David H. Huddleston

Hydrodynamic Modeling of St. Louis Bay Estuary and Watershed Using EFDC and HSPF

Abstract St. Louis Bay estuary is a vital water body in the Mississippi Gulf Coast Region and greatly affects the water quality in the Mississippi Sound. As the first step of total maximum daily load (TMDL) study, a hydrodynamics model was developed by integrating Hydrological Simulation Program Fortran (HSPF) and Environmental Fluid Dynamics Code (EFDC). In this application the EFDC model was configured to simulate time-varying surface water elevation, velocity, salinity, and water temperature. The HSPF was applied to compute the fresh water discharge from the upstream watersheds. The model reasonably simulated the tidal range and phase. The simulated water temperature and salinity showed good and fairly good agreement with observations. The calculated correlation coefficients between computed and observed velocity were lower compared with those for water level, temperature, and salinity, but the magnitudes of simulated velocity were found to be in the range of observed data. The wind data was found to have strong impacts on velocity simulation by modeling verification tests. Near the study area, there is wind data available only at one station, which has been applied to the entire modeling domain. The lack of high-resolution wind data makes it very difficult to simulate the velocity distribution well. It is anticipated and recommended that the development of this model be continued to synthesize additional field data into the modeling process.

Fecal coliform modeling under two flow scenarios in St. Louis Bay of Mississippi

St. Louis Bay, along with its two major tributaries, Wolf River and Jourdan River, are included in the Mississippi 1998 Section 303(d) List for violation of the designated water use of recreation and shellfish harvesting. Fecal coliform was identified as one of the pollutants that caused the water quality impairment. In order to facilitate the total maximum daily loads (TMDL) development, the fecal coliform dynamics was investigated under 2 flow scenarios with a calibrated and validated modeling framework by integration of Environmental Fluid Dynamic Code (EFDC) and Hydrological Simulation Program Fortran (HSPF). EFDC was used to model the hydrodymics and fecal coliform transportation in the Bay and the tributaries, whereas HSPF was applied to compute the flow and fecal coliform loadings from the watersheds. The total amount of precipitation in the dry year simulation corresponds to a 50-year return period of low flow condition, and a 10-year return period of high flow condition for wet weather simulation. For EFDC modeling, the fecal coliform sources considered were the contributions from the 2 upper watersheds (no tidal influence), the 28 small surrounding watershed, and 12 municipal, industrial, and domestic point sources. When simulating the fecal coliform loadings from the 2 upper watersheds using HSPF, the simulated non-point source loadings of fecal coliform included wildlife, land application of hog and cattle manure, land application of poultry litter, and grazing animals. The EFDC modeling results indicated that the wet weather exerted greater stress on fecal coliform water quality conditions. The number of exceedance of fecal coliform water quality standard in wet year simulation is much higher than that in dry year simulation. The impact of the upper rural watersheds loads on fecal coliform levels in the St. Louis Bay is much less significant than that from the surrounding urban runoff. Fecal coliform TMDL development should be based on high flow conditions since the decision makers are more concerned about worse scenarios. This fecal coliform modeling research would provide useful information of critical condition selection for TMDLs development in similar coastal areas.

Modeling nutrient dynamics under critical flow conditions in three tributaries of St. Louis Bay

