增刊1 冯海霞等：基于HJ-1A/1B CCD数据的干旱监测
363
为进一步分析PDI和MPDI对干旱变化的敏感情况，
MPDI分布图上随机在2010-03-12与2010-01-03的PDI、
选取了20个样点，样点的PDI、MPDI差值如图8。
的PDI、MPDI和NDVI时间序列曲线变化，仍可正确地监测研究区实际的旱情变化和农作物的长势情况。
4 结论与讨论
1）利用HJ-1A/ABCCD数据和PDI、MPDI等干旱
指数，可以快速实现对喀斯特脆弱生态环境的旱情监测。
2）MPDI对干旱变化的响应比PDI敏感，MPDI在监测植被覆盖较好的西南喀斯特地区的作物长势和干旱情况时比PDI更为有效。
3）降水量是影响监测效果的主要因素，它对MPDI的影响比PDI大，结合研究区的PDI、MPDI和NDVI时间序列曲线变化，仍可正确监测研究区实际的旱情变化和植被长势情况。
4）HJ-1A/AB 数据兼具MODIS数据重访周期短和TM数据分辨率高的特点、且具有覆盖范围广、数据免费等优势，是农情遥感监测应用研究中重要的数据源，在干旱监测与评估方面具有广阔的应用前景。
遥感反映的浅层地表的干旱情况，对深层土壤水分的变化监测较弱，且受天气状况的影响较大，比如小雨会造成PDI、MPDI值得急剧下降，但对深层土壤水分影响较小。微波因具有较强的穿透力，成为遥感干旱监测的重要发展的方向。
[参 考 文 献]
[1] 阿布都瓦斯提·吾拉木. 基于n维光谱特征空间的农田干旱遥感监测[D]. 北京大学，2006，5－138.
Abduwasit Ghulam. Remote Monitoring of Farmland Drought Based on n-Dimentional Spectral Feature Space[D], doctoral dissertation, Peking University, 2006, 5－138. (in Chinese with English abstract)
[2] Jackson R D, Idso S B. Canopy temperature as a crop water
stress indicator[J]. Water Resources Research, 1981, 17: 133－138.
[3] Jackson R D, Slaler P N, Pinter P J. Discrimination of growth
and water stress in wheat by various Vegetation indices through clear and turbidatmosphere[J]. Remote Sensing of Environment, 1983, 3: 187－208. [4] 陈维英，肖乾广，盛永伟. 距平植被指数在1992年特大干旱监测中的应用[J]. 环境遥感，1994，9(2)：106－112.
Chen Weiying, Xiao Qianguang, Sheng Yongwei. Application of the anomaly vegetation index to monitoring heavy drought in 1992[J]. Remote Sensing of Environment, 1994, 9(2): 106－112.
[5] Kogan F N. Remote sensing of weather impacts on vegetation in
nonhomogeneous area[J]. International Journal of Remote sensing, 1990, 11: 1405－1420.
[6] Peters A J, Walter-Shea E A, Lei J, et al. Drought monitoring
with NDVI-based standardized vegetation index[J]. Photogrammetric Engineering and Remote Sensing, 2002, 65 (1): 71－75.
[7] Kogan F N. Application of vegetation index and brightness
temperature for drought detection[J]. Advances in Space Research, 1995, 15: 91－100.
图8 样点的PDI、MPDI差值对比图 Fig.8 difference value of PDI、MPDI
从图8可见，MPDI因为考虑了植被盖度的影响，相同的地点相同时间的MPDI比PDI小，且MPDI的差值较PDI的差值大；说明相同的干旱变化，MPDI的变化幅度比PDI大，即MPDI对干旱的变化情况比PDI敏感。
研究区2010年的大旱使植被的生长受到了严重影响，植被枯萎、甚至死亡，植被盖度急剧下降，造成研究区2010年的MPDI值比2009年同期的MPDI值高，且逼近PDI的值，MPDI的值越逼近PDI值时，表明旱情越严重。
经以上分析，MPDI对干旱的变化情况比PDI敏感，在监测植被覆盖区的干旱情况时比PDI更为有效。 3.3 长时间干旱监测中PDI、MPDI的影响因素分析
影响PDI、MPDI监测精度的因素较多，如土壤基线的确定、植被覆盖度的计算、气候条件的变化等，但对固定的研究区，土壤线基本稳定，植被覆盖度的突变也不大，影响其监测效果的主要因素是降水。降水量较小时，对缓解旱情作用较小，却能引起PDI、MPDI值的急剧降低，影响监测效果。
从PDI、MPDI的时间序列变化曲线图6，除了2010-04-09的4号点外，变化幅度较大的另一个点是2009年的第4个点（2009-04-14），根据气象资料，2009-04-12安顺地区有24.3 mm的降水；2009-04-14研究区的PDI的值比2009-03-14降低0.011，而MPDI降低0.027；2010-04-08研究区有9.3 mm的降水，2010-04-09研究区的PDI值比2010-03-12降低0.073，而MPDI降低0.075，即相同的降水条件下，MPDI的变化幅度大于PDI。这与光学传感器和植被的特性有关，光学遥感的有效数据是晴空时的，较降水时间有一定的滞后性，降水后，土壤蒸发较快，植被因为截流、自身吸收等原因，使水分保持时间较长久，因此MPDI对降水的响应比PDI敏感。
结合3.1基于PDI、MPDI和NDVI旱情监测分析可知，较小的降水量虽然影响监测的效果，但结合研究区
364
农业工程学报 2011年
image[J]. Spacecraft Engineering, 2009, 18(6): 133－137. (in Chinese with English abstract)
[17] 陈雪洋，蒙继华，吴炳方，等. 基于HJ-1 CCD的夏玉米
FPAR遥感监测模型[J]. 农业工程学报，2010，26（增刊1）：241－245.
