基于NDVI的宁夏盐池县的植被盖度动态变化及其影响因素的研究
摘要
盐池县位于宁夏回族自治区东部,介于北纬37°04′～38°10′,东经106°30′～107°47′之间，北部与毛乌素沙漠相连，南部与黄土高原相靠，地处干旱和半干旱过渡地带，在地理位置上是从黄土高原向鄂尔多斯台地的过渡地带、宁夏盐池县在植被上是从干草原向荒漠的过渡地带，在资源利用上是从农区向牧区过渡地带，处于属典型的农牧交错区，是宁夏唯一的牧业县。在自然和人为的共同作用下，该区以草原退化问题严重威胁着当地经济的发展和生态安全。自2002年底盐池县全境实施禁牧封育，为了更好地对该区草原植被状况发展进行实时动态监测，了解草原植被变化趋势，依据国家标准(GB19377-2003)及遥感数据，从盐池县草原植被覆盖情况出发，选择草原植被覆盖状况作为评价指标，利用植被指数（NDVI）及象元二分法（亚像元分解法），建立盐池县草原植被覆盖遥感监测体系，对该区域草原覆盖状况变化遥感监测和评价，主要取得以下成果：
(1)本文从盐池县为例，根据其草原植被覆盖实际情况出发，依据遥感数据和国家标准(GB19377-2003)为基础，以盐池县草原植被覆盖度为评价的指标，结合遥感植被指数（NDVI）及亚像元分解法（象元二分法），对盐池县草原植被覆盖度进行了遥感估测计算，并依据野外调查数据进行验证，进行盐池县草原覆盖度连续11年的状况变化遥感监测及评价，表明此模型进行植被覆盖度遥感估测是有效切实可行的。
(2)2000年到2010年间，11年间盐池县草原植被覆盖明显好转，极低覆盖植被草原减少了1462.39km~2；中覆盖草原、中高覆盖草原、高覆盖草原增加了586.39km~2，说明虽然整体趋于好转。但状况不甚乐观，低和极低覆盖草原都向上级别转移，说明该县荒漠化在此期间发生逆转，但低覆盖草原部分保留面积较大，说明在这一期草原整体向良性方向发展，但速度缓慢。在人为因素和自然因素的共同作用下，持续地加强草原生态建设方面的投入力度，草原整体状况会有质的变化。应用此方法对宁夏盐池县植被覆盖度取得了较高精度。
(3)在自然及人为因素下，不同草原植被覆盖的面积和等级都在不断演变，而在该县运用马尔科夫链，在2000年、2005年及2009年遥感数据的辅助下，能客观的反应该县的整体草原覆盖状况，因此马儿科夫链能有效应用于草原覆盖的动态变化评价，结果可靠。
(4)利用GIMMS和MOD13Q1数据集，对盐池县草原植被覆盖时空变化特征分析的研究表明： GIMMS和MOD13Q1数据集，可用于一定尺度草原植被覆盖变化的研究，且效果良好，也是对此数据集在草原资源监测应用中的大胆探索与信息挖掘。
Yanchi County at the eastern part of the Ningxia Hui Autonomous Region, betweenlatitude37°04'~38°10', longitude106°30'~107°47' between the northern part of theMu Us desert connected to the south and the Loess Plateau, rely on, is located in aridandsemi-arid transition zone on the geographical location of the transition zone from the LoessPlateau to the Ordos tableland Yanchi County in vegetation from steppe to desert transitionzone in the transition zone from rural areas to the pastoral areas of resource utilization,in atypical pastoral area, is the only animal husbandry county in Ningxia. Grassland degradationin the area under the combined action of natural and man-made, a serious threat to localeconomic development and ecological security. Yanchi County, since the end of2002,implementation of the grazing prohibition, in order to better the area of the development ofthe grassland vegetation status in real-time dynamic monitoring to understand the trends ofgrassland vegetation, according to the national standard (GB19377-2003) and remote sensingdata from Yanchi County grassland vegetation coverage, select the grassland vegetation coveras the evaluation index, vegetation index (NDVI) and the dichotomy pixel (sub-pixeldecomposition), the establishment of Yanchi County grassland vegetation cover remotesensing monitoring system, the region grassland cover conditions changes in remote sensingmonitoring and evaluation, the following results:
(1) Yanchi County proceed according to the grassland vegetation cover the actualsituation, based on remote sensing data and the national standard (GB19377-2003), based onYanchi County grassland vegetation coverage for the evaluation of indicators, combined withremote sensing vegetation index (NDVI) and sub-pixel decomposition method (like thedichotomy), Yanchi County grassland vegetation coverage of remote sensing estimationcalculation and validation based on field survey data, Yanchi County grassland coverage for11consecutive years, the situation changes remote sensingmonitoring and evaluation,indicating that this model of remote sensing estimation of vegetation coverage is effectivepractical.
