Data and model archive for multiple linear regression models for prediction of weighted cyanotoxin mixture concentrations and microcystin concentrations at three recurring bloom sites in Kabetogama Lake in Minnesota
Multiple linear regression models were developed using data collected in 2016 and 2017 from three recurring bloom sites in Kabetogama Lake in northern Minnesota. These models were developed to predict concentrations of cyanotoxins (anatoxin-a, microcystin, and saxitoxin) that occur within the blooms. Virtual Beach software (version 3.0.6) was used to develop four models: two cyanotoxin mixture (MIX) models and two microcystin (MC) models. Models include those using readily available environmental variables (for example, wind speed and specific conductance) and those using additional comprehensive variables (based on laboratory analyses). Many of the independent variables were averages over a certain time period prior to a sample date, whereas other independent variables were lagged between 4 and 8 days. Funding for this work was provided by the U.S Geological Survey ? National Park Service Partnership and the U.S. Geological Survey Environmental Health Program (Toxic Substance Hydrology and Contaminant Biology). The resulting model equations and final datasets are included in this data release while an associated child item model archive includes all the files needed to run and develop these VB models.
|Data and model archive for multiple linear regression models for prediction of weighted cyanotoxin mixture concentrations and microcystin concentrations at three recurring bloom sites in Kabetogama Lake in Minnesota
|Victoria G Christensen, Erin A Stelzer, Megan J Haserodt
|USGS Digital Object Identifier Catalog
|Upper Midwest Water Science Center