Mapping Prosopis juliflora Using Satellite Data in a Part of Indian Thar Desert

摘要

Prosopis juliflora, an invasive species, is capable of growing in a wide variety of soils and climatic conditions. By aggressive encroachment, it degrades ecosystem by reducing species diversity. An attempt has been made to map the invasive plant in Indian desert district - Jodhpur using satellite data. Hyperspectral, high spatial resolution data were used to study the spectral separability of P. juliflora with other five prominent native species viz., Acacia Senegal, Calotropis procera, Euphorbia cauducifolia, Ziziphus mauritiana and Capparis decidua. Multispectral satellite data corresponding to SPOT-Multispectral (SPOT-XS) and IRS-Linear Imaging Self-scanning Sensor-3 (IRS-LISS3) were synthesized from hyperspectral data of the species obtained in natural conditions. Vegetation indices (VIs) and texture measures were derived from these data. Based on overall accuracy of classification, the best combination of spectral and feature bands were chosen. The best data combination and the methodology were employed over the XS bands of SPOT (1987-89) data and LISS3 bands of IRS (2003) data. The overall accuracy of 86.98% and 89.86% were obtained from SPOT-XS and IRS-LISS3, respectively. User's accuracy (UA) of 62.24% and 70.15% for P. juliflora class were obtained from the SPOT and IRS datasets, respectively. P. juliflora has shown aggressive encroachment and increased 61.68 sq.km area during the last fifteen years.