Current Research
A New Approach to Sediment TMDL Watersheds in the Southern Piedmont
Many streams in the southern Piedmont region are impaired because of high concentrations of suspended sediment and scheduled for development of Total Maximum Daily Load (TMDL) implementation plans. For these waters, it’s not clear if the source is upland erosion from agricultural sources or bank erosion of historic sediment deposited in the flood plains during the 19th and 20th century when cotton farming was extensive. If the TMDL implementation plan addresses the wrong source, stakeholders will become disillusioned with the TMDL process. A new approach is needed to determine if bank erosion of legacy sediment is a significant source and how to implement sediment TMDLs in streams that are in an unstable stage of channel evolution. We propose using the North Fork Broad River, a sediment-impaired watershed in Georgia, as a case study for the new approach. Our hypothesis is that the North Fork Broad River is in an unstable stage and that bank erosion of legacy sediment is a major source of the current, high sediment load. We will use sediment fingerprinting (Cs-137, N-15, trace elements, C, N, S and P) and mixing models to identify the primary sources of erosion. Our preliminary results indicate that Cs-137 can be used to distinguish upland sources of erosion from bank erosion of legacy sediment and that bank erosion is the dominant source of the sediment in the North Fork Broad River. We will use geomorphic assessment (visual ratings at fixed channel intervals and a helicopter video survey) and computer modeling (AGNPS combined with CONCEPT) of channel evolution to determine if the stream is stable or unstable and the best approach to reduce sediment load. Our outreach and education component will promote this approach to stakeholders within the watershed and TMDL specialists, educators, and students throughout the southern region. The priority question we address is what are the hydrologic and geomorphic conditions needed to restore aquatic ecosystems impacted by sediment pollution.
Total Maximum Daily Load for sediment- North Fork Broad River
The North Fork Broad River was included in Water Quality in Georgia, 1998-1999, Attachment I, Addition of Waters to Georgia's Section 303(d) list, for impacted biota and habitat. Sediment was determined to be the pollutant of concern. The current research project focuses to study sediment sources and to model and evaluate the effect of best management practices (BMPs) on sediment loading in the North Fork Broad River in Georgia. The methodology includes a paired watershed monitoring and modeling sediment in selected watersheds pre- and post-BMP implementation. The Soil and Water Assessment Tool (SWAT), a geographic information system (GIS) based hydrologic model will be used to simulate flow and sediment transport in the watershed. Results from this study may help reduce sediment loads on the North Fork Broad River through better decisions on sediment control. The project also aims to better define the sediment loading in the watershed and recommend for the revision of the existing Total Maximum Daily Load (TMDL) for sediment. Use of cutting edge technology such as sediment fingerprinting and remote sensing to identify the critical sediment source areas within the watershed are also in the research agenda.
Data
Nutrient Trading- Etowah River
According to a recent study, Lake Allatoona in north Georgia will be unfit for drinking or recreational purposes within ten years, unless measures are taken to control nonpoint sources of sediment and P in the watershed. Our goal is to develop a scientifically-based framework for trading phosphorus credits between point and non-point sources to meet the phosphorus load restrictions that have been imposed on Lake Allatoona by the Georgia Environmental Protection Division and further reductions that are likely to be imposed in a TMDL to be developed by the end of 2003. The overall objective is to establish a framework for trading phosphorus (P) credits between point sources and agricultural/forestry non-point sources in the Lake Allatoona watershed. Specific project objectives are to: 1) use stream monitoring to determine the P and sediment loading from typical agricultural operations with a range of BMP implementation, 2) use this monitoring data plus historical data to calibrate a watershed-scale model for the basin, 3) use uncertainty analysis of the model to develop scientifically-based P trading ratios for point and agricultural non-point sources, 4) analyze various frameworks for P trading in the watershed, and 5) create a watershed advisory council and use it to engage stakeholders. In 2005 (the second year of our project), we continued to make good progress. Our work has focused on objectives one, two, and five. In regard to the first objective, we established monitoring sites on 12 headwater streams (3 forest and 9 agriculture land use) in the Upper Etowah basin and collected storm and base flow samples for most of the year. Unit area total phosphorus (P) loads were generally less than 1 kg/ha-yr for forested streams and ranged from less than 1 to about 80 kg/ha-yr in the agricultural streams. In regard to the second objective, we have completed the development and calibration of a SWAT model of transport of P and sediment to Lake Allatoona using land use data from 1992 (the calibration data is from the period 1992-1996). The model consists of simulations for the six major tributaries that flow into Lake Allatoona. The model indicated that on average 80% of the P load that entered the streams reached Lake Allatoona, indicating that stream sediments and biota assimilated about 20% of the P load. In regard to objective five, we established a basin stakeholder advisory group and held a first meeting on this group. In the coming year we will use our calibrated SWAT model to estimate current P loads to Lake Allatoona using 2002 land use data. We will also model the effect of selected best management practices in reducing sediment and P loads to the lake (including exporting manure from the watershed) and assess the uncertainty in these reductions. This work will address objective three (determine the uncertainty in nonpoint source reductions). We will also explore the possibility of establishing a system for transporting poultry litter out of the watershed to areas deficient in P and banking credits for this export. This will address objective four (establishing a framework for trading phosphorus credits).
