Abstract:Floods are natural disasters caused mainly due to heavy or excessive rainfall. They induce massive economic losses in Bangladesh every year. Physically-based flood prediction models have been used over the years where simplified forms of physical laws are used to reduce calculations' complexity. It sometimes leads to oversimplification and inaccuracy in the prediction. Moreover, a physically-based model requires intensive monitoring datasets for calibration, accurate soil properties information, and a heavy computational facility, creating an impediment for quick, economical and precise short-term prediction. Researchers have tried different approaches like empirical data-driven models, especially machine learning-based models, to offer an alternative approach to the physically-based models but focused on developing only one machine learning (ML) technique at a time (i.e., ANN, MLP, etc.). There are many other techniques, algorithms, and models in machine learning (ML) technology that have the potential to be effective and efficient in flood forecasting. In this study, five different machine learning algorithms- exponent back propagation neural network (EBPNN), multilayer perceptron (MLP), support vector regression (SVR), DT Regression (DTR), and extreme gradient boosting (XGBoost) were used to develop total 180 independent models based on a different combination of time lags for input data and lead time in forecast. Models were developed for Someshwari-Kangsa sub-watershed of Bangladesh's North Central hydrological region with 5772 km2 drainage area. It is also a data-scarce region with only three hydrological and hydro-meteorological stations for the whole sub-watershed. This region mostly suffers extreme meteorological events driven flooding. Therefore, satellite-based precipitation, temperature, relative humidity, wind speed data, and observed water level data from the Bangladesh Water Development Board (BWDB) were used as input and response variables.For comparison, the accuracy of these models was evaluated using different statistical indices - coefficient of determination, mean square error (MSE), mean absolute error (MAE), mean relative error (MRE), explained variance score and normalized centred root mean square error (NCRMSE). Developed models were ranked based on the coefficient of determination (R2) value. All the models performed well with R2 being greater than 0.85 in most cases. Further analysis of the model results showed that most of the models performed well for forecasting 24-hour lead time water level. Models developed using XGBoost algorithm outperformed other models in all metrics. Moreover, each of the algorithms' best-performed models was extended further up to 20 days lead time to generate forecasting horizon. Models demonstrated remarkable consistency in their performance with the coefficient of determination (R2) being greater than 0.70 at 20 days lead-time of forecasting horizon in most cases except the DTR-based model. For 10- and 5-days lead time of forecasting horizon, it was greater than 0.75 and 0.80 respectively, for all the model extended. This study concludes that the machine algorithm-based data-driven model can be a powerful tool for flood forecasting in data-scarce regions with excellent accuracy, quick building and running time, and economic feasibility.

A computationally efficient flash flood early warning system for a mountainous and transboundary river basin in Bangladesh

Forecasting And Providing Warnings Of Flash Floods For Ungauged Mountainous Areas Based On A Distributed Hydrological Model

Improvement of flood susceptibility mapping by introducing hybrid ensemble learning algorithms and high-resolution satellite imageries

A Dynamic Early Warning Model for Flash Floods Based on Rainfall Pattern Identification

Development of Flood Forecasting System for Someshwari-Kangsa Sub-watershed of Bangladesh-India Using Different Machine Learning Techniques

Empowering real-time flood impact assessment through the integration of machine learning and Google Earth Engine: a comprehensive approach

Urban Flood Analysis and Early Warning Solution for Disaster Resilience

Flood Hazard Assessment of the Bangshi River Floodplain using an Integrated 1D/2D Coupled Hydraulic Model

Modelling and validation of flash flood inundation in drylands

Enhancing real-time flood forecasting and warning system by integrating ensemble techniques and hydrologic model simulations

Hybrid deep learning models for multi-ahead river water level forecasting

Global prediction of extreme floods in ungauged watersheds

Integrated Flood Forecasting and Warning System against Flash Rainfall in the Small-Scaled Urban Stream

Improving operational flood forecasting in monsoon climates with bias-corrected quantitative forecasting of precipitation

Random forest classifications for landuse mapping to assess rapid flood damage using Sentinel-1 and Sentinel-2 data

Flash flood potential risk zonation mapping using GIS-based spatial multi-index model: a case study of Sunamganj District, Bangladesh

Flash Flood Simulation for Ungauged Catchments Based on the Distributed Hydrological Model

An IoT-based system for monitoring and forecasting flash floods in real-time

Enhancing flood prediction in Southern West Bengal, India using ensemble machine learning models optimized with symbiotic organisms search algorithm

Hydrological-hydrodynamic nexus for predicting flood plain inundation: a case study in Pagladiya-Putimari catchment, India

A synergic approach using the model and remote sensing data for flood monitoring in under-observed transboundary rivers