基于MODIS_NDVI的新疆巴音布鲁克草原植被覆盖度时空变化特征及影响因素分析

发布时间：2018-01-06 11:32

本文关键词：基于MODIS_NDVI的新疆巴音布鲁克草原植被覆盖度时空变化特征及影响因素分析　出处：《河北师范大学》2016年硕士论文　论文类型：学位论文

【摘要】：草原占地球陆地总面积的52.17%,是全球面积最大的陆上土地覆盖类型。草原植被是畜牧业的发展的基础具有防风固沙、保持水土、涵养水源、维护生物多样性和旅游观赏等生态功能。草原植被覆盖度是描述草原生态系统健康以及生态承载力的重要参数之一,植被覆盖度的高低会显示区域生态环境的质量,从而指示区域内部生态系统功能和生态系统服务能力的强弱。新疆巴音布鲁克草原是中国第二大草原,拥有独特的高寒草甸草原景观,是新疆天山地区重要的生态屏障。近年来,气候变化和超载放牧所引起的草原退化现象日益严重,因此,研究巴音布鲁克草原植被覆盖度时空变化特征及影响因素,意义重大。本文运用回归分析、趋势分析、缓冲区分析等方法,以MODIS遥感数据和野外样地实测数据为基础,建立了估算植被覆盖度的巴音布鲁克草原MODIS光谱估算模型。根据该模型对研究区植被覆盖度进行等级划分,对典型地物要素进行样线和缓冲区分析,并对研究区2001-2015年的植被覆盖度时空变化特征和主要影响因素进行研究。主要结论如下：(1)构建植被覆盖度光谱估算模型。利用实测SOC_NDVI估测植被覆盖度VC的地面光谱模型是简单的线性函数,R2为0.865；MODIS_NDVI和实测SOC_NDVI之间也为线性关系,R2为0.731；基于上述两个方程反演了估算植被覆盖度的MODIS光谱估算模型,经检验,MODIS光谱估算模型相关系数R为0.882,相关性显著,标准误差SE为11.55%,总体预测精度为88.92%。因此可得出,基于地面光谱模型建立的MODIS光谱估算模型预测精度较高,可估算巴音布鲁克草原生长季的植被覆盖度,具有较好的科学性和应用性。(2)基于所建模型划分研究区草原等级。根据本文得到的MODIS光谱估算模型,将研究区遥感影像NDVI灰度图转化为单波段的植被覆盖度图,并按植被覆盖度的高低将试验期草原划分为五个等级,即极高、高、中、低和极低覆盖草原,它们在研究区中所占比例分别为：18.87%、25.61%、31.28%、13.86%和10.38%。在此基础上从等级划分、地物要素和分区三个方面进行植被覆盖度分析,清晰地呈现了研究区植被覆盖度的内部差异性。(3)对研究区典型地物要素进行样线分析和缓冲区分析。以研究区植被覆盖度图为基础,在研究区内地物要素分布特征明显的区域设置A、B、C、D4条样线和4个缓冲区,依次对应矿山、道路、旅游景点、居民聚居点。通过分析总结得出,典型地物要素对草原植被覆盖度均有不同程度的影响,且距离地物要素距离越近,植被覆盖度越低,对草原影响程度从大到小依次为：矿山居民点道路旅游景点。(4)分析研究区草原植被覆盖度的时空格局。利用MODIS光谱估算模型计算得到研究区2001-2015年的植被覆盖度图,整体上东部和西部低而南部和北部高,且随着时间序列植被覆盖度呈弱增长趋势。全区尺度上,2006-2010年较2001-2005年,低植被覆盖度的面积增加,极高植被覆盖度的面积降低,整体来看植被覆盖度为下降趋势；2011-2015年较前两个时段,低、中植被覆盖度的面积降低而高、极高植被覆盖度的面积增加,植被出现恢复势态。像元尺度上,研究区内76.38%的区域植被覆盖度呈增长态势,23.62%的区域植被覆盖度呈降低态势,增长最大值为1.88%,下降最大值为1.59%。(5)分析时空格局变化的主要影响因素。巴音布鲁克植被覆盖度变化是自然因素和人类活动共同作用的结果。通过分析降水、气温、海拔高度、畜牧量与研究区植被覆盖度之间的相关性,得出相关性R的绝对值从大到小依次为：降水(0.916)气温(0.571)畜牧量(0.50)海拔高度(0.115),其中植被覆盖度与畜牧量呈负相关关系。巴音布鲁克草原因其特有的水分、气候以及土壤条件决定了其特有的植被覆盖度空间分布和变化特征,降水对植被的生长具有重要的作用,植被覆盖度的增加主要取决于降水量的增加,而延长植被生长期会增加植被覆盖度,这依赖于气温的升高。地形因素对植被覆盖度的影响还需进一步验证,人类活动主要通过草原放牧以及旅游行为对植被覆盖度造成影响。
[Abstract]:Grassland accounted for 52.17% of the total land area of the earth, is the world's largest onshore land cover types. The grassland vegetation is the basis for the development of animal husbandry has the wind and sand, soil and water conservation, water conservation, biodiversity maintenance and tourism and other ecological functions. The grassland vegetation coverage is one of the important parameters to describe the grassland ecosystem health and the ecological carrying capacity, low vegetation coverage will show the quality of the regional ecological environment, thereby indicating function and ecosystem service capacity of the regional ecological system internal strength. Xinjiang is second Chinese in Bayinbuluk grassland steppe, alpine meadow grassland has a unique landscape, is an important ecological barrier of Xinjiang Tianshan area. In recent years, climate change and overgrazing caused by grassland degradation is becoming increasingly serious. Therefore, the research in Bayinbuluk grassland vegetation coverage in the temporal and spatial characteristics And the influence factors, of great significance. This paper used regression analysis, trend analysis, buffer analysis method, using the MODIS remote sensing data and the field measured data as the basis, establish the estimation model to estimate the MODIS spectrum in Bayinbuluk grassland vegetation coverage. According to the model of forest vegetation coverage in the study area is classified, analyzed sample line and buffer of typical objects, and 2001-2015 years of research on vegetation coverage factors of temporal and spatial variation characteristics and the main influence is studied. The main conclusions are as follows: (1) construction of vegetation coverage spectrum estimation model. The ground spectral estimation model of vegetation coverage by VC SOC_NDVI test is a simple linear function, R2 0.865; MODIS_NDVI and SOC_NDVI for the linear relationship between the measured and the R2 was 0.731; the two estimate equation inversion