On-site Wastewater Systems: Design Hydraulic Loading Rates
We have also been interested in using numerical models to investigate water flow in on-site wastewater system (OWS) trenches. An estimate of the final steady wastewater infiltration rate is needed to evaluate the suitability of soils for installing OWSs and to determine the "design hydraulic loading rate" (HLRD). Our objective has been to develop a method for estimating the HLRD based on soil and biomat hydraulic properties. We have been using HYDRUS to model two-dimensional water flow from a trench in 12 soil textural classes. To simulate normal operating condistions, we assumed 5 cm of wastewater was ponded in the trench. We used two sets of boundary conditions at the bottom of the soil profile: a deep water table and a shallow water table. We also tested how well a simple equation developed by Bouma et al. (1975) estimated the bottom flux. With a modification to account for unsaturated flow within the biomat, the Bouma equation produced remarkably accurate estimates of trench bottom flux for all soil textural classes using a deep water table. Once the bottom flux is known, it can be converted to a HLRD by using a safety factor (we took 50% of the steady trench bottom flux as our safety factor). Water contents surrounding a trench in a sand profile are shown below after 30 days of infiltration. Only half of the trench is modeled since we assume that water flow is symmetrical around the trench centerline. The model space is 300 cm wide and 100 cm deep. The soil surface is at the top of the model space.
To estimate the trench bottom flow using the modified Bouma equation, we have developed an Excel Spreadsheet (Bouma Calculator) and a guidance word document.
Evaluation of Gravel-Filled Drainfield Trench Hydraulics and Treatment Efficiency
Drainfield sizing in Georgia is based on a design flow rate of 150 gpd/bedroom and a percolation rate for the soil where the drainfield will be installed. Percolation rates are typically estimated based on soil properties and converted to a long-term acceptance rate (LTAR) that incorporates a hydraulic reduction caused by biomat formation. The LTAR sizing method assumes that the wastewater loading rate is less than the LTAR of the trench bottom and that little wastewater exfiltrates through the trench sidewalls. Currently sidewall area is considered a safety factor when determining drainfield size and is not considered in the drainfield design. This is an unknown safety factor with no scientific support.
Biomats are recognized as having a major impact on both long-term hydraulic function and treatment efficiency of soil-based OWTSs. Many studies have addressed the reduction of wastewater infiltration rate with biomat development on intact core samples taken from mature OWTSs. Few studies, however, have addressed biomat-induced changes in infiltration as newly installed drainfield trenches become operational or biomat development for trench sidewalls. There are also no studies that show the most effective biomat density to wastewater infiltration ratio for optimal wastewater treatment.
As land prices increase and pressure increases to minimize drainfield size, incorporation of a known safety factor applied to accommodate sidewall exfiltration should be considered. Managing biomat formation rates and biomat density may also be an effective method for achieving optimal wastewater infiltration and treatment. An experimental OWTS installed in a Pacolet soil in Griffin, GA will be evaluated over the next 18 months. We will monitor soil hydraulic conditions both in the trenches and at several distances and depths down-gradient from the trenches using tensiometers and time-domain reflectometry (TDR). Wastewater treatment efficiency by the soil will also be monitored in and around the trenches using suction lysimeters. Data collected from this study will be incorporated in to a simulation model (HYDRUS-2D) to evaluate design loading rates for drainfields installed in loamy and clayey soils common in the Piedmont.
UGA Botanical Gardens- Nitrate contamination
Sampling of the Orange Trail Creek at the State Botanical Gardens by the Upper Oconee Watershed Network in April of 2004 found relatively high concentrations of nitrate-nitrogen. The College of Agricultural and Environmental Sciences subsequently funded a study by UGA scientists to determine the source of nitrate. Seven groundwater monitoring wells were installed in the area surrounding the creek and sampled during the summer of 2005. Soil core samples were taken at each site where monitoring wells were installed and several additional areas where shallow rock prevented installing ground water wells. Additional stream samples were also taken. Nitrate-nitrogen concentrations in the monitoring wells ranged from 4.3 to 19.7 mg/L. It appears that there are three sources of nitrate in the local area. One apparent source is one or more of the four wastewater lagoons at the UGA Swine Farm. The lagoon(s) appear to have leaked wastewater into fractures beneath the lagoon(s) where groundwater moves rapidly to springs that are headwaters of the eastern branches of the Orange Trail Creek. A second apparent source is the spray field at the UGA Swine Farm where wastewater from the lagoons has been applied. Groundwater movement from this area is likely to be the source of high nitrate in the northeastern branch of Orange Trail Creek. A third source appears to be coming from the hill north of Orange Trail Creek where the Poultry Science Complex is located and causing high nitrate concentrations in the north west branch of Orange Trail Creek. The identity of this source is unknown, but it may be poultry litter that was stacked in this area in the past, or wastewater pumped from a lagoon farther north and applied to a spray field on the south face of the hill. The nitrate in the creek and local ground water does not impose a threat to human health in that no drinking wells are installed in this area. However, steps should be taken to reduce nitrate concentrations where Orange Trail Creek enters the Oconee River (a constructed wetland is suggested), to properly close down the lagoons on the Swine Farm, and to identify the exact source of nitrate at the Poultry Science Complex.