The two components of the study reflect assessment and change analysis of a tropical wetland in Sri Lanka. The first section explains spatial classification using pixel level-disaggregated image analysis and refined aggregated image analysis and comparison of information extracted by all methods to analyse a better classifier. The second section illustrates change analysis calibrating the land change modeller (LCM) [IDRISI-Andes]. Key observations: a) visual interpretation provides comprehensive blueprint of the wetlandscape compared to supervised and unsupervised classifiers b) change in landscape pattern reflect substantial transition in wetland use. Validation using field coordinates and socioeconomic data showed kappa value (%) of 87.
|Number of pages||18|
|Journal||Italian Journal of Remote Sensing / Rivista Italiana di Telerilevamento|
|Publication status||Published - Jun 2012|