(2)2000to2010,11years Yanchi County grassland vegetation cover significantlyimproved, very low vegetation cover grassland reduced1462.39km~2; in coverage grassland,high coverage grassland, high coverage grassland increased by586.39km~2, indicating that although as a whole tends to improved. But the situation is not very optimistic, low and verylow coverage grassland up level transfer of the county desertification be reversed during thisperiod, but low coverage grassland part to retain the larger, indicating that in this periodgrassland as a whole to the positive direction, but slow. The common role of the human andnatural factors, continue to enhance the grassland ecological construction investment,grassland overall situation will be a qualitative change. The application of this methodachieved a high accuracy vegetation coverage in Yanchi County.
(3) under the natural and human factors, the size and grade of the different grasslandvegetation cover in the evolving use of Markov chain in the county,2000,2005and2009remote sensing data of the auxiliary objective the reactions of the county's overall grasslandcover condition, so the horse Cove chains can be effectively applied to the dynamic changesin evaluation of grassland cover, the results are reliable.
(4) the GIMMS, and MOD13Q1data set, covering the grassland vegetation of YanchiCounty Temporal and spatial analysis studies have shown that: the GIMMS, and MOD13Q1data sets can be used to a certain scale of grassland vegetation cover change, and to goodeffect, this data set bold explorations in the grassland resource monitoring applications andinformation mining.
(1)本文从盐池县为例，根据其草原植被覆盖实际情况出发，依据遥感数据和国家标准(GB19377-2003)为基础，以盐池县草原植被覆盖度为评价的指标，结合遥感植被指数（NDVI）及亚像元分解法（象元二分法），对盐池县草原植被覆盖度进行了遥感估测计算，并依据野外调查数据进行验证，进行盐池县草原覆盖度连续11年的状况变化遥感监测及评价，表明此模型进行植被覆盖度遥感估测是有效切实可行的。
(2)2000年到2010年间，11年间盐池县草原植被覆盖明显好转，极低覆盖植被草原减少了1462.39km~2；中覆盖草原、中高覆盖草原、高覆盖草原增加了586.39km~2，说明虽然整体趋于好转。但状况不甚乐观，低和极低覆盖草原都向上级别转移，说明该县荒漠化在此期间发生逆转，但低覆盖草原部分保留面积较大，说明在这一期草原整体向良性方向发展，但速度缓慢。在人为因素和自然因素的共同作用下，持续地加强草原生态建设方面的投入力度，草原整体状况会有质的变化。应用此方法对宁夏盐池县植被覆盖度取得了较高精度。
(3)在自然及人为因素下，不同草原植被覆盖的面积和等级都在不断演变，而在该县运用马尔科夫链，在2000年、2005年及2009年遥感数据的辅助下，能客观的反应该县的整体草原覆盖状况，因此马儿科夫链能有效应用于草原覆盖的动态变化评价，结果可靠。
(4)利用GIMMS和MOD13Q1数据集，对盐池县草原植被覆盖时空变化特征分析的研究表明： GIMMS和MOD13Q1数据集，可用于一定尺度草原植被覆盖变化的研究，且效果良好，也是对此数据集在草原资源监测应用中的大胆探索与信息挖掘。
Yanchi County at the eastern part of the Ningxia Hui Autonomous Region, betweenlatitude37°04'~38°10', longitude106°30'~107°47' between the northern part of theMu Us desert connected to the south and the Loess Plateau, rely on, is located in aridandsemi-arid transition zone on the geographical location of the transition zone from the LoessPlateau to the Ordos tableland Yanchi County in vegetation from steppe to desert transitionzone in the transition zone from rural areas to the pastoral areas of resource utilization,in atypical pastoral area, is the only animal husbandry county in Ningxia. Grassland degradationin the area under the combined action of natural and man-made, a serious threat to localeconomic development and ecological security. Yanchi County, since the end of2002,implementation of the grazing prohibition, in order to better the area of the development ofthe grassland vegetation status in real-time dynamic monitoring to understand the trends ofgrassland vegetation, according to the national standard (GB19377-2003) and remote sensingdata from Yanchi County grassland vegetation coverage, select the grassland vegetation coveras the evaluation index, vegetation index (NDVI) and the dichotomy pixel (sub-pixeldecomposition), the establishment of Yanchi County grassland vegetation cover remotesensing monitoring system, the region grassland cover conditions changes in remote sensingmonitoring and evaluation, the following results:
