The Utilization of MODIS and Landsat TM/ETM + for Cotton fractional yield estimation in Burewala

Abstract

Estimates of crop yield are desirable for managing agricultural lands. Remote sensing is the one technology that can give an unbiased view of large areas, with spatially explicit information distribution and time repetition, and has thus been widely used to estimate crop yield and offers great potential for monitoring production, yet the uncertainties associated with large-scale crop yield estimates are rarely addressed. In this study, we tried to estimate cotton cropped area using the supervised classification; planting dates for 11 years (1998 to 2009) of Landsat imagery, and fractional yield using MODIS (Terra) the normalized difference vegetation index (NDVI), and enhanced vegetation index (EVI) in an intensive agricultural region of Burewala, Punjab province of Pakistan. Vegetation indices are widely used for assessing and monitoring ecological variables such as vegetation cover and above-ground biomass. Monitoring the spatial distribution of cotton yield helps identifying sites with yield constraints. The newly available satellite images from the MODIS sensor provide enhanced atmospheric correction, cloud detection, improved geo-referencing, comprehensive data quality control and the enhanced ability to monitor vegetation development. The high temporal resolution of the MODIS datasets can provide an efficient and consistent way for biomass and fractional yield monitoring and assessment. The reflected radiation provides an indication of the type and density of canopy. The condition, distribution, structure and the development of the vegetation through the phenological stages can affect the relation between yield and NDVI. The high spatial resolution Landsat images were applied to extract the area under cotton cultivation within the landscape and to determine the cotton fraction among other land uses within the coarse spatial resolution MODIS pixels.