Publication: Real-Time Flood Inundation Map Generation Using Decision Tree Machine Learning Method: Case Study of Kelantan River Basins
Date
2023
Authors
Sidek L.M.
Basri H.
Marufuzzaman M.
Deros A.M.
Osman S.
Hassan F.A.
Journal Title
Journal ISSN
Volume Title
Publisher
Springer Science and Business Media Deutschland GmbH
Abstract
Spatial investigation and rainfall affect the most in runoff and flood modelling compared to other influencing flood sources. Therefore, flash flood and short-term flood prediction require numerical rainfall estimation, which employs falls, mudflow, melted ice, etc. To forecast unexpected flood occurrences, faster flood prediction necessitates computational prediction models such as Machine Learning (ML) algorithms, which are extensively utilized around the world. So, in this research, a real-time flood inundation map (FIM) is used to develop the ML-based visualization (ML_V) method to characterize the collected dataset. Additionally, to predict the flood depth, a trained Decision Tree (DT)-based sorting algorithm is used in this method. This DT-based model takes 4 random rainfall data to train and predict the flood depth of the study area
the Kelantan River basin in Malaysia, which needs further processing for ML_V. The results showed that the precision of the forecasted map was around 80% which is compared with another Gaussian Na�ve Bias ML algorithm. However, in imbrication with the ArcGIS maps, the forecasted map detached the �out of boundary� images and generated a clearer map. It is obvious that this ML_V model was anticipated to read the final output, which is involved during the flood guidance statement (FGS) to broadcast the data to the community. � The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023.
the Kelantan River basin in Malaysia, which needs further processing for ML_V. The results showed that the precision of the forecasted map was around 80% which is compared with another Gaussian Na�ve Bias ML algorithm. However, in imbrication with the ArcGIS maps, the forecasted map detached the �out of boundary� images and generated a clearer map. It is obvious that this ML_V model was anticipated to read the final output, which is involved during the flood guidance statement (FGS) to broadcast the data to the community. � The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023.
Description
Keywords
Decision tree , Flood forecasting , Flood hazard map , Flood impacts , Kelantan , Machine learning , Risk assessment