SPATIAL PREDICTION OF FLOOD IN KUALA LUMPUR CITY OF MALAYSIA USING LOGISTIC REGRESSION

Abstract. Flooding is one of the most prevalent natural disasters affecting people worldwide. a devastating disaster in Malaysia regarding number affected, socioeconomic damage, severity, and scale impact. Urban flooding currently major concern due to possible consequences frequency with which it occurs urban areas as urbanization population increase. Due paved surfaces, roads, high population, buildings that prevent water infiltration movement nearby river, floods pose significant threat sustainability lives properties city. The recent Kuala Lumpur December 2021 January 2022 affected many buildings, infrastructure, lives. As result, this city needs model susceptibility flood-prone for an early warning system against future flood hazards Lumpur. This because can never be eradicated but minimized managed. Therefore, study integrates geospatial technology statistical (logistic regression) assess Ten conditioning factors such altitude, slope, TWI, drainage density, distance LULC, NDVI, NDWI, rainfall MNDWI were used predict susceptible flood. prediction shows overall accuracy 0.84, precision 0.91, recall 0.72, F1-score 0.80. Distance MNDWI, LULC are critical variables showed significance prediction. Thus, stakeholders should prioritize planning increase avoid effects.