A new hydrologic model evaluation framework

点击翻译按钮获取中文摘要

The ability to make sound, science-based, decisions in response to a changing climate and changing landscape is critically important for protecting Canada's water resources. One means we have to support planning in response to the unknown impacts of these changes is via the use of mathematical simulation models, which help support decision-making by predicting flows and water levels, determining recharge to groundwater, estimating water availability, and/or supporting water quality investigations. Although such hydrologic models of water resources systems are regularly used to inform policy at various levels of government and help resource managers make important operating decisions, the current methods used to evaluate the suitability of these models to support decision-making are insufficient, particularly in a rapidly changing world where there is no guarantee the future is similar to the past. The current model evaluation or testing procedures used to evaluate the utility of a model, (i.e., model validation) is currently defined to look backwards by comparing model predictions to past data and thus unable to properly assess whether a given hydrological model is actually useful for its intended purpose. This research program will identify fundamentally different, forward-looking, decision-oriented hydrologic model testing/evaluation procedures. New methods will be developed for improving models through identification of model development choices that maximize the reliability of a model to correctly inform decisions. The novel model evaluation research program is therefore expected to help shape the future state-of-practice and state-of-the-science with regard to hydrologic modelling in support of decision-making. The new model evaluation framework will also be applicable to other types of environmental or water resources simulation models.