Water Resources Engineer Mead & Hunt, Wisconsin, United States
Abstract Submission: Engineers and scientists can strategically develop automated tools for use with 2D (and combined 1D/2D) models that help ensure effective model outputs/outcomes and eliminate questions about modeling processes. And since many hydraulic models store input data and parameters in plain text files, the possibilities for automation are vast. In this project case study, engineers built automated tools that improved the computational efficiencies of a probabilistic hydraulic analysis for a 150-square-mile study area near a flood risk management dam. The combined 1D/2D model required a large matrix of simulations, including variable dam operational scenarios, differing inflow distributions between three flooding sources, and detailed output analysis at hundreds of separate infrastructure locations. Dam operational scenarios were computed with visual basic for applications (VBA), accommodating a wide range of input variables the user could query from historical records or directly input for synthetic scenarios. Detailed model calibration required flow-varied roughness factors. This detailed calibration was approached programmatically with Python to meet project deadlines. Engineers also used Python to determine the optimal 2D computational parameters for various flow conditions and compute tens of thousands of post-modeling duration and elevation exceedance values. By learning how to build automated tools and using them to overcome modeling challenges, engineers can run analysis models more consistently and perform probabilistic analyses more efficiently.
Learning Objectives/Expected Outcome (Optional) : By learning how to build automated tools and using them to overcome modeling challenges, engineers can run analysis models more consistently and perform probabilistic analyses more efficiently.
