Remote Sensing Applications for Aromatic Basmati Rice Discrimination and Mapping

Multispectral data from aerial and satellite platforms, equipped with optical and microwave sensors, have been utilised for mapping and tracking rice cultivation in different regions of the world. However, very few studies have been conducted to map aromatic Basmati rice grown in a specific geo-environmental region of India and Pakistan. Two case studies on Basmati rice mapping using different methodologies have been presented. The first operational study, carried out from 2007 to 2012, utilised optical sensor data from Resourcesat-1 for mapping Basmati rice in the entire Basmati-growing regions of India. Data analysis using supervised maximum likelihood classification discriminated between traditional and evolved Basmati varieties from other high-yielding (HYV) varieties due to their phenological stage differences. An increasing trend in the Basmati rice area in India was observed due to its profitability to farmers. The second study used an integration of Sentinel-2 optical and Sentinel-1 microwave SAR data in two districts of Haryana during the kharif season of 2023, leveraging a Random Forest (RF) algorithm implemented within the Google Earth Engine (GEE) platform. The fortnightly composited images of MSI reflectance bands, red-edge and green chlorophyll indices, and SAR VH-VV difference at different phenological growth stages were analysed, which showed significant differences between long- and short-duration Basmati, early Basmati, and HYV varieties. A significant improvement in classification accuracy was observed. These studies also showed that traditional Basmati varieties are now grown only in a few districts. Among the evolved Basmati, long-duration varieties are being replaced by shorter-duration varieties, mainly due to their lower irrigation requirements and post-rice crop diversification possibilities.