A Machine Learning-Based Surrogate Model for the Identification of Risk Zones Due to Off-Stream Reservoir Failure

Abstract

Approximately 70,000 Spanish off-stream reservoirs, many of them irrigation ponds, need to be evaluated in terms of their potential hazard to comply with the new national Regulation of the Hydraulic Public Domain. This requires a great engineering effort to evaluate different scenarios with two-dimensional hydraulic models, for which many owners lack the necessary resources. This work presents a simplified methodology based on machine learning to identify risk zones at any point in the vicinity of an off-stream reservoir without the need to elaborate and run full two-dimensional hydraulic models. A predictive model based on random forest was created from datasets including the results of synthetic cases computed with an automatic tool based on the two-dimensional numerical software Iber. Once fitted, the model provided an estimate on the potential hazard considering the physical characteristics of the structure, the surrounding terrain and the vulnerable locations. Two approaches were compared for balancing the dataset: the synthetic minority oversampling and the random undersampling. Results from the random forest model adjusted with the random undersampling technique showed to be useful for the estimation of risk zones. On a real application test the simplified method achieved 91% accuracy.

ACROPOLIS: A Graphical User Interface for Classification of Risk for Off-Stream Reservoirs Using Machine Learning

“…Therefore, the classification process demands significant resources, including financial means, engineering expertise, and time, which are often lacking for owners. To address this challenge, Silva-Cancino et al (2022) [7] developed a machine learning (ML) algorithm that automates the risk identification process of the AoI. This solution eliminates the need to develop a two-dimensional hydraulic model, as suggested by the guidelines, offering a streamlined and efficient alternative.…”

“…In this paper, we introduce ACROPOLIS, a user-friendly software designed to classify off-stream reservoirs following Spanish regulations [2], using the ML algorithm developed by Silva-Cancino et al (2022) [7] to determine the hazard classification (A, B, or C) based on the provided input data. This paper is divided as follows.…”

“…The development of the ML-based approach for risk identification zones for off-stream reservoirs was presented in international conferences [23,24]. The tool was also presented to private and public dam owners, which provided positive feedback.…”

A Probabilistic Approach for Off-Stream Reservoir Failure Flood Hazard Assessment

“…That is, for example, the use of machine learning (ML), a common technique used to predict the behaviour of structures [55,56], morphodyncamic evolution [57], hydrological purposes [58], the flood extent [59], and the peak discharge due to breaching of embankment dams [60], among others. Silva-Cancino et al [61] presented a methodology based on ML to identify risk zones at any point in the vicinity of an off-stream reservoir. A ML-based surrogate model was trained with data obtained from 1200 two-dimensional hydraulic synthetic cases.…”