An Improved Method for Producing High Spatial-Resolution NDVI Time Series Datasets with Multi-Temporal MODIS NDVI Data and Landsat TM/ETM+ Images

摘要: Due to technical limitations, it is impossible have high resolution in both spatial and temporal dimensions for current NDVI datasets. Therefore, several methods are developed produce (spatial temporal) time-series datasets, which face some limitations including computation loads unreasonable assumptions. In this study, an unmixing-based method, Linear Mixing Growth Model (NDVI-LMGM), proposed achieve the goal of accurately efficiently blending MODIS data multi-temporal Landsat TM/ETM+ images. This method firstly unmixes changes different land cover types then uses unmixed predict Landsat-like dataset. The test over a forest site shows accuracy (average difference: −0.0070; average absolute 0.0228; relative 4.02%) efficiency NDVI-LMGM (31 seconds using personal computer). Experiments more complex landscape long-term demonstrated that performs well each stage vegetation growing season robust regions with contrasting variations. Comparisons between (i.e., Spatial Temporal Adaptive Reflectance Fusion (STARFM), Enhanced STARFM (ESTARFM) Weighted (WLM)) show accurate efficient than methods. will benefit surface process research, requires dense dataset resolution.