重旱面积有所减小，正常和轻旱面积增加，松原、前郭、长岭三地旱情减轻尤为明显，是旱情较轻的时间；到了9月，研究区干旱情况大体可分为两部分，大安以西的地区（包括大安）仍然会以中旱和重旱为主，而大安以东的地区则以正常和轻旱为主，部分地区受秋旱影响；
4）2000-2008年的年际变化分析表明，2000-2008年研究区全区主要以轻旱和中旱为主。研究区内由东向西分别呈现出由轻旱向重旱的过渡性变化，其中研究区西南部通榆地区出现重旱年份最多。2000-2008年年间，2007年的中旱和重旱面积比最大，高达71.5%，旱情最为严重；2005年中旱和重旱面积比最小，两者和为38.7%，是旱情最轻的一年。
Drought is one of major natural disasters that happens frequently inworld-wide in recent years, they not only disturbed social and economicsustainable development seriously, but also affected agriculture especially.The western region of Jilin Province where is the major agriculture andlivestock production in Jilin province is affected by drought frequently,because of the natural geographic and climate. Compare with thetraditional agricultural drought monitoring, remote sensing is more andmore applied in agricultural drought monitoring for its fast, efficient, lowcost and other characteristics. Land surface temperature and vegetationindex has the close relationship with the vegetation physiology andgrowth, so they have become the important indexes of agriculturaldrought monitoring. But only one index for drought monitoring willcause a problem which reduces the accuracy and usefulness of thedrought monitoring because short-term vegetation subjected to waterstress can maintain the original green time. So the methods which use twoindexes become an important direction of current drought monitoring.
In this paper, using MODIS land surface temperature data andsurface reflectivity data between May and September from2000to2008with GIS, RS and GPS technology to retrieved TVDI in the westernregion of Jilin Province, and analyze the relationship between TVDI with Ts, NDVI, surface soil moisture and meteorological data as well as itsspatial variability inter-month, inter-annual change. The conclusions areas follows:
1) The study area of Ts-NDVI feature space dry and wet edge arenot a straight line at all, and there are similar shapes of dry or wet edge indifferent feature space; the slope of dry and wet edge changes with theNDVI values.
2) There are some relationships between TVDI with Ts, NDVI,samples of soil moisture and precipitation; we can conclude thattemperature vegetation drought index is reasonable to indicate theevaluation of drought. From the correlation of TVDI and Ts, NDVI andthe samples of soil moisture showed that TVDI and Ts has a significantpositive correlation, TVDI and NDVI has a negatively correlation, TVDIand soil moisture has a significant negative correlation. TVDI with theprecipitation showed that the data change trend has a contrary trend withthe increase of TVDI values.
3) From month change of the western region of Jilin Province from2000to2008, we can find that there was little change in the wet areas ofnorthern of region; the drought was middle dry and severe dry in thewestern region of Jilin Province at May and June and it was easy to causethe spring drought, especially in Baicheng, Taonan, Tongyu. With thedevelopment, the middle dry and severe dry reduced at July and August, and the normal and light dry increased at the same time, the drought ofSongyuan, Qianguo and Changling eased particularly; The droughtconditions can be roughly divided into two parts when entered toSeptember. One part was the west of Da’an (include Da’an), the type ofdrought was still middle dry and severe dry, another part was the eastregion of Da’an, the drought dominated normal and light drought.
4) The inter-annual variability from2000to2008analysis showedthat, the region was mainly in light drought and middle drought. And thedrought changed with middle drought to severe drought from east to west,especially the southwest Tongyu had the most years of the severe drought.There were the biggest areas of the middle drought and the severedrought in2007, two areas were accounted for71.5percent of the region,and the drought is the most serious in this year. The areas of middledrought and severe drought were the least areas in2005, they wereaccounted for38.7percent, and this year is the lightest drought than otheryears.