Chen Xueyang, Meng Jihua, Wu Bingfang, et al. Monitoring corn FPAR based on HJ-1 CCD[J]. Transactions of the CSAE, 2010, 26(Supp.1): 241－245. (in Chinese with English abstract)
[18] 张竞成，顾晓鹤，王纪华，等. 基于HJ-CCD与TM影像
的水稻LAI估测一致性分析[J]. 农业工程学报，2010，26(7)：186－194.
Zhang Jingcheng, Gu Xiaohe, Wang Jihua, et al. Analysis of consistency between HJ-CCD images and TM images in monitoring rice LAI[J]. Transactions of the CSAE, 2010, 26(7): 186－194. (in Chinese with English abstract) [19] 张学艺，李剑萍，秦其明，等. 几种干旱监测模型在宁夏的对比应用[J]. 农业工程学报，2009，25(8)：18－24. Zhang Xueyi, Li Jianping, Qin Qiming, et al. Comparison and application of several drought monitoring models in Ningxia, China[J]. Transactions of the CSAE, 2009, 25(8): 18－24. (in Chinese with English abstract)
[20] Vermote E F, Tanre D, Deuze J L, et al. The Second
Simulation of the Satellite Signal in the Solar Spectrum (6S) User’s Guide[J]. France: Laboratoire d’Optique Atmospherique, 1997.
[21] 杜鑫，陈雪洋，蒙继华，等. 基于6S模型的环境星CCD数据大气校正[J]. 国土资源遥感，2010，2：22－26.
Du Xin, Chen Xueyang, Meng Jihua1, et al. Atmospheric correction of HJ-1CCD data based on 6S model[J]. Remote Sensin for Land and Resource, 2010, 2: 22－26. (in Chinese with English abstract)
[22] 阿布都瓦斯提·吾拉木，秦其明，朱黎江. 基于6S模型
的可见光、近红外波段数据的大气校正[J]. 北京大学学报：自然科学版，2004，40(4)：611－618.
Abduwasit Ghulam, Qin Qiming, Zhu Lijiang. 6S model based atmospheric correction of visible and near-infrared data and sensitivity analys[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2004, 40(4): 611－618. (in Chinese with English abstract)
[23] Abduwasit. Ghulam, Qin Qiming, Zhu Lin, et al. Satellite
remote sensing of groundwater: quantitative modelling and uncertainty reduction using 6S atmospheric simulations[J]. International Journal of Remote Sensing, 2004, 25(23): 5509－5524.
[24] Calson et al. Satellite remote sensing of land use changes in
and around San Jose, Costa Rica[J]. Remote Sensing of Environment, 1999, 70: 247－256. [25] 徐希孺. 遥感物理[M]. 北京：北京大学出版社，2005，26
－120.
[26] 赵英时. 遥感应用分析原理与方法[M]. 北京：科学出版社.
2003，36－39.
[8] 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.
Han Lijuan, Wang Pengxin, Wang Jindi, et al. The study of NDVI-Ts Feature Space[J]. Science in China(Series D: Earth Sciences), 2005, 35(4): 371－377. (in Chinese with English abstract) [9] 詹志明，秦其明，阿布都瓦斯提·吾拉木，等. 基于NIR-Red
光谱特征空间的土壤水分监测新方法术[J]. 中国科学（D辑），2006，36(11)：1020－1026.
