ASSESSMENT OF SOME CROP YIELD MODELS USING GEOMATICS TOOLS

Abstract

in the life of an economy agriculture plays a crucial role. It is considered as the backbone of our economic system. It provides raw material, food and also provides employment opportunities to a very large proportion of population.The countries like India which needsaccurate, reliable and timely information on different types of crops grown and their acreages. So monitoring of each and every crop condition is important for the economic development of our nation. For this remote sensing has proved itself to be very efficient in monitoring the growth of agricultural crops and in irrigation scheduling. So keeping all this in mind the main aim of this study, is to develop various models which can easily predict the crop yield and to compare those models to decide which model is providing more precise results under given conditions.For this IRS IC LISS 111 sensor data for last 12 years (2000-2012) is used of flowering stage, last 12 years meteorological data such as max temp, rnin temp, sunshine, rainfall and agro-meteorological indices such as Growing Degree Days (GDD), Temperature Difference (TD) and Helio Thermal Unit (HTU) has been used of Ludhiana district which is collected from School of Climate Change & Agricultural Meteorology. Maximum Likely hood supervised classification is used for obtaining the wheat class from the multispectral image for calculating the acreage. The Acreage estimated using the LISS III data is around 248 thousand hectare which is about 3% deviated when compared with the actual acreage data which was published in Statistical Abstract of Punjab. Actual wheat yield data is retrieved from statistical abstract of Punjab and NDVI is computed for the pixels whose location is obtained using GPS. Now using multi-variate regression model different wheat yield models are developed using meteorological, agrometeorological and remote sensing data such as spectral yield model, agro-meteorological model, agro-meteorological trend yield model and agro-meteorological spectral yield model. With the help of these models wheat yield is predicted and to validate the predicted yield it is compared with the actual wheat yield data and relative deviation is calculated. The result shows that the yield predicted using agrometeorological spectral model provides more precise result 0 ASSESSMENT OF SOME CROP YIELD MODELS USING GEOMA TICS TOOLS which is 1% deviated from the actual yield as compared to other models as this model has an advantage of using spectral data in prediction. With the help of this study, it is found that the predicted yield of wheat yield for Ludhiana district for 20 12-13 is 4780.68 kg/ hectare using agro-met trend yield model.