Scott A. Wells

Challenges and Opportunities for Integrating Lake Ecosystem Modelling Approaches

A large number and wide variety of lake ecosystem models have been developed and published during the past four decades. We identify two challenges for making further progress in this field. One such challenge is to avoid developing more models largely following the concept of others (‘reinventing the wheel’). The other challenge is to avoid focusing on only one type of model, while ignoring new and diverse approaches that have become available (‘having tunnel vision’). In this paper, we aim at improving the awareness of existing models and knowledge of concurrent approaches in lake ecosystem modelling, without covering all possible model tools and avenues. First, we present a broad variety of modelling approaches. To illustrate these approaches, we give brief descriptions of rather arbitrarily selected sets of specific models. We deal with static models (steady state and regression models), complex dynamic models (CAEDYM, CE-QUAL-W2, Delft 3D-ECO, LakeMab, LakeWeb, MyLake, PCLake, PROTECH, SALMO), structurally dynamic models and minimal dynamic models. We also discuss a group of approaches that could all be classified as individual based: super-individual models (Piscator, Charisma), physiologically structured models, stage-structured models and trait-based models. We briefly mention genetic algorithms, neural networks, Kalman filters and fuzzy logic. Thereafter, we zoom in, as an in-depth example, on the multi-decadal development and application of the lake ecosystem model PCLake and related models (PCLake Metamodel, Lake Shira Model, IPH-TRIM3D-PCLake). In the discussion, we argue that while the historical development of each approach and model is understandable given its ‘leading principle’, there are many opportunities for combining approaches. We take the point of view that a single ‘right’ approach does not exist and should not be strived for. Instead, multiple modelling approaches, applied concurrently to a given problem, can help develop an integrative view on the functioning of lake ecosystems. We end with a set of specific recommendations that may be of help in the further development of lake ecosystem models.

A general model of the activated sludge reactor with dispersive flow—I. model development and parameter estimation

Abstract A general dynamic model, including biological processes, hydraulics, oxygen transfer and temperature, was developed for the activated sludge reactor. Transport of wastewater constituents was described by the 1-D advection–dispersion equation with a source term written principally according to the Activated Sludge Model No. 1 (ASM1). Oxygen transfer was strongly related to the air flowrate and this feature was included in the oxygen mass balance. The temperature model was based on a heat balance of the reactor. Experimental methods for parameter estimation were demonstrated. The dispersion coefficient and the oxygen transfer coefficient were estimated from tracer studies and off-gas measurements, respectively. Values of the kinetic and stoichiometric coefficients estimated from batch experiments were generally found to agree with the default values used in the ASM1.

A general model of the activated sludge reactor with dispersive flow—II. model verification and application

Abstract A general dynamic model of the activated sludge reactor, including biological processes, hydraulics, oxygen transfer and temperature, was verified against both steady-state and dynamic field data. These data originated from the Rock Creek wastewater treatment plant located in Hillsboro, Oregon (USA). During simulations only one set of the model parameters was used. The model was capable of predicting dissolved oxygen concentrations and consumption, ammonia concentrations, and nitrate+nitrite concentrations with time and distance along the activated sludge reactor. Uncertainty analysis enabled the classification of model parameters (kinetic and stoichiometric coefficients, and wastewater components) according to model sensitivity. The maximum growth rate of nitrifiers, μ A,max , and the ammonia concentration, S NH , were the most sensitive coefficient and wastewater component, respectively, when the sum of ammonia, nitrate+nitrite, dissolved oxygen concentrations and total oxygen uptake rate was taken into account. Application of the model was demonstrated by analyzing the impact of changing the operational mode (from A/O process to 4-stage Bardenpho) on dissolved oxygen conditions and nitrogen removal.

