Improving Automatic Calibration of Computationally Intensive Groundwater Models
This research aims to develop advanced but practical methodologies for automatic calibration of computationally intensive groundwater simulation models. Such models are largely used in many consulting engineering companies, but they may not be completely calibrated because of the time limits and their very large computational demands. The methodologies developed during this internship increase the computational efficiency of automatic calibration algorithms by using domain specific knowledge in groundwater modelling. The general frameworks for these methodologies have been already developed in the intern’s PhD thesis and tested across different automatic calibration practices on other models such as surface water models. This internship benefits the model development practice in the partner organization as the modellers are provided with advanced tools and techniques to efficiently make use of their available computational budgets when developing and calibrating their models. The computational efficiency gains enable the modellers to develop more accurate groundwater models of the real-world…..TOBECONTINUED
View Full Project DescriptionBryan Tolson
Matrix Solutions Inc (Guelph, ON)
Engineering
University of Waterloo
Accelerate