The Right Scale for the Question Asked: A Challenge for Great Lakes Hydrology and Restoration Forecasts
Abstract
Water is a primary metric for viewing Great Lakes health and restoration. Our ability to characterize and forecast hydrologic conditions continues to improve, but challenges remain. One notable issue is the scale of the Great Lakes system. Because of its size, constructing a tractable model that encompasses the entire watershed requires extensive simplification. Questions that require higher spatial and temporal resolution, however, often cannot be well-simulated using the coarse resolution of Great Lakes watershed-scale tools. In work conducted for the Great Lake Restoration Initiative, a regional Precipitation-Runoff Modeling System (PRMS) model has been constructed for the entire Lake Michigan watershed. The model was parameterized using Hydrologic Response Units of about 207 square kilometers, and 148 stream gages ranging in HUC 8-12 scales. This representation of the system is useful for annual and perhaps seasonal/monthly characterizations of water fluxes to Lake Michigan. Likewise, this scale of forecast can be useful for certain predictions, such as lake stage forecasts where Lake Michigan is the basinwide receptor - a receptor that is a natural integrator of large space and time. Many questions need a more refined characterization of the dynamics and processes operating within the system. For these, such a coarse representation of space and time is not appropriate. Many Great Lakes issues, such as localized Areas of Concern (AOCs) and Harmful Algal Blooms (HABs), are not well represented by basinwide averages. For example, annual flows are not useful for assessing the resilience of a sand-capping restoration activity to future climate variability. In this work, a spatially detailed PRMS model has been refined from the regional PRMS model to construct more representative forecasts for these types of questions. The sub-basin model needs to answer the smaller scale question, but must also be consistent with the conceptualization and parameterization of the regional framework. These issues underscore the need to use all the tools in the toolbox, and not direct inordinate effort to unrealistic goals, such as a single simulator that represents all spatial and temporal scales. That is, there can be no expectation of "the" model for a Great Lake, only a family of models tuned for constructing forecasts for specific questions.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.H54E..08H
- Keywords:
-
- 1800 HYDROLOGY;
- 1807 HYDROLOGY / Climate impacts;
- 1847 HYDROLOGY / Modeling;
- 1860 HYDROLOGY / Streamflow