This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ Preprints) and either DOI or URL of the article must be cited.
Lake trophic state indices have long been used to provide a measure of the trophic state of lakes. Over time it has been determined that these indices perform better when they utilize multiple metrics and provide a continuous measurement of trophic state. We utilize such a method for trophic state that is based upon a Proportional Odds Logistic Regression (POLR) model and extend this model with a Bayesian multilevel model that predicts nutrient concentrations from universally available GIS data. This Bayesian multilevel model provides relatively accurate measures of trophic state and has an overall accuracy of 60%. The approach illustrates a method for estimating a continuous, mutli-metric trophic state index for any lake in the United States. Future improvements to the model will focus on improving overall accuracy and use variables that are more sensitive to change over time.
This is an early draft version of a research note explaining an extended application of a Bayesian Probabilistic Odds Logistic Regression model for a lake trophic state index.