Impact of Phosphorus Loading from the Watershed on Water Quality Dynamics in Lake Tenkiller, Oklahoma, USA”

Tenkiller Lake is a Corps of Engineers reservoir operated by the Tulsa District located in Oklahoma. Because of concern over possible impacts to the reservoir of poultry waste applied to agricultural lands in the basin, a hydrodynamic and water quality model of Tenkiller Lake was constructed to evaluate how the reservoir would respond to changes in nutrient loading from the watershed. The CE-QUAL-W2 model was calibrated to field data over the period of January 1, 2005 through September 2007. Comparisons of field data to model predictions were made for water level, temperature, NH4-N, NO3-N, PO4-P, total N, Total P, chlorophyll a, and dissolved oxygen at multiple reservoir sites at various depths. The model was used to predict the response of the reservoir to changes in Total P loading in the watershed over a 50-year period. Several scenarios were run with the model including no changes in nutrient loading from the watershed, natural conditions, and cessation of further land application of poultry waste in the watershed. Results and recommendations for achieving water quality targets in Lake Tenkiller were outlined. The cessation scenario showed that even after 50 years, the reservoir would still not have recovered to a natural pristine state. High residual Total P loading in the basin from past agricultural practices would continue to degrade water quality.

Modeling the Response of Dissolved Oxygen to Phosphorus Loading in LakeSpokane

ABSTRACT Wells SA, Berger CJ. 2016. Modeling the response of dissolved oxygen to phosphorus loading in Lake Spokane. Lake Reserv Manage. 32:270–279. Mathematical models of hydrodynamics and water quality are often used to determine the assimilative capacity of a waterbody when waterbodies violate state water quality standards. A model of the Spokane River and Lake Spokane in eastern Washington was developed to evaluate the assimilative capacity of the waterbody by setting a total maximum daily load (TMDL). A CE-QUAL-W2 model of this system was developed to establish the TMDL limits for the critical low-flow year of 2001. A recent paper evaluated this model and raised several issues about the validity of this model application as a regulatory tool related to its ability to predict total phosphorus, dissolved oxygen, and chlorophyll a relationships. This paper analyzes the validity of their critiques. For example, the critique used an incorrect formula to calculate total phosphorus inflows into Lake Spokane and used a volume-weighted minimum hypolimnetic dissolved oxygen (DOmin) that was not representative of hypolimnetic conditions. They also assumed incorrectly that the hydrologic, meteorological, operational, and sediment conditions of the 2001 TMDL model would be representative of conditions in other years. Although the water quality model of Lake Spokane can be improved, the critique does not invalidate the model as a tool to evaluate how the lake responds to nutrient environmental controls.

Meeting Temperature Requirements for Fisheries Downstream of Folsom Reservoir, California

Folsom lake, located near Sacramento California USA, is a deep-storage reservoir that provides municipal water, power generation and cold water for salmonid fish in the lower American River. The reservoir has discrete temperature control shutters on the three powerhouse intakes that allow the operator to choose different water levels at each intake to accommodate downstream temperature requirements. A complex model of the reservoir was developed using the CE-QUAL-W2 model (Cole and Wells, 2013) and was calibrated to historical operations over a 10-year time period. Absolute mean temperature errors in model profiles and in downstream temperature were 0.57 o C and 0.59 o C, respectively, well less than the goal of <1 o C. Variability in meteorological data sources and leakage through the temperature control shutters at the dam were challenges during model calibration. A complex operational model tool was developed using the CE-QUAL-W2 model to automatically determine how best to select outlet shutter position in order to reach the downstream temperature goals for fish. The goal was to develop an operational tool that best allows for meeting overall goals of using the limited stored cold water pool to meet downstream temperature objectives. The model proved successful running long-term simulations that can be used to evaluate operations based on modified or forecasted hydrological and meteorological inputs.

Hydrodynamic and Water Quality River Basin Modeling using CE-QUAL-W2 Version 3

The CE-QUAL-W2 is a two-dimensional (longitudinal-vertical) water quality and hydrodynamic computer simulation model that was originally developed for deep, long, and narrow waterbodies. The current model, Version 2, has been used in over 200 river, reservoir/lake, and estuary applications throughout the U.S. and abroad. Version 2, though, cannot accommodate systems that have a significant sloping water surface since the vertical coordinate system is aligned with gravity and vertical accelerations are neglected. The governing equations for CE-QUAL-W2 were re-derived so that it could be applied to entire river basins including river-estuary, lake-river, and reservoir-river systems with channel slopes. This re-derivation is one of many improvements for the Version 3 code. Other improvements include improved numerical schemes, improved and additional water quality algorithms, ability of the user to add hydraulic structures between model segments and the ability to model the effect of hydraulic structures on gas transfer. Test cases for this new code include a 244 km section of the Lower Snake River in the Northwestern USA; the Bull Run River basin composed of 3 water supply reservoirs and 2 river sections with a 2.2% and a 1.4% average slope in the Oregon Cascade mountains, USA; and the Columbia Slough system in Portland, OR, USA composed of 33 separate lake systems connected by hydraulic structures and a fresh-water tidal region.