引文
[1]刘良明.基于EOS-MODIS数据的遥感干旱预警模型研究[D].武汉大学博士学位论文，2004.
[2]李恩泽.吉林省西部地区农业生态地理环境研究报告[R].长春：吉林省科技厅，2001:15—40．
[3]刘桂环,董锁成,彭唬.吉林省西部生态环境脆弱区农业发展模式探讨[J].中国生态农业学报，2006，14(1):215-219．
[4]任春颖,张柏,王宗明等.吉林省西部半干旱地区土地利用方式对土壤养分的影响——以吉林省通榆县为例[J].干旱地区农业研究，2010，28(4).
[5]王黎明.吉林省西部区域遥感蒸散模型研究及其应用[D].吉林大学博士学位论文，2005.
[6]胡红武,胡梅,龙玲,等.区域干旱遥感监测研究综述[J].安徽农业科学，2008，36(33):14817-14819.
[7]张红卫,陈怀亮,申双和,等. NDVI-ST特征空间及干湿边变化特征[J].气象科技，2010，38(1):86-95.
[8]徐霞.基于TVDI指数的农业干旱遥感监测应用研究[D].新疆大学，2009年.
[9] Waston K, Rowen L C, Offield T W. Application of thermal modeling in thegeologic interpretation of IR images [J].Remote Sensing ofEnvironment,1971,3:2017-2014.
[10] Waston K, Pohn H A. Thermal inertia mapping from satellites discrimination ofgeologic units in Oman [J].Journal of Research GeologySurving,1974,2(2):147-158.
[11]闫娜,杜继稳,李登科,延军平.干旱遥感监测方法研究应用进展[J].灾害学，2008，23(4):117-121.
[12] PRICE J C. Using spatial context in satellite data to infer regional scaleevapotransportation [J].IEEE Transactions on Geoscience and RemoteSensing,1990,28:940-948.
[13]韩丽娟,王鹏新,王锦地等.植被指数-地表温度构成的特征空间研究[J].中国科学D辑，2005，35(4):371-377.
[14]韩晓勇.基于MODIS和地物光谱的植被水分定量遥感研究——以大兴安岭林区为例[D].长安大学,2009.
[15]闫娜娜.基于遥感指数的旱情监测方法研究[D].中国科学院遥感应用研究所,2005.
[16] Carlson T N, Gillies R R, Perry E M. A method to make use of thermal infraredtemperature and NDVI measurement s to infer surface soil water content andfractional vegetation cover [J].Remote Sensing Reviews,1994,9:161-173.
[17]覃志豪,高懋芳,秦小敏,等.农业旱灾监测中的地表温度遥感反演方法——以MODIS数据为例[J].自然灾害学报，2005，14(4):64-71.
[18] Nemain R R, Running S W. Estimation of regional surface resis-tance toevapotranspiration from NDVI and t hermal-IR AVHRR data [J].Journal ofApplied Meteorology,1989,28:276-284.
[19] Moron M S, Clarke T R, Inoue Y, et al. Estimating crop water deficit using therelation between surface air temperature and spectral vegetation index[J].Remote Sensing of Environment,1994,49:246-263.
[20]高磊,覃志豪,卢丽萍.基于植被指数和地表温度特征空间的农业干旱监测模型研究综述[J].国土资源遥感，2007，73(3):1–7.
[21] Goetz S J. Muti2sensor analysis of NDVI surface temperature and biophysicalvariables at a mixed grassland site [J].International Journal of RemoteSensing,1997,18(1):71-94.
[22] Lambin E F, Ehrlich D. The surface temperature-vegetation index for land coverand land cover change analysis [J].International Journal of RemoteSensing,1996,17:463-487．
[23] SANDHOL T I, RASMUSSEN K, ANDERSEN J. A simple interpretation of thesurface temperature/vegetation index space for assessment of surface moisturestatus [J].Remote Sensing of Environment,2002,79:213-224.
