Science Inventory

Assessing Lake Trophic Status: A Proportional Odds Logistic Regression Model

Citation:

Nojavan, F., B. Kreakie, Jeff Hollister, AND S. Qian. Assessing Lake Trophic Status: A Proportional Odds Logistic Regression Model. Association for the Sciences of Limnology and Oceanography (ASLO), Santa Fe, NM, June 05 - 10, 2016.

Impact/Purpose:

This work revisits Carlson’s classic lake trophic status modelling. Our proportional odds logistic regression model accurately models all lake trophic class using exclusively secchi depth and eco-region.

Description:

Lake trophic state classifications are good predictors of ecosystem condition and are indicative of both ecosystem services (e.g., recreation and aesthetics), and disservices (e.g., harmful algal blooms). Methods for classifying trophic state are based off the foundational work of Carlson from 1977. This work presents an updated lake trophic classification model using secchi disk depth and eco-region. Secchi disk depth and eco-region are extremely simple and cost effective to measure without the need to develop an extensive biological monitoring program, and therefore are effective measures to predict lentic productivity. The ordered categorical lake trophic state is modeled using a proportional odds logistic regression model. The model contributes to literature in developing an index that is multivariate, varies across a continuum (i.e. continuous) and is used to classify lakes in probabilistic terms. The model was developed and validated using National Lake Assessment 2007 data set. The overall accuracy was 0.67 and the balanced accuracy ranged between 0.74 and 0.79 among classes.

Record Details:

Record Type: DOCUMENT (PRESENTATION/ABSTRACT)
Product Published Date: 06/27/2016
Record Last Revised: 06/27/2016
OMB Category: Other
Record ID: 320139

Organization:

U.S. ENVIRONMENTAL PROTECTION AGENCY

OFFICE OF RESEARCH AND DEVELOPMENT

NATIONAL HEALTH AND ENVIRONMENTAL EFFECTS RESEARCH LABORATORY

ATLANTIC ECOLOGY DIVISION

POPULATION ECOLOGY BRANCH