Yield estimation in farmers' fields at Alipur Block of Delhi by using crop model and satellite data

Abstract

A study was undertaken to validate the Wheat Growth Simulator (WTGROWS) in the farmers' fields of Alipur Block of Delhi and linking satellite derived vegetation index with the simulation model to estimate the wheat yield. Date of sowing, management practices and cultivars varied widely among the study sites. Leaf area index (LAI), phenological development and agronomic management (fertilizers and irrigation) were monitored at regular intervals for the 25 field sites selected in the study area. Above ground biomass and grain yield were recorded at harvest. Using the parameters derived for these sites, WTGROWS was run for each of the individual 25 sites. Crop phenology, temporal course of LAI and grain yield of each site was compared with the actual observations. The simulated and actual LAI temporal profile matched well for sites with different dates of sowing, excepting larger deviation noticed in the later stages of the crop growth. The simulated pre-anthesis duration and total above ground biomass were also correlated well with the observed values being mostly within ± 15%. There were large discrepancies in simulated and observed grain yield. A satellite image near anthesis of IRS 1D LISS-3 was acquired for the study area. The sites were identified on the image and their vegetation indices were derived. Average grey value in Infrared (IR) and Red (R) band, Ratio Vegetation Index (RVI), Soil Adjusted Vegetation Index (SAVI) andNormalized Difference Vegetation Index (NDVI) were giving significant relation with measured LAI of 5 th February which corresponded to crop anthesis stage. The relation between vegetation indices and LAI was logarithmic in nature. This logarithmic relation was incorporated into the WTGROWS to force the LAI to the equation-derived value at particular growth stage and model yield was computed and compared with actual observations.