Improved thermal structure simulation and optimized sampling strategy for Lake Erie using a data assimilative model

Document Type

Article

Publication Date

2-2020

Department

Department of Civil, Environmental, and Geospatial Engineering; Great Lakes Research Center

Abstract

Lake Erie has experienced substantial environmental issues (e.g., hypoxia, harmful algal blooms) for decades, which are closely related to the lake’s thermal characteristics. While three-dimensional (3D) hydrodynamic models have been widely applied to Lake Erie, challenges remain due to model representation of physical processes, errors and uncertainty in boundary conditions and forcing terms. The Great Lakes region has a relatively dense and long-term observational record, and these observational data have been used for model initialization and verification, but have not been incorporated into 3D model simulations through data assimilation (DA) to create reanalysis products or improve short-term forecasts. In this work, we developed and evaluated DA to improve thermal structure simulation of Lake Erie. Moored instrument data and satellite data are incorporated into a data-assimilative hydrodynamic model for analysis and evaluation. Results show that DA can effectively improve the model performance to create reanalysis fields when the DA formulation is appropriately developed in recognition of the dynamic complexities and anisotropic error covariances of Lake Erie. The data assimilative model also improves forecasting accuracy and restrains forecasting uncertainty to an acceptable level on a timescale of 1–7 days after being unleashed from DA. Lastly, data sampling strategies based on an error correlation map are examined. Results show the method can effectively reduce the sampling effort while still achieving similar model skills with potential for optimal design of an observation network or field sampling strategy.

Publication Title

Journal of Great Lakes Research

Link to Full Text

Share

COinS