Two-dimensional hydrodynamic models numerically solve full Shallow Water Equations (SWEs). Despite their high accuracy, these models have long simulation run times and therefore are of limited use for exploratory or real-time flood predictions. We investigated the possibility of improving flood modelling speed using Machine Learning (ML). We propose a new method that replaces the computationally expensive parts of the hydrodynamic models with simple and efficient data-driven approximations. Our hypothesis is that by integrating ML with physics-based numerical methods, we can achieve improved generalization performance: that is, the trained model for one case study can be used in other studies without the need for new training. We tested two ML approaches: for the first, we integrated curve fitting, and, for the second, artificial neural networks (ANN) with a finite volume scheme to solve the local inertial approximation of the SWEs. The data-driven models approximated the Momentum Equation, which explicitly solved the time derivative of flow rates. Water depths were then updated by applying a water balance equation. We also tested two different training datasets: the simulated dataset, generated from the results of hydrodynamic model, and the random dataset, generated by directly solving the momentum equation on randomly sampled input data. Various combinations of input features, for example, water slope and depth, were explored. The proposed models were trained in a small hypothetical case and tested in a different hypothetical and in two real case studies. Results showed that the curve-fitting method can be implemented successfully, given sufficient training and input data. The ANN model trained with a random dataset was substantially more accurate than that of the model trained with the simulated dataset. However, it was not successful in the real case studies. The curve-fitting method resulted in better generalization performance and increased the simulation speed of the local inertial model by 23%. Future research should test the performance of ML in terms of an increase in stable time step size and approximation of the full SWEs. 相似文献
Geotechnical and Geological Engineering - Soil–reinforcement interaction is a major factor in the analysis and design of reinforced earth structures. In current research the effects of... 相似文献
Natural Hazards - Satellite based thermal anomaly occurs as a substantial precursor for strong earthquakes, as the need for earthquake precursor detection has very important for impending main... 相似文献
Geotechnical and Geological Engineering - In this study, peak particle velocity (PPV) values for driving three piles with diameters of 40 cm, 50 cm, and 70 cm in a clayey... 相似文献
Polycyclic aromatic hydrocarbons (PAHs) are types of hazardous contaminants, which their ingestion could cause severe consequences on human health. Leakages from storage tanks, underground pipelines, and evaporation ponds are the main sources of soil and groundwater contaminations at the Tehran Oil Refinery area (TOR site), located in south of Tehran, Iran. In this study, soil samples were collected from different locations at and adjacent to a polluted stream in the south of the refinery. The samples were analyzed for two hazardous PAH compounds, namely benzo[a]anthracene and acenaphthene. The clean up levels due to the accidental ingestion of contaminated soils at the site were also investigated in accordance to the U.S.EPA guidelines. Comparing the soil analysis results indicated that the benzo[a]anthracene concentrations in the samples varied from 53 to 299 mg/kg, which were higher than the clean up level of 1.17 mg/kg. Thus, soil remediation is required for this contaminant. The acenaphthene analysis results denoted that the average concentration of this contaminant was below the clean up level of 116.67 mg/kg, indicating that no treatment for this contaminant is necessary at the TOR site. Also, because the slope of the ground extends to the south of the stream, which stimulates the migration of the contaminants in this direction due to advection and dispersion mechanisms, the average of benzo[a]anthracene concentrations in south samples was higher than north samples (i.e., Cavg(S) = 160 ppm, Cavg(N) = 113 ppm). Various treatment techniques such as thermal desorption, soil vapor extraction (SVE), and solidification/stabilization (S/S) were investigated for this site. Due to moderate to high plasticity and relatively low permeability of the soil and low volatility of benzo[a]anthracene, however S/S method is recommended as a practical approach for the remediation of the soil at the site. 相似文献
Supervised injection facilities (SIFs) are medical facilities where injection drug users can inject their illicit drugs under the supervision of nurses and doctors. Currently, there is only one legal SIF in operation in North America and it has been operating in Vancouver, British Columbia for over a decade. The purpose of this study is to determine whether the current facility needs to be expanded to other locations in British Columbia, Canada. We employ mathematical modeling to estimate the number of new human immunodeficiency virus (HIV) and hepatitis C infections prevented based on the available secondary data. Additionally, we also estimate the number of prevented overdose deaths attributable to the SIF. With very conservative estimates, it is predicted that establishing two SIFs locations outside Vancouver in British Columbia’s capital city, Victoria, is cost-effective, with a benefit-cost ratio of 1.25:1. It appears that expanding SIFs to Victoria could offer significant savings for local health care institutions. 相似文献
A new closure and a modified detrainment for the simplified Arakawa–Schubert (SAS) cumulus parameterization scheme are proposed. In the modified convective scheme which is named as King Abdulaziz University (KAU) scheme, the closure depends on both the buoyancy force and the environment mean relative humidity. A lateral entrainment rate varying with environment relative humidity is proposed and tends to suppress convection in a dry atmosphere. The detrainment rate also varies with environment relative humidity. The KAU scheme has been tested in a single column model (SCM) and implemented in a coupled global climate model (CGCM). Increased coupling between environment and clouds in the KAU scheme results in improved sensitivity of the depth and strength of convection to environmental humidity compared to the original SAS scheme. The new scheme improves precipitation simulation with better representations of moisture and temperature especially during suppressed convection periods. The KAU scheme implemented in the Seoul National University (SNU) CGCM shows improved precipitation over the tropics. The simulated precipitation pattern over the Arabian Peninsula and Northeast African region is also improved.