不同NDVI年值提取方法对秦岭山地植被变化及其气候响应的影响

Impact of Different Extracted Methods of Annual NDVI on Vegetation Cover Change and Their Response to Climate Change in Qinling Mountains

基于2000—2014年秦岭山地MODIS NDVI影像、气温和降水数据,分别利用平均值法(AVM)、最大值合成法(MVC)、平均-最大值合成法(AMVC)、时序重建方法(RAVM)获取NDVI年值,分析了不同方法下NDVI时空变化过程与趋势,及其与气温、降水的相关性,以期为秦岭山地植被动态变化及其气候响应准确评估提供方法基础。结果表明:RAVM获取的NDVI年值效果较好,平均值居中,线性变化趋势极显著。其次为AMVC、MVC和AVM法获取的NDVI年值偏高和偏低。山地植被动态监测中,植被退化是关注的热点问题,AMVC方法计算得到的秦岭山地NDVI线性减少趋势分布面积最大,其次为MVC、RAVM、AVM方法。不同NDVI年值提取方法下,NDVI与气温的相关性具有一定的差异。如相关系数空间分散分布(MVC和RAVM)和相对集中分布(AVM和AMVC),NDVI与气温的正相关为主(AVM和RAVM)和负相关为主(MVC和AMVC)。但4种不同方法下NDVI与降水的相关性均以正相关为主,表明不同方法表现植被与降水关系具有一致性,而植被与气温的关系不稳定。

On the basis of the MODIS NDVI images,temperature and precipitation data from 2000 to2014 in Qinling Mountains,the spatial and temporal variation of NDVI and the correlation between NDVI and temperature,precipitation were analyzed with 4 different extraction methods on annual NDVI value,such as AVM（Average Value Method）,MVC（Maximum Value Composition）,AMVC（Average-Maximum Value Composition）and RAVM method（Reconstruction-Average Value Method）.This study would provide an information to accurate assessment of NDVI change in Qinling Mountains.The results showed that the RAVM was the best to extract annual value of NDVI,and the linear change trend was very significant,followed by the AMVC,while the annual value of NDVI extracted by MVC and AVM were higher and lower.Vegetation degradation was a hot issue for the monitoring of Qinling Mountains.The linear reduction trend area calculated by AMVC was the largest,followed by the MVC,RAVM and AVM.The correlation between NDVI and temperature was different to a certain extent,such as the spatial distribution of coefficient was relatively scattered（MVC and RAVM）or concentrated（AVM and AMVC）,and the positive（AVM and RAVM）or negative correlation（MVC and AMVC）.But the correlation between NDVI and precipitation was positive for the 4 different methods,which reflect the correlation between NDVI and precipitation was stable and unstable between NDVI and temperature.