[24]卓义.基于MODIS数据的蒙古高原荒漠化遥感定量监测方法研究[D].内蒙古师范大学,2007.
[25] Koike T. Study on spatial and temporal variability of surface soil wetness onTibetan Plateau by using the Satellite-based microwave radiometer [J].Annual JHydrualic Eng,1997,41:915-919.
[26]刘茜.基于MODIS数据的黑龙江省农业干旱遥感监测研究[D].南京信息工程大学,2011.
[27]康为民,罗宇翔,郑小波等.贵州温度植被干旱的指数（TVDI）特征及其遥感干旱的监测应用[J].贵州农业科学，2008，36(4):27-30.
[28]张顺谦.基于温度植被干旱指数的四川伏旱遥感监测与影响评估[J].农业工程学报,2007，23(9):141-146.
[29]盛绍学,胡雯,马晓群等.安徽省农业干早遥感监测指标的确定及应用[J].中国农业气象，2001，22(4):36-39.
[30]王鹏新,龚健雅,李小文.条件植被温度指数及其在干旱监测中的应用[J].武汉大学学报，2001，26(5):412-418.
[31]齐述华.干旱监测遥感模型和中国干旱时空分析[D].北京：中国科学院遥感应用研究所，2004.
[32]陈鹏.基于植被指数和地表温度的农业旱情监测适用性研究[D].南京信息工程大学,2011.
[33]姚春生,张增祥,汪潇.使用温度植被干旱指数法(TVDI)反演新疆土壤湿度[J].遥感技术与应用，2004，19(6):473-478.
[34]杨曦,武建军,闫峰等.基于地表温度-植被指数特征空间的区域土壤干湿状况[J].生态学报，2009，29(3):1205-1216.
[35]李红军,郑力,雷玉平等.植被指数-地表温度特征空间研究及其在旱情监测中的应用[J].农业工程学报，2006，22(11):170-174.
[36] Christelle Vancutsem, Pietro Ceccato, Tufa Dinku, et al. Evaluation of MODISland surface temperature data to estimate air temperature in differentecosystems over Africa [J].Remote Sensing ofEnvironment,2010,(114):449-465.
[37] Ronglin Tang, Zhao-Liang Li, Bohui Tang. An application of the Ts-VI trianglemethod with enhanced edges determination for evapotranspiration estimationfrom MODIS data in arid and semi-arid regions:Implementation and validation[J].Remote Sensing of Environment,2010,(114):540-551.
[38] Inge Sandholt, Kjeld Rasmussen, Jens Andersen. A simple interpretation of thesurface temperature/vegetation index space for assessment of surface moisturestatus [J].Remote Sensing of Environment,2002,(79):213-224.
[39] Xianwei Wang, Hongjie Xie, Huade Guan, et al. Different responses ofMODIS-derived NDVI to root-zone soil moisture in semi-arid and humidregions [J].Journal of Hydrology,2007,(340):12-24.
[40]任春颖.吉林省西部半干旱地区土地利用方式对土壤养分的影响——以吉林省通榆县为例[J].干旱地区农业研究,2010,28(4):206-212.
[41]王惠清,吴忠义等编著.吉林省气候[M].北京：气象出版社，1997．
[42]王季平等编.吉林年鉴[M].长春：吉林年鉴社，2003
[43]吉林省地方志编纂委员会.吉林省志.第一卷总述[M].长春：吉林人民出版社，2004.
[44]王雅强.吉林省西部夏季降水与暴雨时空分布规律研究[D].东北师范大学硕士论文，2006.
[45]王丽涛,周艺,乔德军等.我国北方森林草原地区火险等级遥感评估及时空特征分析[J].遥感信息,2011,3:44-49
[46]赵英时等.遥感应用分析原理与方法[M].北京：科学出版社,2003.
[47]郭鹏.基于MODIS的青藏高原地表参数时空变化研究[D].中国石油大学(华东),2010.
