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.

Abstract

The most recent report of the Intergovernmental Panel on Climate Change (IPPC) presented medium confidence in projected increases of precipitation and run-off in some regions, while in others these are projected to decrease. However, the intensification of the climate is more generally projected to increase, which may lead to larger floods. Other studies have assessed the projected precipitation and have agreed with the IPPC assessment. These changes in climate present a concern about the risk of failure of large dams, given that most of these structures were built during the second half of the 20th century with different methods and often with limited hydrological data, resulting in uncertainty on how the design discharge capacity of the spillways can provide sufficient safety levels against overtopping in both the current and future climate. This study analyses the impacts of climate change on the failure of dams due to overtopping at a global scale The analysis is based on current and projected hydrological data obtained from the a PCR-GLOBWB Global Hydrological Model and five Global Climate Models (GCMs) selected from CMIP5 (Climate Model Intercomparison Project, phase 5). In this research an analysis of the design flood for a sample of about 1400 dams across the world under current, historical and future scenarios of climate change is made, and compared to the original design flood, by building a synthetic spillway design (using re-analysis data from CRU TS 3.2 and ERA-Interim datasets). Results from this study show that changes can be expected for the spillway discharge capacity. A consistent trend of increasing difference of the spillway capacity between the RCP4.5 and RCP8.5 scenarios and the baseline runs shows that there is a direct impact of climate change on the increase of dam failures rates due to overtopping. East Asia, South Asia, Central North America and Western North America are the regions facing the biggest rise in spillway discharge.