Zhan Zhiming, Qin Qiming, Abduwasit Ghulam, et al. A New method of soil moisture monitoring based nir-red spectral reflectance space[J]. Science in China(Series D: Earth Sciences), 2006, 36(11): 1020－1026. (in Chinese with English abstract) [10] Abduwasit Ghulam, Qin, Qiming, Zhan Zhiming. Modified
perpendicular drought index (MPDI): a real-time drought monitoring method[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2007, 62: 150－164. [11] 李素菊. 环境与灾害监测预报小卫星星座[J]. 中国减灾，2004，(2)：19.
Li Suju. The environment and disaster monitoring and forecasting smallsatellite constellation[J]. Disaster Reduction in China, 2004(2):1 9. (in Chinese with English abstract) [12] 李传荣，唐伶俐，胡坚，等. HJ-1光学卫星应用潜力[J]. 科
技导报，2008，26(13)：56－59.
Li Chuanrong, Tang Lingli, Hu Jian, et al. Potential applications of HJ-1 optical satellites[J]. Science and Technology Review, 2008, 26(13): 56－59. (in Chinese with English abstract) [13] 代守仑，王肇宇，白照广，等. HJ-1A、1B卫星研制综述[J]. 航天器工程，2009，18(6)：12－16.
Dai Shoulun, Wang Zhaoyu, Bai Zhaoguang, et al. A summary of HJ-lA/1B satellite development[J]. Spacecraft Engineering, 2009, 18(6): 12－16. (in Chinese with English abstract)
[14] 孙林，柳钦火，陈良富，等. 环境与减灾小卫星高光谱成
像仪陆地气溶胶光学厚度反演[J]. 遥感学报，2006，10(5)：770－776.
Sun Lin, Liu Qinhuo, Chen Lisngfu, et al. The application of HJ-1 hyperspectral haging radian eter to retrieve aerosol optical thickness over land[J]. Journal of Remote Sensing, 2006, 10(5): 770－776. (in Chinese with English abstract) [15] 陈世荣，孙颧，张宝军. 环境减灾-1lA、1B卫星在干旱监
测中的应用研究及实现[J]. 航天器工程，2009，18(6)：138－141.
Chen Shirong, Sun Hao, Zhang Baojun. Application research and implementation of drought monitoring by HJ-1A/AB satellites[J]. Spacecraft Engineering, 2009, 18(6): 138－141. (in Chinese with English abstract)
[16] 王桥，杨煌，吴传庆. 环境减灾-1A卫星超光谱数据反演
叶绿素a浓度的模型研究[J]. 航天器工程，2009，18(6)：133－147.
Wang Qiao, Yang Yi, Wu Chuanqing. Study of retrieving models for chlorophyll-a concentration based on HJ-1A HIS
增刊1 冯海霞等：基于HJ-1A/1B CCD数据的干旱监测
365
Drought monitoring based on HJ-1A/1B CCD data
※
Feng Haixia1,2, Qin Qiming1, Jiang Hongbo1, Dong Heng1, Zhang Ning1,Wang Jinliang1, Liu Mingchao1
(1. Institute of Remote Sensing and GIS, Peking University, Beijing 100871, China; 2. Department of Civil Engineering, Shandong Jiaotong University, Jinan 250023, China)
Abstract: HJ-1A/1B is the first small satellite constellation in China for environment and disaster monitoring and forecasting. This paper is aimed to discuss the potential application of HJ-1A/1B CCD data in drought monitoring in southwestern Karst area in China. The Anshun area in Guizhou province which suffered greatly from drought disaster in 2010 was selected as study area. The Normalized Difference Vegetation Index (NDVI), Perpendicular Drought Index (PDI) and Modified Perpendicular Drought Index (MPDI) were used in a time-series analysis based on multi-temporal HJ-1A/1B data to monitor the crop growth and drought status. The characteristic, suitability and influence factors of each model were also discussed. The results showed that PDI, MPDI performed well by utilizing HJ-1A/1B data in drought monitoring, in which MPDI presented a better performance than PDI, especially in densely vegetated area. Precipitation is a major influencing factor in drought monitoring, and it has more impact on MPDI than PDI. The combination of drought monitoring index and vegetation index (NDVI) in a time-series analysis could give a better description of crop growth and drought condition. This work is hoped to facilitate application of HJ-1A/1B data in crop and drought monitoring in western Karst area of China and contribute to the decision-making of Chinese government when confronting drought disasters.
Key words: drought, precipitation monitoring, HJ-1, PDI, MPDI, time-series