River body extraction and classification using enhanced models of modified normalized water difference index at Yeh Unda River Bali

Abstract

Subak is a Balinese social organization that regulates irrigation systems in traditional ways. Monitoring of river flow using remote sensing technology has been expected to determine new irrigation channels. In this study, we conducted water body extraction using remote sensing technology based on water indices at Yeh Unda river, Bali. Landsat 8 OLI satellite images were used as the main dataset. We used Enhanced Modified Normalized Difference Water Index (EMNDWI) based on Modified Normalized Difference Water Index (MNDWI) that has been introduced as the conventional method to extract water bodies. Experimental results from the proposed method was compared to the results of MNDWI method. The proposed method was able to detect narrow river with brown water color, although there is a bit of missed detection caused by distraction on surrounding areas like trees, bushes and shadow. The results of the extraction at the selected area show that EMNDWI is a viable method for segmentation of river with brown water colors and narrow bodies. The results of water body extraction may be used by Subak officers to build a new water management system traversed by the Yeh Unda river and to analyze overflowing in the river.