(1) Yanchi County proceed according to the grassland vegetation cover the actualsituation, based on remote sensing data and the national standard (GB19377-2003), based onYanchi County grassland vegetation coverage for the evaluation of indicators, combined withremote sensing vegetation index (NDVI) and sub-pixel decomposition method (like thedichotomy), Yanchi County grassland vegetation coverage of remote sensing estimationcalculation and validation based on field survey data, Yanchi County grassland coverage for11consecutive years, the situation changes remote sensingmonitoring and evaluation,indicating that this model of remote sensing estimation of vegetation coverage is effectivepractical.
(2)2000to2010,11years Yanchi County grassland vegetation cover significantlyimproved, very low vegetation cover grassland reduced1462.39km~2; in coverage grassland,high coverage grassland, high coverage grassland increased by586.39km~2, indicating that although as a whole tends to improved. But the situation is not very optimistic, low and verylow coverage grassland up level transfer of the county desertification be reversed during thisperiod, but low coverage grassland part to retain the larger, indicating that in this periodgrassland as a whole to the positive direction, but slow. The common role of the human andnatural factors, continue to enhance the grassland ecological construction investment,grassland overall situation will be a qualitative change. The application of this methodachieved a high accuracy vegetation coverage in Yanchi County.
(3) under the natural and human factors, the size and grade of the different grasslandvegetation cover in the evolving use of Markov chain in the county,2000,2005and2009remote sensing data of the auxiliary objective the reactions of the county's overall grasslandcover condition, so the horse Cove chains can be effectively applied to the dynamic changesin evaluation of grassland cover, the results are reliable.
(4) the GIMMS, and MOD13Q1data set, covering the grassland vegetation of YanchiCounty Temporal and spatial analysis studies have shown that: the GIMMS, and MOD13Q1data sets can be used to a certain scale of grassland vegetation cover change, and to goodeffect, this data set bold explorations in the grassland resource monitoring applications andinformation mining.
引文
查勇,JayGao.2003,国际草地资源遥感研究新进展[J].地理科学进展,22(6):607-617.
陈宝瑞,辛晓平,朱玉霞等.2007,6:内蒙古荒漠化年际动态变化及与气候因子分析[J].遥感应用,39-44.
池宏康,周广胜,许振柱等.2007,草地植被盖度的近距离遥感测定[J].草业学报:105-110.
董永平,吴新宏等.2005草原遥感监测技术[M].北京:化学工业出版社,1-12.
方红亮,田庆久.1998,高光谱遥感在植被监测中的研究综述.遥感技术与应用,13(l):62-69.
高清竹,李玉娥,林而达.2005,藏北地区草地退化的时空分布特征[J].地理学报:965-973
胡良军,邵明安.2001,论水土流失研究中的植被覆盖度量指标[J].西北林学院学报,16(1):40-43.
黄秋燕,于刚,丁琳.2008,都安县植被覆盖度动态变化的遥感反演及驱动力分析[J].南方国土资源:34-37.
黄文广,刘晓东等.2011,禁牧对草地覆盖度的影响—以宁夏盐池县位例[J].草业科学,28(08):1502-1506.
李建龙,黄敬峰,王秀珍.1997草地遥感[M].北京:气象出版社,91-102.
李晓兵,史培军.基于NOAA/AVHRR数据的中国主要植被类型NDVI变化规律研究.植物学报,1999,41(3).
林年丰,汤洁,斯蔼等.2006,松嫩平原荒漠化的EOS-MODIS数据研究[J].第四纪研究:265-272.
