A Multiple Model Approach for Determining the Phase II Total Nitrogen TMDL To the Neuse River Estuary

The Environmental Protection Agency (EPA) conditionally approved the first phase of the TMDL to the Neuse River estuary in July 1999. It called for a second phase, to be completed in 2001, that would incorporate the latest monitoring data and models. The North Carolina Division of Water Quality (DWQ) received approval from EPA on the second phase in March 2002.The second phase acknowledges that a substantial amount of model prediction uncertainty exists and that the reduction target may be changed in the future. Nevertheless, the recommended reduction amount is 30% from the average 1991-1995 total nitrogen load delivered to the estuary. The State of North Carolina and EPA-Region IV used three water quality response models of the estuary to determine the reduction target. The models include:A CE-Qual W2 application to the Neuse estuary, also known as the Neuse Estuary Eutrophication Model (NEEM). The NEEM is a process-based, material balance simulation model that tracks an assortment of water quality variables related to eutrophication. Its compartments are in two dimensions: longitudinal segments along the estuary and vertical layers through the water column – the model laterally (across the estuary) aggregates its calculations/predictions.A probability network model called the Neuse Estuary Bayesian Ecological Response Model (Neu-BERN). While initially structured to focus on broader ecological endpoints, later effort was put into this model to enable it to predict algal response (chlorophyll a) to various nitrogen loads. This predictive link can be thought of as a multiple regression model.A Water Analysis Simulation Program (WASP) application to the Neuse estuary. EFDC/WASP is similar to CE-QUAL W2 in that it is also a process-based simulation model. The key difference between the two models is that WASP has compartments in three dimensions, rather than two for CE-QUAL W2. In addition to having longitudinal segments and vertical layers, WASP divides the estuary laterally (i.e. from bank to bank).In addition to the water quality response models, DWQ used analyses of riverine total nitrogen loading between 1995 and 1998 by Duke University researchers to arrive at the reduction target.Though the predictions and decisions contained in the phase II TMDL are based on the best currently available information, there is substantial uncertainty in them. For this reason, DWQ intends to follow an adaptive approach to managing the estuary. In other words, DWQ will use the models to guide decision making, but continuing observation of the watershed and estuary, as nitrogen controls are implemented (e.g., the Neuse Rules, and other measures such as wetland restoration and establishment of conservation easements), is expected to be the best approach for determining the appropriate level of management.