[48]麻金继.基于MODIS图像海岸带气溶胶光学特性反演模型研究[D].中国科学院合肥物质科学研究院,2007.
[49]白天路.基于遥感和地面实测水分数据的小流域土壤水分模拟[J].西北大学，2010.
[50]毕力格.基于MODIS数据的内蒙古干旱监测[D].内蒙古师范大学,2009.
[51] Qin Z, Dall`Olmo G, Karnieli A, et al. Derivation of split window algorithm andit s sensitivity analysis for retrieving land surface temperature from NOAA2AVHRR data [J].Journal of Geophysical Research,2001,106(D19):22655.
[52]韩晓勇.《基于MODIS和地物光谱的植被水分定量遥感研究——以大兴安岭林区为例》[D].长安大学，2009.
[53]韩岭.地震灾害遥感特征分析与信息提取技术研究[D].成都理工大学，2011.
[54]姚春生.使用MODIS数据反演土壤水分研究[D].中国科学院研究生院硕士学位论文，2003.
[55]李喆,谭德宝,秦其明.基于特征空间的遥感干旱监测方法综述[J].长江科学院院报，2010，27(1):37-40.
[56]赵莉荣,武伟,刘洪斌等.基于温度植被干旱指数法的农业干旱研究[J].西南师范大学学报(自然科学版)，2009，34(2):80-84.
[57] Christelle Vancutsem, Pietro Ceccato, Tufa Dinku, et al. Evaluation of MODISland surface temperature data to estimate air temperature in differentecosystems over Africa [J].Remote Sensing ofEnvironment,2010,(114):449-465.
[58] Xianwei Wang, Hongjie Xie, Huade Guan, et al. Different responses ofMODIS-derived NDVI to root-zone soil moisture in semi-arid and humidregions [J].Journal of Hydrology,2007,(340):12-24.
[59]齐述华,王长耀,牛铮.利用温度植被旱情指数(TVDI)进行全国旱情监测研究[J].遥感学报，第7卷第5期，2003，420-427.
[2]李恩泽.吉林省西部地区农业生态地理环境研究报告[R].长春：吉林省科技厅，2001:15—40．
[3]刘桂环,董锁成,彭唬.吉林省西部生态环境脆弱区农业发展模式探讨[J].中国生态农业学报，2006，14(1):215-219．
[4]任春颖,张柏,王宗明等.吉林省西部半干旱地区土地利用方式对土壤养分的影响——以吉林省通榆县为例[J].干旱地区农业研究，2010，28(4).
[5]王黎明.吉林省西部区域遥感蒸散模型研究及其应用[D].吉林大学博士学位论文，2005.
[6]胡红武,胡梅,龙玲,等.区域干旱遥感监测研究综述[J].安徽农业科学，2008，36(33):14817-14819.
[7]张红卫,陈怀亮,申双和,等. NDVI-ST特征空间及干湿边变化特征[J].气象科技，2010，38(1):86-95.
[8]徐霞.基于TVDI指数的农业干旱遥感监测应用研究[D].新疆大学，2009年.
[9] Waston K, Rowen L C, Offield T W. Application of thermal modeling in thegeologic interpretation of IR images [J].Remote Sensing ofEnvironment,1971,3:2017-2014.
[10] Waston K, Pohn H A. Thermal inertia mapping from satellites discrimination ofgeologic units in Oman [J].Journal of Research GeologySurving,1974,2(2):147-158.
[11]闫娜,杜继稳,李登科,延军平.干旱遥感监测方法研究应用进展[J].灾害学，2008，23(4):117-121.
[12] PRICE J C. Using spatial context in satellite data to infer regional scaleevapotransportation [J].IEEE Transactions on Geoscience and RemoteSensing,1990,28:940-948.
[13]韩丽娟,王鹏新,王锦地等.植被指数-地表温度构成的特征空间研究[J].中国科学D辑，2005，35(4):371-377.