Analytical and field verification of a 3D hydrodynamic and water quality numerical scheme based on the 2D formulation in CE-QUAL-W2

ABSTRACT A new 3D hydrodynamic model was developed to simulate water quality transport in surface waterbodies. The governing equations are the continuity equation, free surface equation, momentum equation and transport equation. The 2D numerical scheme of CE-QUAL-W2 was expanded in three dimensions and modified to solve for the free surface elevation. A time splitting technique was employed to solve the momentum and transport equation. The numerical formulation of the 3D scheme used a novel solution, which resulted in a tri-diagonal matrix form for solving the free surface equation rather than a more computationally intensive penta-diagonal matrix solution. In addition, the hydrodynamic and water quality equations were solved at the same time step in order to allow feedback between water quality and hydrodynamics. The verification of the model hydrodynamics and temperature was performed by comparing the model predictions to known analytical solutions and field data from Lake Chaplain, Washington, USA. There was good agreement of the solution of the hydrodynamic equations to analytical solutions and field data.

Prediction of GHG Emissions from a New Reservoir

Not until the 1990s was the significance of carbon emissions from reservoirs to greenhouse gas (GHG) accumulation in the atmosphere realized. Currently, hydroelectric projects proposed for World Bank funding must estimate their net GHG footprint, which in most cases has been estimated based on field measurements from similar reservoirs. Here we describe the development and application of CE-QUAL-W2 for prediction of future water quality and net GHG emissions from the proposed Amaila reservoir and downstream Kuribrong River in Guyana, South America. Sediment diagenesis was simulated, and the model enhanced to include the decomposition of a submerged tropical forest. The project site is located upstream of Amaila Falls in the remote highlands of the Guiana Shield. The water quality of the upper Kuribrong River exhibits low pH, low alkalinity, and somewhat high organic carbon concentrations, which is very different from reservoirs in the lowlands of French Guiana and Brazil where GHG fluxes have been measured. High CO 2 concentrations in the reservoir were caused primarily by the low pH and high CO 2 concentrations in the river inflows. Much of this CO 2 was emitted to the atmosphere within the reservoir, and most of the rest was passed through the dam. Net CO 2 emissions for the average flow year with vegetation harvested or burned were 30,799 T/yr. Net emissions for the no removal of vegetation were 71,824 T/yr. Methane concentrations in the dam outflow were less than 0.8 mg/L after the first year and 0.4 mg/l after 5 years.

Automatic Calibration of Water Quality Models for Reservoirs and Lakes

One of the most important purposes of surface water resource management is to develop predictive models to assist in identifying and evaluating operational and structural measures for improving water quality. To better understand the effects of external and internal nutrient and organic loading and the effects of reservoir operation, a model is often developed, calibrated, and used for sensitivity and management simulations. The importance of modeling and simulation in the scientific community has drawn interest towards methods for automated calibration. This paper addresses using an automatic technique to calibrate of a 2-dimensional, hydrodynamic, and water quality model CEQUAL-W2 (Cole and Wells, 2013). CEQUAL-W2 is a two-dimensional (2D) longitudinal/vertical hydrodynamic and water quality model for surface water bodies, modeling eutrophication processes such as temperature-nutrient-algae-dissolved oxygen-organic matter and sediment relationships. The numerical method used for calibration in this paper is the particle swarm optimization method developed by Kennedy and Eberhart (1995) and inspired by the paradigm of birds flocking. The objective of this calibration procedure is to choose model parameters and coefficients affecting temperature, water elevation, chlorophyll a, dissolved oxygen, and nutrients (such as NH4, NO3, and PO4). A case study is presented for the Karkheh Reservoir in Iran with a capacity of more than 5 billion cubic meters that is the largest dam in Iran with both agricultural and drinking water usages. This algorithm is shown to perform very well for determining model parameters for the reservoir water quality and hydrodynamic model. Implications of the use of this procedure for other water quality models are also shown.