of MODIS spectrum estimation model based on vegetation coverage, After inspection, MODIS spectral estimation model of correlation coefficient R is 0.882, the correlation is significant, the standard error of SE is 11.55%, the overall prediction accuracy of 88.92%. therefore can be obtained, high precision MODIS spectra of building surface spectral estimation model based on the model can estimate the Ba Abrookt native long season of vegetation coverage, good science and application of. (2) based on the model, the study area is divided according to the grassland level estimation model for MODIS spectra obtained in this paper, the study area of remote sensing image NDVI grayscale into a single band of the vegetation coverage map, and according to the degree of vegetation coverage in the test period of grassland is divided into five levels, namely high, high, in, low and very low coverage grassland, they accounted for in study area are respectively: 18.87%, 25.61%, 31.28%, 13.86% and 10.38%. on the basis of the classification, the three aspects of objects and partitions for planting Coverage analysis, clearly presents the internal differences of vegetation coverage research. (3) analyze the sample line analysis and buffer of typical objects in study area. To study vegetation coverage map based on the distribution characteristics of the study area features obvious regional set A, B. C, D4 lines and 4 buffer, followed by the corresponding mines, roads, tourist attractions, community. Through the analysis concluded that the typical features elements covering different influence degrees of grassland vegetation, and the distance between objects closer, the vegetation coverage is low, the degree of influence on the grassland to small is as follows: the mining settlements road tourist attractions. (4) analysis of temporal and spatial patterns of grassland vegetation coverage. The calculated study area 2001-2015 estimation model using MODIS spectral vegetation map, the East and West and South and low North high, and with the time series of vegetation coverage showed a weak growth trend. The scale, 2006-2010 is 2001-2005, an area of low vegetation coverage increased, extremely high vegetation coverage area decreased, overall vegetation coverage decreased; 2011-2015 years than the previous two periods, low vegetation the coverage area is reduced and the high, very high vegetation coverage area increased, the vegetation recovery situation. The pixel scale, vegetation coverage in the study area of 76.38% growth, vegetation coverage decreased 23.62% growth trend, the maximum value is 1.88%, the largest decline in value of 1.59%. (5) the main factors affecting the analysis of temporal and spatial patterns change. Vegetation coverage change Bayinbuluk is the interaction of natural factors and human activities. Through the analysis of the precipitation, temperature, altitude, animal husbandry and the vegetation cover degree The absolute value of the correlation, the correlation of R from big to small in order: (0.916) precipitation temperature (0.571) animal husbandry (0.50) altitude (0.115), the vegetation and animal husbandry are negatively correlated. In Bayinbuluk grassland because of its unique water, climate and soil conditions determines its unique vegetation cover spatial distribution and change characteristics, plays an important role in the growth of precipitation on vegetation, increase vegetation coverage increase mainly depends on precipitation, and prolong the vegetation growth period will increase the vegetation coverage increased, which depends on the temperature. The terrain factors impact on vegetation coverage also need further verification, human activities mainly through grazing and tourist behavior on the impact of vegetation coverage.

【学位授予单位】：河北师范大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：S812

【相似文献】