植被覆盖区VIIRS与MODIS遥感指数的相关性

The relationship between VIIRS and MODIS remote sensing index in vegetation coverage area

连续一致的遥感植被指数是获取地表植被动态变化的基础和前提。MODIS遥感指数具有较高时间分辨率且记录时间长,是快速、大面积获取植被信息的重要数据。然而,MODIS处于超期服役阶段,VIIRS是MODIS传感器的继承和发展,研究2种传感器同类植被指数的关系,以实现二者的联合应用,具有重要的应用价值。采用探索性数据分析方法,探究植被覆盖区VIIRS NDVI和MODIS NDVI的定量关系,结果发现:农田、林地、草地的VIIRS NDVI和MODIS NDVI均表现出显著的线性相关,相关系数最高可达0.96(P<0.01),均值在0.8以上。经验证,面向农田、林地和草地的综合模型能精确反映植被区VIIRS NDVI和MODIS NDVI的线性关系。农田、林地、草地的MODIS NDVI的时序特征和综合模型反演NDVI的时序特征具有一致性,综合模型可有效应用于两者间的转换和时序应用。

Consistent remote sensing（ RS） vegetation indices are the basic and prerequisite for obtaining the dynamic changes of surface vegetation. MODIS remote sensing indices had long recording time and the high temporal resolution. MODIS RS indices were the important data for rapid and large-scale asquisition of vegetation information. But MODIS was in the extended service phase,and VIIRS sensor was inheritance and development of MODIS sensor. Studying the relationship between VIIRS sensor and MODIS sensor on the same vegetation index to achieve the joint application,which had important application value. The exploratory data analysis methods were used to explore the quantitative relationship between VIIRS NDVI and MODIS NDVI in vegetation coverage area. The result showed that there was significant linear correlation between VIIRS NDVI and MODIS NDVI in farmland,forestland and grassland. The correlation coefficient was as high as 0. 96（ P0. 01）,and the mean value was above 0. 8. The comprehensive model for farmland,forestland and grassland could accurately reflect the linear relationship between VIIRS NDVI and MODIS NDVI in the vegetation area. The time series characteristics of MODIS NDVI in farmland,forestland and grassland were consistent with the invelsion characteristics of NDVI in comprehensive model. So the comprehensive model can be used for timing application and NDVI converting between VIIRS and MODIS effectively.