Citation:

Kreakie, B., Bryan Milstead, AND J. Kiddon. Using National Coastal Assessment Data to Model Estuarine Water Quality at Large Spatial Scales. Coastal & Estuarine Research Federation (CERF) 24th Biennial Conference, Providence, Rhode Island, November 05 - 09, 2017.

Impact/Purpose:

With ever increasing pressures put on our coastal environments, monitoring data are invaluable to understanding how these environments are responding to these pressures. Monitoring data also make it possible to make predictions about how environments may respond to future conditions. By using USEPA-collected environmental data, this research examines the environmental drivers to changes in water quality. We show that chlorophyll a concentrations (a proxy for coastal eutrophication) are robustly predicted by total nitrogen, total phosphorus, and subregion (local spatial variation). We also show that 2010 and 2015 chlorophyll a concentrations are much higher relative to other sample years.

Description:

The water quality of the Nation’s estuaries is attracting scrutiny in light of population growth and enhanced nutrient delivery. The USEPA has evaluated water quality in the National Coastal Assessment (NCA) and National Aquatic Resource Surveys (NARS) programs. Here we report on a Random Forest modelling investigation of the survey data to identify the predictor variables affecting surface chlorophyll concentrations in designated regions, paying particular attention to the nutrient measures employed (TN and TP vs DIN and DIP) and regional scale. We also examine model results for indications of change in chlorophyll concentrations over time. The water quality data used for RF modelling were collected at over 7800 randomly selected sites from 2000 to 2006 (NCA) and in 2010 & 2015 (NARS). The sites were sampled during the summer using consistent collection and assessment methods. Water quality measures included temperature, salinity, pH, Secchi depth, dissolved oxygen, chlorophyll a, dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphorous (DIP). Total nitrogen (TN) and total phosphorus (TP) were measured nationwide in 2005 and later. Our work identified 26 sub-regions to investigate the importance of local distinctions. The modelling technique used here is a machine learning algorithm that produces unbiased estimates of error and robust measures of variable importance by bootstrapping predictor variables and employing training/test data subsets. The model results are robust (model statistics: mean squared error = 0.22 and adjusted R2 = 0.685). Based on the percent mean decrease in accuracy, TN, TP, and TN/TP ratio are important predictor variables for chlorophyll, while DIN and DIP are relatively unimportant. The finer-scale subregions are significant predictors, while the large-scale regions are not. Chlorophyll levels are significantly greater in 2010 and 2015 relative to other years.

Record Details: