Soil Macronutrients Analysis of Vaijapur Taluka from Spectral Signatures Using Hyperspectral Non-Imaging Data

As we know soil is a vital part of the environment, survival and humanistic welfare .Soil is formed by minerals, organic matter and living organism components. From the point of view of agriculture, the soil offers support to plants and acts as a reservoir of water and nutrients Soil physical, biological and chemical properties plays a very crucial role in agricultural sector. Accelerated low cost and predictable assessment of soil quality under agricultural management is necessary to accomplish convenient observation of the effects of various management practices on soil conditions to avoid soil degradation and ensure feasible soil productivity and also soil security for improvement of agriculture sector . The objective of this study is to find soil macronutrient contents from top surface (0-20cm) of agricultural soil. FieldSpec4 Spectroradiometer (350-2500 nm) is a rapid, non-destructive, and cost-effective, time consuming tool for predicting the soil properties such as soil texture, water, carbon, nitrogen, phosphorus to better crop production .We also analyze soil samples using traditional laboratory methods .To achieve these objectives some preprocessing techniques and several statistical procedures were implemented using partial least squares-regression (PLSR).Partial least square regression gives us results for successfully estimation of soil characteristics such as Soil Moisture (RMSE :0.00097 R-square :0.99780 calibration and validation RMSE :0.0011 R-square 0.9971), For Nitrogen(RMSE :0.0012 R-square :0.9974 calibration and validation RMSE :0.0013 R-square 0.9972) Carbon(RMSE :0.0017 R-square :0.9943 calibration and validation RMSE :0.0019 R-square 0.9928), Phosphorus(RMSE :0.0015 R-square :0.9953 calibration and validation RMSE :0.0016 R-square 0.9947).All preprocessing, statistical analysis and classification have done using R-tool, ViewSpec and Uncrambler 10.5 software