刘广峰,吴波,范文义等.2007,基于像元二分模型的沙漠化地区植被覆盖度提取—以毛乌素沙地为例[J].水土保持研究:268-271.
卢敬华,陈伟.1998,植被覆盖与土壤水分的动力学模型.成都气象学院学报,13(3),209-216.
潘晓玲,2001,干早区绿洲生态系统动态稳定性的初步研究.第四纪研究,21(4):345-351.
乔峰,张生英,张克斌等.2006,13(3):宁夏盐池植被覆盖动态变化遥感监测[J].水土保持研究:181-183,186.
秦伟,朱清科,张学霞等.2006,植被覆盖度及其测算方法研究进展［J］.西北农林科技大学学报:自然科学版,(9):163-170.
孙吉雄.2000,草地培育学[M].北京:中国农业出版社,1-2.
仝川,杨景荣.2002,锡林河流域草原植被退化空间格局分析[J].自然资源学报,17(5):201-206."
涂军,熊燕,石德军.1999,青海高寒草甸草地退化的遥感技术调查分析[J].应用与环境生物学报,1999,5(2):131-135.
王洪波,杨发林,2005:宁夏草业[M].银川:宁夏人民出版社,3-27.
卫亚星等.2008,基于MODISNDVI时序数据的青海省草地分级[J].30(5):688-693.
杨吉龙,李家存,杨德明,2001,高光谱分辨率遥感在植被监测中的应用综述.世界地质,20(3):307-312.
詹小国,王平.2001,基于RS和GIS的三峡库区水土流失动态监测研究.长江科学院院报,18(2):41-44.
张宏斌.2007基于多源遥感数据的草原植被状况变化研究.硕士论文[D].北京:中国农业科学院,3-4.
张克斌，李瑞，夏照华等.2006,4(6):宁夏盐池植被盖度变化及影响因子[J].中国水土保持科学.18-22.
张云霞,李晓兵,陈云浩.2003,草地植被盖度的多尺度遥感与实地测量方法综述[J].地球科学进展.18（01）:85-93.
章文波,符素华,刘宝元.2001,目估法测量植被覆盖度的精度分析[J].北京师范大学学报(自然科学版),37(3):402-408.
中国草地资源,中华人民共和国农业部畜牧兽医司,全国畜牧兽医总站主编.1996,北京:中国科学技术出版社,56-59.
邹亚荣,张增祥,赵晓丽.GIS支持下我国干旱区草地资源动态分析[J].环境科学研究,2003,16(1):19-26.
Purevdor JTS,Tateishi R,Ishiyama T,et.1998,Relat ionships betweenpercent veget at ion cover and veget ation indices [J].International Journal of Remot e Sensing,19(18):3519-3535.
Anatoly A Gitelson,Yoram J Kaufinan,Robert Stark,Don Rundquist.2002,Novel algorithms for Remoteestimation of vegetation fraction.Remote Sensing of Environment,80(l):76-87.
Dymond JR,Stephens PR,Newsome PF et al.1992,Percent vegetation cover of a degrading Rangeland fromSPOT.International Journal of Remote Sensing,13(11):1999-2007.
Zhou Q,Robson M,Pilesjo P.1998,On the ground estimation of vegetationcover in Aust ralian rangelands[J].Int ernat ional Journal ofRemot e Sensing,9:1815-1820.
Wilson AD,Abraham NA,BarrattR,et.1987,Evaluation ofmethods of assessing veget at ion change in th esem-i arid rangelandsof southern Aust ralia [J].Aust ralian Rangelan d Journal,9:5-13.
Curran PJ,Williamson HD.1986,Sample size f or ground and remot elysensed dat a[J].Remote S ensing ofEnvironment,20:31-41.
Gutman G,Ignatov A.1998,The derivation of the green vegetation frac ion from NOAA/AVHRR dat a foruse in numerical weather prediction models[J].International Journal of Remote Sensing,19(8):1533-1543.
Jasinsky MF,Eagleson PS.1989(27)The structure of Red,NIR scattergrams of semi-vegetatedlandscapes.Transactions on Geoscience and Remote Sensing,441-451.
