Detecting Land Use/Land Cover change using Landsat Imagery: Jumgal District, Kyrgyzstan

Abstract

One of the important problems of Geographical Science in the XXI century is the study of quantitative and qualitative changes observed in the land use. In this study, we present two datasets of land use/cover in the Jumgal District derived from Landsat TM and OLI/TIRS images in 1996 and 2014. We used these data to investigate changes in land use/cover between 1996 and 2014. The satellite data has been geocoded to the Universal Transverse Mercator (UTM) coordinate system, Zone 43 North, World Geodetic System (WSG-84) coordinate datum. The information provided by satellites to quantify the various parameters of land use of the district has been evaluated by applying various image-processing techniques. The images were classified using Maximum Likelihood classification method and mapped using ArcGIS 10.1. From 1996 to 2014, the water body’s area increased by 530 ha (98.1 %); agriculture area decreased by 9720 ha (14.4 %); mixed forest decreased by 7010 ha (33.3 %); wetland decreased by 1080 ha (9.5%); grassland increased by 540 ha (0.2 %); barren land increased by 18360 ha (14.3 %); and perennial snow/ice decreased by 1620 ha (16 %). The overall accuracies for 1996 and 2014 were respectively 97 % and 92 % with Kappa Statistics of 96 %, and 90 %. User and producers’ accuracies of each information class ranged from 77 % to 100 % in most cases. From these findings, it is highly recommended that the local government initiate measures to protect the forest in Jumgal district whose continuous retreat would expose the land to erosion and other related environmental concerns.