Capturing the Great Lakes water balance in a Bayesian Network

Session: 53. - Great Lakes Water Level Fluctuations and Water Management

Joeseph Smith, Cooperative Institute for Great Lakes Research, [email protected]
Scott Steinschneider, Department of Biological and Environmental Engineering, Cornell University, [email protected]
Jacob Bruxer, Environment and Climate Change Canada, [email protected]
Andrew Gronewold, NOAA, GLERL, [email protected]

Abstract

The Great Lakes - St. Lawrence River Adaptive Management Committee (GLAM), along with the International Joint Commission, identified a need to examine "how the [Laurentian Great Lakes] system may be changing over time and whether any modifications to the regulation plan(s) may be warranted to address what is learned... including emerging issues and/or to address changing conditions". With respect to Great Lakes hydrology, the system can be viewed through the water balance (and level) of the lakes. To make defensible conclusions about hydrologic changes in the system, a closed water balance is desired. This talk discusses a Bayesian Network that assimilates multiple estimates of the Great Lakes water balance and its components, simulation of which (through Markov chain Monte Carlo methods) produces new, balance closing estimates. Additionally, we present high level results of this Large Lakes Statistical Water Balance Model (L2SWBM), which was used to generate a new record of the Great Lakes water balance from 1950 through 2015.

1. Keyword
hydrologic cycle

2. Keyword
Great Lakes basin

3. Keyword
modeling

4. Additional Keyword
water balance

5. Additional Keyword
regional modeling

6. Additional Keyword
Bayesian Network