洞庭湖流域植被物候特征参数遥感提取

Extraction of the Vegetation Phenological Characteristic Parameters with Remote Sensing in Dongting Lake Basin

以洞庭湖流域为研究区域,对MODIS13Q1植被产品数据的所有像元进行EVI时间序列重构后,采用滑动平均法、动态阈值法、最大变化斜率法以及Logistic函数拟合法等4种方法进行物候参数提取。对比分析后,选择了效果最佳的滑动平均法对2005年及2015年洞庭湖流域的EVI时间序列的遥感影像进行植被物候参数提取,分析近10年来研究区域植被物候特征变化。结果表明:滑动平均法提取的植被物候参数较准确,适用性较强;利用遥感手段能及时准确地监测大区域尺度植被物候变化,为研究区域气候变化等提供参考。

This study had used MODIS 13Q1 vegetation product data in Dongting Lake Basin as the research data, and reconstructed all the pixels EVI time series in the study area. The phenological parameters extraction with the moving average method, dynamic threshold method, the greatest change slope method, the Logistic function fitting method had been made. With comparing and analyzing the four kinds of extracting phenological parameters, phenological parameters were extracted for 2005 and 2015 in Dongting Lake Watershed of EVI time series of remote sensing images by moving average method, and then we analyzed the changes of the phenological characteristic parameters in the study area in the past 10 years. The results showed that the moving average method was well used to extract the phenology parameters in the Dongting Lake basin, and the results were more accurate. And the remote sensing method could be used to monitor the change of vegetation phenology in large scale in a timely and accurate manner, which provided an important reference for the study of regional climate change.