[14]韩晓勇.基于MODIS和地物光谱的植被水分定量遥感研究——以大兴安岭林区为例[D].长安大学,2009.
[15]闫娜娜.基于遥感指数的旱情监测方法研究[D].中国科学院遥感应用研究所,2005.
[16] Carlson T N, Gillies R R, Perry E M. A method to make use of thermal infraredtemperature and NDVI measurement s to infer surface soil water content andfractional vegetation cover [J].Remote Sensing Reviews,1994,9:161-173.
[17]覃志豪,高懋芳,秦小敏,等.农业旱灾监测中的地表温度遥感反演方法——以MODIS数据为例[J].自然灾害学报，2005，14(4):64-71.
[18] Nemain R R, Running S W. Estimation of regional surface resis-tance toevapotranspiration from NDVI and t hermal-IR AVHRR data [J].Journal ofApplied Meteorology,1989,28:276-284.
[19] Moron M S, Clarke T R, Inoue Y, et al. Estimating crop water deficit using therelation between surface air temperature and spectral vegetation index[J].Remote Sensing of Environment,1994,49:246-263.
[20]高磊,覃志豪,卢丽萍.基于植被指数和地表温度特征空间的农业干旱监测模型研究综述[J].国土资源遥感，2007，73(3):1–7.
[21] Goetz S J. Muti2sensor analysis of NDVI surface temperature and biophysicalvariables at a mixed grassland site [J].International Journal of RemoteSensing,1997,18(1):71-94.
[22] Lambin E F, Ehrlich D. The surface temperature-vegetation index for land coverand land cover change analysis [J].International Journal of RemoteSensing,1996,17:463-487．
[23] SANDHOL T I, RASMUSSEN K, ANDERSEN J. A simple interpretation of thesurface temperature/vegetation index space for assessment of surface moisturestatus [J].Remote Sensing of Environment,2002,79:213-224.
[24]卓义.基于MODIS数据的蒙古高原荒漠化遥感定量监测方法研究[D].内蒙古师范大学,2007.
[25] Koike T. Study on spatial and temporal variability of surface soil wetness onTibetan Plateau by using the Satellite-based microwave radiometer [J].Annual JHydrualic Eng,1997,41:915-919.
[26]刘茜.基于MODIS数据的黑龙江省农业干旱遥感监测研究[D].南京信息工程大学,2011.
[27]康为民,罗宇翔,郑小波等.贵州温度植被干旱的指数（TVDI）特征及其遥感干旱的监测应用[J].贵州农业科学，2008，36(4):27-30.
[28]张顺谦.基于温度植被干旱指数的四川伏旱遥感监测与影响评估[J].农业工程学报,2007，23(9):141-146.
[29]盛绍学,胡雯,马晓群等.安徽省农业干早遥感监测指标的确定及应用[J].中国农业气象，2001，22(4):36-39.
[30]王鹏新,龚健雅,李小文.条件植被温度指数及其在干旱监测中的应用[J].武汉大学学报，2001，26(5):412-418.
[31]齐述华.干旱监测遥感模型和中国干旱时空分析[D].北京：中国科学院遥感应用研究所，2004.
[32]陈鹏.基于植被指数和地表温度的农业旱情监测适用性研究[D].南京信息工程大学,2011.
[33]姚春生,张增祥,汪潇.使用温度植被干旱指数法(TVDI)反演新疆土壤湿度[J].遥感技术与应用，2004，19(6):473-478.
[34]杨曦,武建军,闫峰等.基于地表温度-植被指数特征空间的区域土壤干湿状况[J].生态学报，2009，29(3):1205-1216.
[35]李红军,郑力,雷玉平等.植被指数-地表温度特征空间研究及其在旱情监测中的应用[J].农业工程学报，2006，22(11):170-174.
[36] Christelle Vancutsem, Pietro Ceccato, Tufa Dinku, et al. Evaluation of MODISland surface temperature data to estimate air temperature in differentecosystems over Africa [J].Remote Sensing ofEnvironment,2010,(114):449-465.
