Fodder crops assessment using multi-temporal Landsat-8 data by NDVI based classification in Haryana state of India

Abstract

This study highlights the application of multi-temporal Landsat-8 imageries to identify and discriminate fodder crops from food crops and estimate the area utilized for fodder cultivation in three districts of Haryana during 2016 in Rabi season. Atmospherically, corrected NDVI based spectral-temporal profiles showed that the time of sowing plays an important role to differentiate fodder crops from other crops. ISODATA unsupervised classification approach was used to image classification. Accuracy assessment was carried out between four classes (fodder crops, plantation, forest and other crops) and 91.49% as overall accuracy of classification was observed. The total area under fodder cultivation was estimated as 6.37 per thousand hectares (ha) with 3.60 per thousand hectares in Kurukshetra, 1.94 per thousand hectares in Ambala and 0.83 per thousand hectares in Yamunanagar. Thus it covered an area of approximately 4855 square kilometers which included Ambala, Krukshetra and Yamunanagar districts of Haryana.