【作者】 滑永春；

【导师】 彭道黎；

【作者基本信息】 北京林业大学 ， 森林经理， 2010， 博士

【摘要】 植被覆盖变化是遥感、环境学、气象学等诸多学科所关注的一个重要领域,不仅是遥感应用中需要解决的一个重要环节,也是气候变化分析和地表碳循环等诸多领域研究的基础。本文选择京津风沙源工程治理区作为研究区,以MODIS-NDVI时间序列影像为主要数据源(2000年、2004年、2008年),系统研究了研究区的植被覆盖变化过程与驱动机制。对5种不同的植被指数的研究表明,归一化植被指数(NDVI)探测低盖度植被的能力最强,全球环境指数(GEMI)消除土壤背景影响的效果最好,综合考虑研究区区植被的特点及其植被—土壤相互作用的光谱反映特征后认为,NDVI是研究区植被变化度量的首选植被指数,其次是GEMI。通过NDVI和自然因子之间回归分析,得出,造成研究区内植被空间分布格局不均等的主要原因是降雨,在西北干旱半干旱的地区植被等级最低,特别容易造成荒漠化现象。植被盖度和降雨,10°积温,干燥度,DEM,经度,温度,纬度等自变量因子回归关系为：植被盖度=0.917*(降雨)-0.026*(10°积温)+0.22*(干燥度)+0.286*(DEM)+0.221*(经度)+0.037*(温度)从2000年到2008年,京津风沙源工程治理区植被覆盖度总体呈现增长趋势。尽管不同植被盖度类型间转入转出频繁,治理与破坏同时存在,但总体上来说荒漠化程度在降低,植被由低等级转为高等级,植被状况明显好转。通过四个分区的研究发现,除了农牧交错地带沙化土地治理区低植被有增加趋势以外,其余3个分区均有不同程度的减少。其中,北部干旱草原沙化治理区是低盖度植被主要分布区,在浑善达克沙地治理区的翁牛特旗等地区也存在大量的低盖度植被,但该区是低植被向高植被转化最快的地区,燕山丘陵山地水源保护区是植被覆盖最好的地区。通过景观格局指数分析可以看出,整个研究区斑块数在减少,最大斑块指数呈现先减少后增加趋势,景观多样性指数、均匀度指数都在变小,聚集度指数在增加。各镶嵌体比例分配由均匀逐渐趋向于不均匀,各景观要素所占面积比例差异在进一步拉大,低盖度植被向着高盖度植被转移。高盖度植被面积逐渐增加,优势斑块呈越来越明显的变化态势。通过各种促使植被变化的驱动力因素分析得到,从长时间尺度上看,自然和人为因素都驱动着植被覆盖度及其景观的变化,但在短时间尺度上,自然因素具有相对稳定性,而人类活动无疑是植被覆盖度及其景观变化最主要的驱动因素,尤其是本研究区选取的是京津风沙源工程治理区,各种林业工程是相比较其他因素更为重要。更多还原

【Abstract】 Vegetation cover change is an important area of remote sensing, environmental science, meteorology, and many other disciplines to concern to, which is not only an important aspect for remote sensing to deal with, but also base of climate change analysis, carbon cycle of surface area and many other researches. This paper choices Beijing and Tianjin sandstorm source control area as a research area, with MODIS-NDVI images of time series(2000,2004,2008) as main data source, to systematically study vegetation cover change and driving mechanism.Five different kinds of vegetation indices studies have shown that, Normalized Difference Vegetation Index (NDVI) is the strongest to detect low vegetation cover and Global Environment Index (GEMI) is the best to eliminate the effect of soil background influence. Considering the characteristics of vegetation and spectrum of vegetation-soil interaction, NDVI is first vegetation index to measure vegetation change in study area, followed by GEMI.Through regression analysis between NDVI and natural factor, the reason that results in uneven spatial pattern of vegetation in study area is mainly rainfall. In the northwestern arid and semi-arid vegetation, the lowest vegetation level, in particular, likely causes desertification. Regression equation between vegetation cover and independent variables, that is, rainfall,10°accumulated temperature, aridity, DEM, longitude, temperature, latitude is as follows:vegetation cover=0.917 (rainfall)-0.026(10°accumulated temperature)+0.22 (aridity)+0.286 (DEM)+0.221 (longitude)+0.037 (temperature)From 2000 to 2008, in general, vegetation coverage of Beijing and Tianjin sandstorm source control area showed growth. Despite different vegetation cover changed frequently, in the mean tine, treatment and destruction existed, generally desertification degree reduced, vegetation transfered from low grade to high-grade, and vegetation has clearly risen. By comparing results of 4 partitions, In addition to increasing in farming-pastoral desertification control zone, low vegetation had different levels of reduction in other three partitions. Northern arid grassland desertification control area is main distribution area of low vegetation cover, and there were a large number of low vegetation cover in sand, which, transferred fast from low vegetation to high vegetation..Yanshan hill water source protection area was the best area of vegetation cover. Through analysis of landscape pattern indices, in entire study area, number of patch, landscape diversity and evenness index reduced, while aggregation index increased. Proportion of mosaic gradually moved from evenness to non-uniform and difference of proportion of landscape element widened further. Low vegetation cover shifted toward high vegetation, so high-cover vegetation increased gradually.Through a variety of driving force to promote vegetation change, from the point of long time-scale, natural and man-made factors are driving vegetation cover and landscape change. In the short time scale, natural factor has relative stability and human activities are undoubtedly the most important driver of landscape and vegetation cover change. Especially in Beijing and Tianjin sandstorm source control area, all kinds of forestry projects are more important compared to other factors.更多还原