The pesticides originally designed to kill target organisms are dangerous for many other wild species. Since they are applied directly to the environment, they can easily reach the water basins and the topsoil. A dataset of 125 aromatic pesticides with well-expressed aquatic toxicity towards trout was subjected to quantitative structure activity relationships (QSAR) analysis aimed to establish the relationship between their molecular structure and biological activity. A literature data for LC50 concentration killing 50% of fish was used. In addition to the standard 2D-QSAR analysis, a comparative molecular field analysis (CoMFA) analysis considering the electrostatic and steric properties of the molecules was also performed. The CoMFA analysis helped the recognition of the steric interactions as playing an important role for aquatic toxicity. In addition, the transport properties and the stability of the compounds studied were also identified as important for their biological activity.Previous research results indicated that dry weather condition has complicated impacts on nitrogen dynamics; monitored and modeling data showed both increased and decreased levels. In order to facilitate the total maximum daily loads (TMDLs) development at three tributaries of St. Louis Bay estuary, the nitrogen dynamics were investigated for two designed critical flow conditions by integrating Hydrological Simulation Program Fortran (HSPF), Environmental Fluid Dynamics Code (EFDC), and Water Quality Analysis Simulation Program (WASP). The total amount of precipitation during the dry year corresponded to a flow condition with return period of 50 years, and 10-year return period for wet year. The dry year contributed more total nitrogen (TN) loads per unit flow volume. At the upstream tributaries, the computed peak reach-averaged TN concentrations were significantly higher for dry weather simulation than wet conditions, whereas at the near-bay tributary, there were no significant differences in the peak TN concentrations. Hence, for the upstream tributaries, the nitrogen TMDL calculation should be based on dry weather condition since the decision-makers are more concerned about the worse scenario.

Watershed modeling of dissolved oxygen and biochemical oxygen demand using a hydrological simulation Fortran program

Several inland water bodies in the St. Louis Bay watershed have been identified as being potentially impaired due to low level of dissolved oxygen (DO). In order to calculate the total maximum daily loads (TMDL), a standard watershed model supported by U.S. Environmental Protection Agency, Hydrological Simulation Program Fortran (HSPF), was used to simulate water temperature, DO, and bio-chemical oxygen demand (BOD). Both point and non-point sources of BOD were included in watershed modeling. The developed model was calibrated at two time periods: 1978 to 1986 and 2000 to 2001 with simulated DO closely matched the observed data and captured the seasonal variations. The model represented the general trend and average condition of observed BOD. Water temperature and BOD decay are the major factors that affect DO simulation, whereas nutrient processes, including nitrification, denitrification, and phytoplankton cycle, have slight impacts. The calibrated water quality model provides a representative linkage between the sources of BOD and in-stream DO\BOD concentrations. The developed input parameters in this research could be extended to similar coastal watersheds for TMDL determination and Best Management Practice (BMP) evaluation.

A computer-aided visualization tool for stochastic theory education in water resources engineering

In this paper, we propose and demonstrate the proof‐of‐concept for a computer‐aided visualization tool for stochastic theory education in water resources engineering. Using Java Native Interfacing, the tool can wrap a space‐time stochastic model written in any computer language and also not require any specific language compiler during tool usage. This feature also allows the tool to be implemented very easily on any configuration of currently used classroom PCs. We also gauged the merit of a computer‐aided visualization tool in the classroom by conducting a survey of the civil engineering (CE) curricula of US universities. Questionnaires were distributed to the instructors via an online survey. Eighty‐four percent of the universities surveyed were found to offer a general semblance of stochastic theory education in their curriculum for CE. A similar percentage of the total 241 courses that we initially surveyed were found to be available at the graduate level, while 4.5% and 11.5% were either dual‐listed or undergraduate‐level courses, respectively. Forty universities were found to have complete (integrated) courses dedicated to stochastic theory education (or a near‐relative related discipline). 11.2% (27) of these courses were relevant to water resources engineering, while only 9.5% (23 courses) were related to surface water hydrology. Only 62.5% of instructors were active users of some kind of computer‐aided visualization tools for classroom instruction. All instructors believed that a rapid visualization system to represent the effect of input (i.e., an aspect of stochastic theory) on output (i.e., application or representation of variability) would enhance the technology as a learning tool. Surveyed instructors were unanimous in their willingness to integrate such an instruction tool for teaching theory using real‐world examples of water resources engineering. However, 42% felt that such a tool would need to be user‐friendly and graphically very attractive in order to be popular among students. We believe that with the demonstration of proof‐of‐concept of our proposed computer‐aided visualization tool, the effectiveness of modernizing course curricula in CE for undergraduate water resources education can be made more compatible with the needs of the 21st century and that there is indeed sustainable demand in the classrooms for its institutional development.

