Change detection studies in coastal zone features of Goa, India by remote sensing

Abstract

Digital remote sensing data of SPOT-1 (Nov. 1990), IRS-1C (Jan. 2001) and IRS-1D (Jan. 2003) have been subject to maximum likelihood classifier (MLC) to carry out change detection studies in the coastal zone of Goa. The classified images were evaluated on both homogeneous and heterogeneous regions in terms of confusion matrix as well as by field validation. The classification results of these multi-temporal data reveal that MLC gives an average accuracy of 85.69, 77.66, and 88.43% for 1990, 2001 and 2003 images respectively, and the corresponding kappa coefficients are 0.89, 0.84 and 0.86 respectively. Results of MLC indicate that there is an increase in urban land, vegetation and water body reduction in barren land and sandy beach along the coastal zone of Goa. The urbanization is attributed to tourism boom-related activities.