The use of spatial-temporal analysis for noise reduction in MODIS NDVI time series data

Time series of satellite data can be employed for mapping the development of vegetation in space and time. However, noise induced by cloud contamination and atmospheric variability affects data quality. Science Datasets is an integral part of the MODIS Land production chain that focuses on evaluating and documenting the scientific quality of products. This study aims at the reconstruction of time series of MODIS NDVI data based on the reliability of the science data sets and on a spatial-temporal analysis of the low quality pixels. The MOD13Q1 product was analyzed over a period of one year. After identifying the pixel with the lowest guarantee of quality, it is estimated by regression analysis among neighboring pixels classified as high-quality. The combination of the per-pixel quality and spatial-temporal information is a promising method for reconstructing high-quality MODIS NDVI time series.