INVENTORY AND ASSESSMENT OF CORIANDER CROP IN THE STATE OF RAJASTHAN USING MULTITEMPORAL REMOTE SENSING DATA
- Mahalanobis National Crop Forecast Centre, DAC&FW, New Delhi 110 012, India
Keywords: Coriander, Multitemporal, Remote Sensing, Sentinel 2A, Inventory, Classification, Accuracy
Abstract. The present study aims to develop the methodology for inventory and assessment of coriander crop in Rajasthan using remote sensing technique. Sentinel-2A optical data having a spatial resolution of 10 m, from January–March, 2017 were considered for this study keeping in mind the crop calendar. It was found that coriander at its flowering stage gives a distinct light pink colour which helps it to differentiate from other crops. However it is difficult to separate other stages of coriander (early vegetative, mature stage) owing to its similarity in tonal pattern with mustard. The overall accuracy of single date image was found to be 63.29% and Kappa (K^) Coefficient as 0.5532. With the inclusion of multiple dates accuracy increased to 91.14% and Kappa (K^) Coefficient to 0.7436. This was because increase in information increases the possibility to separate crops from each other. This study demonstrates the feasibility of multi-temporal satellite data for accurate coriander crop mapping area estimation in multi-crop scenario with reasonable accuracy at the Block/district level and State level.