Tueker JC.1979,8(2):Red and Photographic infrared linear combination for monitoring vegetation.RemoteSensing of Environment,127-150.
Myneni RB,Asrar G.1994,47(3):Atmospheric effectsand spectral vegetation indices.Remote Sensing ofEnvironment,390-02.
Jordan CF.1969,Derivation of leaf area index from quality of light on the forest floor.Eeology,50:663-666.Richardson AJ and Wiegand CL.1977,Distinguishing vegetation from soil backgroundinformation.Photogrammetic Engineering and Remote Sensing,43:1541-1552.
Rouse JW,Haas R H,Schell JA and Deering DW.1973,Monitoring vegetation systems in the Great Plainswith ERTS.Third erts SymPosium,NASA SP-351,1:309-317.
CriPpen RE.1990,Caleulating the vegetation index faster.Remote Sensing of Environment,34:71-73.
Clevers JGPW.1988,The deviration of a simplified reflectance model for the estimation of leaf areaindex.Remote Sensing of Environment,33:53-70.
Huete AR.1988,A soil adjusted vegetation index (SAVI).Remote Sensing of Environment,25:295-309.
Qi J,Chehbouni A,Huete AR,Derr YH and Sorooshian S.1994,A modified soil adjusted vegetationindex[J].Remote sensing of environment,48(2):119-126.
Baret and Guyot G.1991,Potentials and limits of vegetation indices for LAI and APAR assessment.RemoteSensing of Environment,35:161-173.
陈宝瑞,辛晓平,朱玉霞等.2007,6:内蒙古荒漠化年际动态变化及与气候因子分析[J].遥感应用,39-44.
池宏康,周广胜,许振柱等.2007,草地植被盖度的近距离遥感测定[J].草业学报:105-110.
董永平,吴新宏等.2005草原遥感监测技术[M].北京:化学工业出版社,1-12.
方红亮,田庆久.1998,高光谱遥感在植被监测中的研究综述.遥感技术与应用,13(l):62-69.
高清竹,李玉娥,林而达.2005,藏北地区草地退化的时空分布特征[J].地理学报:965-973
胡良军,邵明安.2001,论水土流失研究中的植被覆盖度量指标[J].西北林学院学报,16(1):40-43.
黄秋燕,于刚,丁琳.2008,都安县植被覆盖度动态变化的遥感反演及驱动力分析[J].南方国土资源:34-37.
黄文广,刘晓东等.2011,禁牧对草地覆盖度的影响—以宁夏盐池县位例[J].草业科学,28(08):1502-1506.
李建龙,黄敬峰,王秀珍.1997草地遥感[M].北京:气象出版社,91-102.
李晓兵,史培军.基于NOAA/AVHRR数据的中国主要植被类型NDVI变化规律研究.植物学报,1999,41(3).
林年丰,汤洁,斯蔼等.2006,松嫩平原荒漠化的EOS-MODIS数据研究[J].第四纪研究:265-272.
刘广峰,吴波,范文义等.2007,基于像元二分模型的沙漠化地区植被覆盖度提取—以毛乌素沙地为例[J].水土保持研究:268-271.
卢敬华,陈伟.1998,植被覆盖与土壤水分的动力学模型.成都气象学院学报,13(3),209-216.
潘晓玲,2001,干早区绿洲生态系统动态稳定性的初步研究.第四纪研究,21(4):345-351.
乔峰,张生英,张克斌等.2006,13(3):宁夏盐池植被覆盖动态变化遥感监测[J].水土保持研究:181-183,186.
秦伟,朱清科,张学霞等.2006,植被覆盖度及其测算方法研究进展［J］.西北农林科技大学学报:自然科学版,(9):163-170.
孙吉雄.2000,草地培育学[M].北京:中国农业出版社,1-2.
仝川,杨景荣.2002,锡林河流域草原植被退化空间格局分析[J].自然资源学报,17(5):201-206."
涂军,熊燕,石德军.1999,青海高寒草甸草地退化的遥感技术调查分析[J].应用与环境生物学报,1999,5(2):131-135.
王洪波,杨发林,2005:宁夏草业[M].银川:宁夏人民出版社,3-27.
卫亚星等.2008,基于MODISNDVI时序数据的青海省草地分级[J].30(5):688-693.
