资源简介

摘要： 基于遥感的植被物候分析具有连续观测、覆盖面广的优点，已经成为全球气候变化背景下植被生理物理变化研究的重要手段。基于青藏高原96个气象站点和对应的MODIS NDVI时序数据，探索分析了2000~2014年间站点、典型植被覆盖和高原整体区域上的植被物候变化趋势及地理环境影响。首先采用三次样条(Spline)函数法、双逻辑斯蒂(D\|L)函数法和奇异谱分析法(SSA)进行NDVI时序数据重构，其次利用导数法(Derivative method，Der)及阈值法(Threshold method，Trs)提取植被物候关键参数(生长季长度LOS、生长季开始点SOS、生长季结束点EOS)，进而对比分析6种方法提取结果的差异和适用条件；再利用M\|K test趋势分析法计算了各站点和区域的物候发展趋势，并探索其与海拔、降水、温度间的相关性；最后将本文遥感LOS与气温阈值法得到的生长期长度指标(GSL)进行了对比验证。结果表明：①阈值法有较强的稳定性及适应性，导数法结果差异较大，整体来看阈值法优于导数法。草原及森林覆被类型中3种物候参数使用SSA、Spline、D\|L与阈值法的组合提取结果较优；②多种方法提取的多种物候参数趋势空间差异性较大，小尺度上趋势比较一致。青藏高原东南部湿润半湿润的灌木草原区及西北荒漠草原区，呈现SOS和EOS推迟、但LOS仍然延长的趋势；青藏高原西南部湿润区呈现SOS延后、EOS提前、LOS缩短的趋势；青藏高原分布广泛的草原区，各物候参数并未呈现显著变化的趋势特征；③温度与物候参数发展趋势呈相关关系，温度升高的情况下呈现SOS提前、EOS延后的现象。由于高原地貌、气候的复杂性，多数站点的物候发展趋势与海拔、降水的相关性并不显著，GSL与LOS的相关性也呈现相似的特征。本文证实在植被物候参数遥感提取中，多种方法差异较大，很少有在全部区域均适用的方法；站点尺度上青藏高原在近15年来植被物候并未呈现明显春季提前或秋季延后趋势。  关键词： 青藏高原;  NDVI时序重构;  植被物候参数;  趋势分析;  阈值法;  导数法     Abstract: In the context of global climate change，vegetation phenology analysis based on remote sensing has become an critical method for studying the characteristics of physical and physiological changes of vegetation.This paper uses the MODIS NDVI time\|series data of 96 meteorological stations over the Tibetan Plateau during 2000\|2014 to explore the development trend of vegetation phenological and geographical environment factors of each meteorological station，typical vegetation coverage and the whole plateau region.Firstly，using three cubic spline function method (Spline)，double logistic function method(D\|L)and singular spectrum analysis (SSA)，NDVI time\|series data is reconstructed，then using the derivative method (Der)and threshold method (Trs)，the key parameters of phenological information is extracted，after that differences and application conditions between the six methods are analyzed and compared.Secondly，using M\|K test trend analysis method，the phenological development trend of each site and area were calculated，the relationship between phenological development trend and altitude，precipitation，temperature is studied.Finally，by the Growing season length(GSL)obtained by temperature threshold method，LOS is compared and verified.in grassland and forest land cover types，SSA，Spline，D\|L combined with threshold method to get the Start of Season(SOS)，end of season(EOS)，Length of season (SOS)respectively is a good combination strategy.(2)The spatial differences of various phenological parameters extracted by different methods are large，and the trend is relatively consistent at small scales.Southeast humid and semi\|humid shrub steppe region and northwestern desert steppe in the Tibetan Plateau，SOS and EOS delayed，but LOS prolonged；southwestern humid region，SOS and EOS delayed，LOS shortened；widely distributed grassland，the phenological parameters did not show significant tendency.(3)Temperature is related to the development trend of phenological parameters.With temperature increasing，the phenomena of SOS advance，EOS lag are presented.Because of the complexity of the plateau landform and climate，there was no significant relationship between phenological development trend for most of the site with the altitude and precipitation，only a few sites have strong correlation，the correlation between GSL and LOS also showed similar characteristics.For remote sensing based phonological analyses，our study identify there is no method existing here that is a adaptive across all the Tibetan Plateau.in addition，at point scale the phenological parameters do not represent a significant earlier or later trend.

相关资源

1. 水质遥感监测的关键要素叶绿素a的反演算法研究进展
2. 基于无人机遥感技术的森林参数获取研究进展
3. 基于无线传感器网络的森林生态系统观测试验平台构建
4. 基于无人机激光雷达的森林冠层高度分析
5. 多源时序国产卫星影像的森林火灾动态监测
6. 基于无人机遥感技术的台风灾害倒伏绿化树木检测

下载说明

·如果您发现该资源不能下载，请通知管理员.gissky@gmail.com

·为确保下载的资源能正常使用，请使用[WinRAR v3.8]或以上版本解压本站资源,缺省解压密码www.gissky.net ，如果是压缩文件为分卷多文件，请依次下载每一个文件，并按照顺序命名为1.rar,2.rar,3.rar...，然后鼠标右击1.rar解压.

·为了保证您快速的下载速度，我们推荐您使用[网际快车]等专业工具下载.

·站内提供的资源纯属学习交流之用,如侵犯您的版权请与我们联系.