A simplified data assimilation method for reconstructing time-series MODIS NDVI data

Abstract

The Normalized Difference Vegetation Index (NDVI) is an important vegetation index, widely applied in research on global environmental and climatic change. However, noise induced by cloud contamination and atmospheric variability impedes the analysis and application of NDVI data. In this work, a simplified data assimilation method is proposed to reconstruct high-quality time-series MODIS NDVI data. We extracted 16-Day L3 Global 1 km SIN Grid NDVI data sets for western China from MODIS vegetation index (VI) products (MOD13A2) for the period 2003–2006. NDVI data in the first three years (2003–2005) were used to generate the background field of NDVI based on a simple three-point smoothing technique, which captures annual features of vegetation change. NDVI data for 2006 were used to test our method. For every time step, the quality assurance (QA) flags of the MODIS VI products were adopted to empirically determine the weight between the background field and NDVI observations. Ultimately, more reliable NDVI data can be produced. The results indicate that the newly developed method is robust and effective in reconstructing high-quality MODIS NDVI time-series.