Predicting yield of lahi (Brassica campestris var. toria) crop using remote sensing in tarai region of uttarakhand

Abstract

Present study was conducted in Uttarakhand state during rabi season for the year 2009-10 in which spectro-meteorological models were developed for predicting the yield of Lahi (Brassica campestris var. toria) crop. Crop management data for Lahi crop and cloud free LANDSAT-ETM+ images of path 145 and row 40 (containing Pantnagar and adjoining region) of the year 2004-05 to 2008-09 were used for the study. Spectro-meteorological models have been developed using NDVI and meteorological data pertaining to some significant crop phenological stages. Eight spectro-meteorological models were developed for Lahi crop using combinations of deviations in NDVI from normal growth curve and weekly weather variables at different crop growth stages. These models expressed the relationship between yield and weather/spectral parameters. The observed yield ranged from 9.65 q ha-1to 20.67 q ha-1and the yield predicted by these 8 models varied from 9.70 q ha-1to 21.07 q ha-1. The highest value of co-efficient of determination (R2 = 0.977) by model 8 shows that NDVI in combination with weather parameters can be used to develop location specific spectro-meteorological models, which can provide better pre-harvest yield forecast.