Science Inventory

Linking multi-media modeling with machine learning to assess and predict lake chlorophyll a concentrations

Citation:

Feng Chang, C., Valerie Cover, C. Tang, P. Vlahos, D. Wanik, J. Yan, J. Bash, AND M. Astitha. Linking multi-media modeling with machine learning to assess and predict lake chlorophyll a concentrations. JOURNAL OF GREAT LAKES RESEARCH. International Association for Great Lakes Research, Ann Arbor, MI, 47(6):1656-1670, (2021). https://doi.org/10.1016/j.jglr.2021.09.011

Impact/Purpose:

The ability to predict water quality in lakes is important because lakes provide essential ecosystem services. While numerical prediction models do not dynamically describe the entirety of air-water-soil interactions, there is an abundance of observations and simulated data available from weather, air quality, agroecosystem, and hydrological models. In this study, we demonstrate how these data can be combined with machine learning to predict in-situ measurements of chlorophyll-α. This, in turn, sheds light on the drivers of HABs, and can inform prediction of events.

Description:

Eutrophication and excessive algal growth pose a threat on aquatic organisms and the health of the public, environment, and the economy. Understanding what drives excessive algal growth can inform mitigation measures and aid in advance planning to minimize impacts. We demonstrate how simulated data from weather, hydrological, and agroecosystem numerical prediction models can be combined with machine learning (ML) to assess and predict chlorophyll a (chl a) concentrations, a proxy for lake eutrophication and algal biomass. The study area is Lake Erie for a 16-year period, 2002–2017. A total of 20 environmental variables from linked and coupled physical models are used as input features to train the ML model with chl a observations from 16 measuring stations. Included are meteorological variables from the Weather Research and Forecasting (WRF) model, hydrological variables from the Variable Infiltration Capacity (VIC) model, and agricultural management practice variables from the Environmental Policy Integrated Climate (EPIC) agroecosystem model. The consolidation of these variables is conducive to a successful prediction of chl a. Aside from the synergistic effects that weather, hydrology, and fertilizers have on eutrophication and excessive algal growth, we found that the application of different forms of both P and N fertilizers are highly ranked for the prediction of chl a concentration. The developed ML model successfully predicts chl a with a coefficient of determination of 0.81, bias of −0.12 μg/l and RMSE of 4.97 μg/l. The developed ML-based modeling approach can be used for impact assessment of agriculture practices in a changing climate that affect chl a concentrations in Lake Erie.