Assessment of water quality conditions in the St. Louis Bay watershed

The water quality data from 14 sampling stations in the St. Louis Bay watershed were analyzed to evaluate the water quality conditions. The differences in water quality parameters between base and storm flow events were compared to identify the pollutant sources. The results indicated that fecal coliform was the primary cause for water quality impairment of the study area. The overall water quality conditions were good in terms of dissolved oxygen, eutrophication, and total suspended solid (TSS). The dominant sources of bio-chemical oxygen demand (BOD) could be from the failing septic system; the majority of the water samples exceeding Mississippi Department of Environmental Quality (MDEQ) target levels were from base flow events. Different from BOD, the majority of the water samples exceeding the water quality criteria and MDEQ target levels were from the storm events for fecal coliform, chemical oxygen demand, total organic carbon, TKN, NO3, NH3, chlorophyll a, and TSS. Based on cluster analysis, the sampling stations were classified into two major categories: upstream and near-coast stations. The major differences between upstream and near-coast stations are elevation, soil texture, and impacts of human activity. The results from this research would provide useful information for total maximum daily load calculation, development of a computational watershed model, and development of best management practices for the St. Louis Bay watershed and similar study area.

Application and evaluation of two nutrient algorithms of hydrological simulation program fortran in Wolf River watershed

This study performs a comparison of two nutrient algorithms of Hydrological Simulation Program Fortran, PQUAL/IQUAL and AGCHEM. Watershed nutrient models with, PQUAL/IQUAL and AGCHEM, were developed and calibrated separately with observed data in the Wolf River watershed. Compared to AGCHEM modules, the PQUAL/IQUAL algorithm was found to have several disadvantages. Examples are: (i) it is a simple loading estimation algorithm, and cannot represent the soil nutrient processes; and (ii) the interactions of modeled nutrient species in the soil cannot be simulated. The AGCHEM modules are capable of explicitly representing the comprehensive nutrient processes in the soil such as fertilization, atmospheric deposition, manure application, plant uptake process, and the transformation processes. Therefore, AGCHEM modules afford the ability to evaluate the alternative management practice and model the interactions between nutrient species. However, our modeling results indicated that the inclusion of AGCHEM modules do not significantly improve the nutrient modeling performance but rather take much more time in model development. The nutrient algorithms selection for total maximum daily loads development depends on the data availability, required modeling accuracy, and available time for model development.

An open-source software for interactive visualization using C++ and OpenGL: Applications to stochastic theory education in water resources engineering

The purpose of this article is to explain the design and implementation of an open‐source engineering education software called Stochastic Theory Education through Visualization Environment (STEVE), version 2.0. In an earlier article, a proof‐of‐concept for a computer‐aided visualization tool (also named STEVE, version 1.0) for stochastic theory education in water resources engineering was articulated [see, Schwenk et al. Comput. Appl. Eng. Educ., 2008, in press). Using Java Native Interfacing, it was shown that STEVE 1.0 could wrap a space–time stochastic model written in any computer language and be independent of any specific language compiler during tool usage. This article describes the general philosophy, software design, and classroom usage for STEVE with significant improvements on visualization and user‐friendliness (hence, rightfully called version 2.0). The software was created using the C++ programming language with the Microsoft Windows Applications Programming Interface (API). OpenGL was used for the visualization display, and the OpenGL Utility Toolkit (GLUT) was used to visualize text inside the OpenGL window. The instructor‐specified simulation program on stochastic theory was written in Fortran 77. The application has user‐friendly options for modifying input data and parameter specifications as desired by the instructor or the student user. STEVE 2.0 has been tested with the Windows XP and Windows Vista operating systems. For the benefit of interested users and software makers, we also provide the software application, a short tutorial and all pertinent source codes as freeware for download on our STEVE homepage at http://iweb.tntech.edu/saswe/steve.html.