Brian L. Tolk James Vogelmann Michelle L. Knuppe Zhiliang Zhu Chengquan Huang 2003 <p>Temporal greenness matrics have been found useful for characterizing vegetation phenology, and have been used to discriminate vegetation cover types and to estimate key vegetation attributes including percent cover and green biomass. So far, however, such matrics have been calculated only from coarse resolution satellite data. Intermediate spatial resolution satellites like Landsat cannot provide the temporal resolutions needed for directly calculating such greenness matrics. In this study, we developed a method to indirectly derive annual integrated NDVI at 30 m spatial resolution using 250 m MODIS data and 30 m Landsat ETM+ imagery. Results showed that more than 90% of the variance of the annual integrated NDVI calculated using one full year’s MODIS data could be explained using as few as 3 appropriately selected observations, demonstrating the feasibility of indirectly estimating the annual integrated NDVI at intermediate spatial resolutions, as normally only limited number of useful observations would be available within the life cycle of a typical project at such spatial resolutions. The developed method was applied to two ETM+ paths/rows, for each of which 3 ETM+ images were acquired in roughly spring, summer and fall/winter seasons around the year 2000. Of the total variance of the MODIS annual integrated NDVI, 81% was explained by the three ETM+ images for one path/row and 74% for the other. </p> application/pdf en American Society for Photogrammetry and Remote Sensing Deriving annual integrated NDVI greenness at 30 m spatial resolution text