Spatiotemporal Analysis Of Flood Inundation Using MNDWI At Rawa Pening 2019-2025

Abstract

Flood inundation around Rawa Pening Lake has intensified in recent years, yet long-term patterns of inundation extent and persistence remain poorly quantified. This study aims to analyse the interannual and spatial dynamics of flood inundation in the Rawa Pening lake-basin system during 2019–2025 and to identify riparian zones that experience recurrent flooding. Sentinel-2 Level-2A imagery was processed within a 3 km riparian buffer to compute the Modified Normalized Difference Water Index (MNDWI), apply a histogram- and visually based threshold for water extraction, and generate annual maximum inundation and seasonal wet-season inundation-frequency metrics. The classification achieved high overall accuracy (93.1%), supporting the reliability of the derived inundation products. The multiyear results reveal strong year-to-year variability in maximum inundation and show that floodwater consistently concentrates in low-lying southern and southeastern riparian sectors, particularly within Banyubiru and Tuntang subdistricts, while western and northern sectors experience smaller and less persistent flooding. The combination of maximum inundation extent and inundation-frequency maps provides complementary insight into extreme events and recurrent flood exposure, enabling a clearer distinction between episodic and chronic flood impacts on riparian agriculture and settlements. These findings demonstrate the value of multitemporal MNDWI analysis and an ecologically defined buffer zone for monitoring lake-basin flooding and offer a spatial basis for prioritizing riparian conservation, land-use regulation, and flood-risk reduction around Rawa Pening.