[37] Ronglin Tang, Zhao-Liang Li, Bohui Tang. An application of the Ts-VI trianglemethod with enhanced edges determination for evapotranspiration estimationfrom MODIS data in arid and semi-arid regions:Implementation and validation[J].Remote Sensing of Environment,2010,(114):540-551.
[38] Inge Sandholt, Kjeld Rasmussen, Jens Andersen. A simple interpretation of thesurface temperature/vegetation index space for assessment of surface moisturestatus [J].Remote Sensing of Environment,2002,(79):213-224.
[39] Xianwei Wang, Hongjie Xie, Huade Guan, et al. Different responses ofMODIS-derived NDVI to root-zone soil moisture in semi-arid and humidregions [J].Journal of Hydrology,2007,(340):12-24.
[40]任春颖.吉林省西部半干旱地区土地利用方式对土壤养分的影响——以吉林省通榆县为例[J].干旱地区农业研究,2010,28(4):206-212.
[41]王惠清,吴忠义等编著.吉林省气候[M].北京：气象出版社，1997．
[42]王季平等编.吉林年鉴[M].长春：吉林年鉴社，2003
[43]吉林省地方志编纂委员会.吉林省志.第一卷总述[M].长春：吉林人民出版社，2004.
[44]王雅强.吉林省西部夏季降水与暴雨时空分布规律研究[D].东北师范大学硕士论文，2006.
[45]王丽涛,周艺,乔德军等.我国北方森林草原地区火险等级遥感评估及时空特征分析[J].遥感信息,2011,3:44-49
[46]赵英时等.遥感应用分析原理与方法[M].北京：科学出版社,2003.
[47]郭鹏.基于MODIS的青藏高原地表参数时空变化研究[D].中国石油大学(华东),2010.
[48]麻金继.基于MODIS图像海岸带气溶胶光学特性反演模型研究[D].中国科学院合肥物质科学研究院,2007.
[49]白天路.基于遥感和地面实测水分数据的小流域土壤水分模拟[J].西北大学，2010.
[50]毕力格.基于MODIS数据的内蒙古干旱监测[D].内蒙古师范大学,2009.
[51] Qin Z, Dall`Olmo G, Karnieli A, et al. Derivation of split window algorithm andit s sensitivity analysis for retrieving land surface temperature from NOAA2AVHRR data [J].Journal of Geophysical Research,2001,106(D19):22655.
[52]韩晓勇.《基于MODIS和地物光谱的植被水分定量遥感研究——以大兴安岭林区为例》[D].长安大学，2009.
[53]韩岭.地震灾害遥感特征分析与信息提取技术研究[D].成都理工大学，2011.
[54]姚春生.使用MODIS数据反演土壤水分研究[D].中国科学院研究生院硕士学位论文，2003.
[55]李喆,谭德宝,秦其明.基于特征空间的遥感干旱监测方法综述[J].长江科学院院报，2010，27(1):37-40.
[56]赵莉荣,武伟,刘洪斌等.基于温度植被干旱指数法的农业干旱研究[J].西南师范大学学报(自然科学版)，2009，34(2):80-84.
[57] Christelle Vancutsem, Pietro Ceccato, Tufa Dinku, et al. Evaluation of MODISland surface temperature data to estimate air temperature in differentecosystems over Africa [J].Remote Sensing ofEnvironment,2010,(114):449-465.
[58] Xianwei Wang, Hongjie Xie, Huade Guan, et al. Different responses ofMODIS-derived NDVI to root-zone soil moisture in semi-arid and humidregions [J].Journal of Hydrology,2007,(340):12-24.
[59]齐述华,王长耀,牛铮.利用温度植被旱情指数(TVDI)进行全国旱情监测研究[J].遥感学报，第7卷第5期，2003，420-427.