Procedia Computer Science · 2019 · 85 citations · 6 references
Natural ResourcesHydrologyWater Resources ManagementHydrologic ForecastingHydrologic ModelingMachine LearningFlood Early WarningFlood ForecastingHydrologic Time Series AnalysisReal-time ForecastingUrban Flood ManagementBayesian Linear RegressionData-driven Hydrologic ForecastingFlood Early Warning SystemsMachine Learning-based Hydrologic ForecastingPredictionRainfall-runoff ForecastingReal-time Flood PredictionWater Level Forecasting
Urban flooding is a major problem in Thailand. An essential countermeasure towards better flooding management is to forecast flood water levels in the real-time manner. Most existing early warning systems (EWS) in Thailand contain a lot of miscalculations when they face with real situations. Towards prediction improvement, this paper presents hydrological modeling augmented with alternative five machine learning techniques; linear regression, neural network regression, Bayesian linear regression and boosted decision tree regression. As the testbed system, the so-called MIKE-11 hydrologic forecasting model, developed by Danish Hydraulic Institute (DHI), Denmark, is used. To test error reduction in runoff forecasting, the water-level records during 2012-2016 data are used for training and the derived model is tested on the record of 2017, in the experiments.
6
Charalampos Doulgeris, Pantazis Georgiou, Dimitris Papamichail · Journal of Environmental Management · 2011 · 102 citations
Natural Resources, Hydrology, Water Resources Management +15