Evaluating the effects of climate and land use change on the future flood susceptibility in the central region of Vietnam by integrating land change modeler, machine learning methods

Abstract

The crucial importance of land cover and use changes and climate changes for worldwide sustainability results from their negative effects on flood risk. In a watershed, a particularly important research question concerning the relationship between land use and climate change and the flood risk is the subject of controversy in the literature. This study aims to assess the effects of land use and climate change on the flood susceptibility in the watershed Nhat Le–Kien Giang, Vietnam using machine learning and Land Change Modeler. The results show that Social Ski Driver Optimization (SSD), Fruit Fly Optimization (FFO), Sailfish Optimization (SFO), and Particle Swarm Optimization (PSO) successfully improve the Support Vector Machine (SVM) model's performance, with a value of the Area Under the Receiver Operating Characteristic curve (AUC) > 0.96. Among them, the SVM-FFO model was better with the value of AUC of 0.984, followed by SVM-SFO (AUC = 0.983), SVM-SSD (AUC = 0.98), SVM-PSO (AUC = 0.97), respectively. In addition, the areas with high and very high flood susceptibility in the study area increased by about 30 km2 from 2020 to 2050 with the SVM-FFO model. Our results underline the consequences of unplanned development. Thus, by applying the theoretical framework of this study, decision makers can take sound more planning measures, such as avoiding construction in areas often affected by floods, etc. Although in this study flood susceptibility is studied in a Central Coast province, the results can be applied to other rapidly developing and flood-prone provinces of Vietnam.