Mapping Land Use and Land Cover in the Upper Ciliwung Watershed Using Landsat Tree Cover (TC) Data

Abstract

<p class='IJASEITAbtract'>Land use and land cover (LULC) mapping using Landsat Tree Cover (TC) data that we employed was digital classification by converting Landsat TC raster data into Landsat TC vector data and determining LULC classes in the attribute table based on percent TC criteria (interval TC- min - TC-max). The classification was adapted from the LCCS classification and partially modified. Compared to conventional digital image classification (supervised and unsupervised classifications), our digital classification method is easier and faster because Landsat TC data does not require pre-processing and reclassification to improve classification accuracy. Landsat TC classification accuracy was assessed against the interpretation of a very high spatial resolution (VHSR) image available in Google Earth (GE). The purpose of the study was to determine the ability of Landsat TC data paired with percent TC criteria of LULC adapted from the LCCS classification and validated with VHSR in GE for mapping LULC in the tropics. This study was conducted in the Upper Ciliwung watershed, which is located in Bogor Regency, West Java Province, Indonesia. LULC mapping using Landsat TC data paired with percent TC criteria of LULC adapted from the LCCS classification and validated with VHSR in GE provided a useful tool for producing LULC map in the Upper Ciliwung watershed. This study classified LULC in the Upper Ciliwung watershed consisting of settlements, closed forests, medium forests, opened forests, mix gardens, tea plantations, shrub lands, grasslands, and rainfed croplands paddy fields, fish fonds, and bare lands with overall accuracy of 91%.