 This study presents a noise mapping method that uses four state flag parameters to estimate cloud possibility index, CPI, for noise elimination in moderate-resolution imaging spectroradiometer, MODIS, time series normalized differential vegetation index, NDVI, products over Sri Lanka. The study shows that CPI-based noise elimination increases the detection accuracy of number of growing cycles from 45 to 60 percent to 85 to 95 percent in vegetated regions and cautions against using time series NDVI data without proper cloud correction mechanism for accurate estimation of spatial distribution and intensity of drought. This article was authored by Saptashi Mondal, Chakalingam Jeganathan, Duarajamanoth, and others.