杨吉龙,李家存,杨德明,2001,高光谱分辨率遥感在植被监测中的应用综述.世界地质,20(3):307-312.
詹小国,王平.2001,基于RS和GIS的三峡库区水土流失动态监测研究.长江科学院院报,18(2):41-44.
张宏斌.2007基于多源遥感数据的草原植被状况变化研究.硕士论文[D].北京:中国农业科学院,3-4.
张克斌，李瑞，夏照华等.2006,4(6):宁夏盐池植被盖度变化及影响因子[J].中国水土保持科学.18-22.
张云霞,李晓兵,陈云浩.2003,草地植被盖度的多尺度遥感与实地测量方法综述[J].地球科学进展.18（01）:85-93.
章文波,符素华,刘宝元.2001,目估法测量植被覆盖度的精度分析[J].北京师范大学学报(自然科学版),37(3):402-408.
中国草地资源,中华人民共和国农业部畜牧兽医司,全国畜牧兽医总站主编.1996,北京:中国科学技术出版社,56-59.
邹亚荣,张增祥,赵晓丽.GIS支持下我国干旱区草地资源动态分析[J].环境科学研究,2003,16(1):19-26.
Purevdor JTS,Tateishi R,Ishiyama T,et.1998,Relat ionships betweenpercent veget at ion cover and veget ation indices [J].International Journal of Remot e Sensing,19(18):3519-3535.
Anatoly A Gitelson,Yoram J Kaufinan,Robert Stark,Don Rundquist.2002,Novel algorithms for Remoteestimation of vegetation fraction.Remote Sensing of Environment,80(l):76-87.
Dymond JR,Stephens PR,Newsome PF et al.1992,Percent vegetation cover of a degrading Rangeland fromSPOT.International Journal of Remote Sensing,13(11):1999-2007.
Zhou Q,Robson M,Pilesjo P.1998,On the ground estimation of vegetationcover in Aust ralian rangelands[J].Int ernat ional Journal ofRemot e Sensing,9:1815-1820.
Wilson AD,Abraham NA,BarrattR,et.1987,Evaluation ofmethods of assessing veget at ion change in th esem-i arid rangelandsof southern Aust ralia [J].Aust ralian Rangelan d Journal,9:5-13.
Curran PJ,Williamson HD.1986,Sample size f or ground and remot elysensed dat a[J].Remote S ensing ofEnvironment,20:31-41.
Gutman G,Ignatov A.1998,The derivation of the green vegetation frac ion from NOAA/AVHRR dat a foruse in numerical weather prediction models[J].International Journal of Remote Sensing,19(8):1533-1543.
Jasinsky MF,Eagleson PS.1989(27)The structure of Red,NIR scattergrams of semi-vegetatedlandscapes.Transactions on Geoscience and Remote Sensing,441-451.
Tueker JC.1979,8(2):Red and Photographic infrared linear combination for monitoring vegetation.RemoteSensing of Environment,127-150.
Myneni RB,Asrar G.1994,47(3):Atmospheric effectsand spectral vegetation indices.Remote Sensing ofEnvironment,390-02.
Jordan CF.1969,Derivation of leaf area index from quality of light on the forest floor.Eeology,50:663-666.Richardson AJ and Wiegand CL.1977,Distinguishing vegetation from soil backgroundinformation.Photogrammetic Engineering and Remote Sensing,43:1541-1552.
Rouse JW,Haas R H,Schell JA and Deering DW.1973,Monitoring vegetation systems in the Great Plainswith ERTS.Third erts SymPosium,NASA SP-351,1:309-317.
CriPpen RE.1990,Caleulating the vegetation index faster.Remote Sensing of Environment,34:71-73.
Clevers JGPW.1988,The deviration of a simplified reflectance model for the estimation of leaf areaindex.Remote Sensing of Environment,33:53-70.
Huete AR.1988,A soil adjusted vegetation index (SAVI).Remote Sensing of Environment,25:295-309.
Qi J,Chehbouni A,Huete AR,Derr YH and Sorooshian S.1994,A modified soil adjusted vegetationindex[J].Remote sensing of environment,48(2):119-126.
Baret and Guyot G.1991,Potentials and limits of vegetation indices for LAI and APAR assessment.RemoteSensing of Environment,35:161-173.