【作者】 李惠敏；

【导师】 武伟；

【作者基本信息】 西南大学 ， 土壤学， 2010， 硕士

【摘要】 植被的生长一般受到多方面因素的影响,即相对固定的地形因子,长时间累计变化的温度、降水等气候因子,人口分布、GDP增加及城市化过程等的人为因子。其中地形因子和气候因子可称为自然因子,它们奠定了植被空间分布的总体格局。人为因子则引起植被的局部变化。从长期来看,自然因子和人为因子都驱动着植被分布格局及其变化,但是在短期内,人类活动则是主要的驱动因素。本文主要研究地形因子、气候因子和人为因子与植被覆盖之间的关系,从而更好的解释植被变化情况,为重庆市生态环境的调节改善提供一定的依据。本研究运用1998-2007年1 km空间分辨率的SPOT-VGTS10数据,提取重庆市归一化植被指数(normalized difference vegetation index, NDVI)影像,共计351景。采用Savitzky-Golay滤波平滑处理、最大值合成法、线性回归分析、时间序列分析等方法研究重庆市NDVI、降水量、温度、人口、GDP等的动态变化情况,构建重庆市的NDVI、气象因子、人为因子的时序数据集；分析重庆市NDVI及各因子的空间分布格局；探讨重庆市NDVI与气象因子,地形因子,人为因子之间的关系。主要研究成果包括：(1)运用Savitzky-Golay滤波技术,可有效地去除NDVI数据中的云、气溶胶及异常值的影响,NDVI时序数据经过滤波平滑后,能更好地反映植被覆盖情况,为研究重庆市植被覆盖变化奠定了基础。(2)重庆市NDVI的年平均值和各年NDVI值都表现出较强的稳定性。从1999年到2007年NDVI总体呈上升趋势,说明重庆市植被量逐年增加。空间分布特征表明,在四个种植业分区(渝东北、渝南、渝中和渝西)各年份NDVI值大小趋势基本保持一致。根据NDVI值排序表现为：渝东北>渝南>渝中>渝西。而各年NDVI则表现出明显的季节性,在1月份和8月份分别为谷值和峰值。另外NDVl年际变化率也表明重庆市近十年植被覆盖量增加,且西部除主城外最为明显。(3)以区县为单位研究各区域旬NDVI与同期降水量、平均气温的相关性发现,NDVI与二者均有较好的相关性,但是NDVI与平均气温相关性远远大于与降水量的相关性。季节NDVI与同期降水量及温度相关性的空间分布结果表明,春季降水量与NDVI在空间上呈正相关；气温对NDVI的影响较为显著,正相关面积占总面积的比例达到72%。夏季NDVl与降水量仅在西北部及中北部个别地区为正相关,其余各地均呈负相关；NDVI与温度总体以正相关为主,在西部地区尤其是主城地区呈现较显著的负相关性；东南部地区及西南部相关性较小。秋季东北部大巴山区NDVI与降水量相关系数较大,其余地区呈负相关；NDVI与温度的相关性则明显呈现出东部地区为正相关,西部地区为负相关的现象。在冬季,东北及东南部各地NDVI受到降水量的影响较大,长江以北大部分地区与降水量呈负相关;与其他季节相比,东北部及东南部地区冬季NDVI受温度的影响相对有所减小,部分地区的植被情况受温度的影响并不大,而西部则呈现很强的正相关性,相关系数达到0.8左右。(4)NDVI与高程有一定的相关性,其中渝东北地区相关性最为明显。而坡度、坡向与NDVI的相关性不明显。(5)重庆市最大化NDVI与人口及GDP均呈负相关,负相关的面积分别达到51.59%和54.39%；通过研究1999年到2006年积分NDVI与城市化的关系表明,城市化对重庆市的植被也有一定的影响,与城市化率呈负相关关系。更多还原

【Abstract】 The growth of vegetation was influenced by many factors such as terrain, temperature, precipitation, population distribution, GDP, and urbanization, etc. Among these factors, terrain and climate are classified as natural factors which formed the overall pattern of spatial distribution of vegetation. Human factor, such as population distribution, GDP, and urbanization, would influence on local changes of vegetation. In the long run, both natural factors and human factors would drive the changes of vegetation patterns. However, in the short term, human activities would be the main driving force in the change of vegetation. Thus, the objectives of this study were to investigate the relationships between the vegetation cover and terrain, between the vegetation cover and climate, and between the vegetation cover and human factors. The results of the study would provide a guideline for regulating the environment in Chongqing.A total of 351 images of SPOT-VGTS10 with a spatial resolution of 1 km from 1998 to 2007 were used to extract normalized difference vegetation index (NDVI) for Chongqing. Several datasets, such as NDVI, climate, and human factors, were developed. Savitzky-Golay smoothing filter, maximum value composites, linear regression analysis, and time series analysis were applied to investigate the inter-annual variations of NDVI, precipitation, temperature, population, and GDP. and constructed the database of NDVI, meteorology, and human factors. The spatial distribution of NDVI and other factors were explored. The correlations between NDVI and meteorology, terrain attributes, and human factors were analyzed. The main results are as follows:(1) Clouds, aerosols and outliers could be effectively removed from the raw NDVI dataset using Savitzky-Golay filtering technology. Smoothed and filtered NDVI time series data would better reflect the presence of the vegetation coverage of Chongqing and provide the data source for studying the vegetation cover change in Chongqing. (2) The annual average NDVI and NDVI values for each year showed strong stability in Chongqing. The amount of vegetation increased yearly in Chongqing. The variations of NDVI were similar in four plantation areas, namely, northeastern, southern, central, and western Chongqing. The order of NDVI values was northeastern> southern> central> western Chongqing. The NDVI for each year showed a seasonal variation, the lowest value was found in January while the highest was in August. Moreover, the variations of inter-annual indicated that the vegetaion increase of Chongqing, especially in the western Chongqing.(3) For each county, NDVI presented strong relationship with mean temperature and weak relationship with precipitation at 10-day intervals. The spatial relationships between NDVI and climate were also investigated at seasonal scale.In spring, positive relationships were found between NDVI and precipitation and between NDVI and temperature. The area percentage of positive relations of NDVI and temperature was 72%. In summer, NDVI was positively related with precipitation in northwestern and northcentral Chongqing. Generally, NDVI showed positive relationship with temperature in the study area. However, significant negative relationships were found in western part and downtown and weak relationships existed in southeastern and southwestern regions. In autumn, NDVI was closely related with precipitation in Daba Mountain areas. NDVI showed positive relationship with temperature in eastern Chongqing and negative relationship in western Chongqing. In winter, NDVI was strongly influenced by precipitation in northeastern and southeastern regions and the correlation coefficient was negative in the north of the Yangtze River. However, the influence of temperature on NDVI in northeastern and southeastern regions was decreased compared with other seasons. NDVI of some areas was not affected by temperature, while NDVI was strongly positive related with temperature in western part.(4) Elevation showed positive relation with NDVI, especially in eastern Chongiqng. However, no relationship was found between slope and NDVI and between aspect and NDVI.(5) There was a negative correlation between maximum NDVI and population and between NDVI GDP in Chongqing, the area percentage was 51.59% and 54.39%, respectively. The integration of NDVI was nagetive related with urbanization ratio indicating the effects of urbanization on the vegetation coverage in the study area.更多还原