Based on data from the Bengkayang Regency Regional Disaster Management Agency (BPBD), in early 2025, 11 sub-districts were affected by flooding, with 12,023 affected people and 3,468 homes submerged. Efforts to minimize the impact of this flood disaster require an effective data-driven monitoring system. A floodwater monitoring system in Bengkayang Regency is essential for effective disaster management, reducing losses and damage, and providing early warnings to the surrounding community. One approach that can be used is remote sensing technology, which can be a solution, especially when combined with machine learning algorithms that can accelerate and improve the accuracy of data analysis. One such machine learning algorithm is the Support Vector Machine (SVM) algorithm. This study has produced a final dataset of five variables: rainfall, slope gradient, land use, VV, and NDWI. This dataset is used for the classification process using the Support Vector Machine algorithm. After preprocessing and dividing the training data by 75% and the test data by 25% of the total 512 data sets. The image classification results using SVM demonstrated quite good performance. The resulting accuracy was 80%, with precision and recall values ranging from 0.67 to 0.98. Based on these results, the model demonstrated excellent ability to identify waterlogging points. The classification results were then integrated into a web-based geographic information system that displays an interactive map of the distribution of waterlogging points.
Cahyaningtyas et al. (Wed,) studied this question.