Sequential Analysis of Quetta Watershed Using Machine Learning Tool

Abstract

Abstract The changes in land use and land cover are the most considerable factors for the decision makers, which can be precisely evaluated by using Geographic Information System (GIS) and Satellite Remote Sensing (SRS) technique. The present study was carried out in the Quetta Basin Watershed for evaluating the land use change of two decades during 1999-2009 and 2009-2018, using the Landsat imageries. The main land use types of study area include water body, agriculture, vegetation, barren and built-up land area. All these land use types were mapped for each time period. The technique of hybrid classification was used for the classification of the imageries by using ISODATA and Maximum Likelihood Algorithm (MLA) in ERDAS Imagine software 2011 v, while change detection maps of different land use land cover classes was generated by using the ArcGIS 10.2 v. The finding of study indicated that widespread and significant changes has occurred in study area during two decades in which remarkable increase occur in built up and vegetation land area, on the other hand reduction was observed in water body, barren and agriculture land area from year 1999 to 2018. During the whole study period, major land shifting was observed in agriculture to vegetation and vegetation to barren. The transformation in land cover use may pose severe risk to watershed resources. Therefore, proper management, effective conservation plans and strategies are necessary to designed for the protection of this natural resource.