A Real-Time Flood Forecasting Hybrid Machine Learning Hydrological Model for Krong H’Nang Hydropower Reservoir

EasyChair Preprint 9393

Abstract

Flood forecasting is critical for mitigating flood damage and ensuring the safe operation of hydroelectric power plants and reservoirs. In this paper, the authors present a hybrid machine-learning hydrological model to enhance the accuracy of real-time flood forecasting. This model is developed based on the combination of the HEC-HMS hydrological model and an Encoder-Decoder-Long Short-Term Memory network. The proposed hybrid model has been applied to the Krong H’nang hydropower reservoir. The observed data from 33 floods monitored between 2016 and 2021 are used to calibrate, validate, and test the hybrid model. Results show that the HEC-HMS-ANN hybrid model significantly improves the forecast quality, especially for long forecasting time steps. The KGE efficiency index, for example, increased from ∆KGE = 16% at time t + 1 to ∆KGE = 69% at time t + 6 hours, similar to other indicators (such as peak error and volume error). The computer program developed for this study is being used at the KrongHnang hydropower to aid in reservoir planning, flood control, and water resource efficiency.

Keyphrases: HEC-HMS, Hydrological hybrid model, KrongH'nang, machine learning, real-time flood forecasting.

@booklet{EasyChair:9393,
  author    = {Phuoc Sinh Nguyen and Truong Huy Nguyen and The Hung Nguyen},
  title     = {A Real-Time Flood Forecasting Hybrid Machine Learning Hydrological Model for Krong H’Nang Hydropower Reservoir},
  howpublished = {EasyChair Preprint 9393},
  year      = {EasyChair, 2022}}