Evaluation of Saturated Buffers as a Conservation Drainage Practice for Treating Agricultural Subsurface Drainage
Saturated buffers are an emerging conservation practice that can help meetIowa’s goals of reducing nutrient export by treating agricultural runoff in subsurface drainage systems. Although these buffers can reduce nutrient discharge effectively, their design and implementation are complicated by specific site conditions, including topography, soil properties, depth of restrictive layer, receiving channel depth, site ownership, and presence of buried utilities. A need exists to gain a better understanding of subsurface flow interactions to increase the suitability of saturated buffers at candidate sites.
To improve the understanding of the function of saturated buffers and aid in their design, this project will develop a model of groundwater flow and nutrient transport through saturated buffers; devise and test an approach for estimating the horizontal hydraulic conductivity; and identify and characterize the key processes that allow saturated buffers to reduce export of nitrogen and phosphorus.
A model of groundwater flow and nutrient transport will be developed, and then evaluated with field measurements from sites with existing saturated buffers. Once confidence in the model is established, scenarios involving different flows, temperature, water levels, and nutrient concentrations will be examined to assess the main processes that remove nutrients. The outcome of this research will be a better understanding of groundwater hydrology, which can be used to develop scientifically sound design guidance for saturated buffers.
For saturated riparian buffers (SRBs) that allow overflow to the stream (as in the six sites we have investigated in this project), our findings indicate that a wider buffer is not necessarily more effective. The optimal width is determined by balancing the need to allow time for nitrate removal with the need to divert as much of the tile flow as possible to the SRB. We find that the optimal width can be smaller than the width specified in USDA-NRCS Code 604.
A three-dimensional, finite-difference groundwater model of a generic SRB was developed to assess the flow of groundwater and the implications for nitrate removal. The study resulted in four main points that will help with enhancing the design of SRBs:
(1) Groundwater flow is three-dimensional near the distribution pipe and the stream and primarily one-dimensional in the rest of the buffer.
(2) The path the water takes in flowing toward the stream depends on where it exits the distribution pipe.
(3) The potential for nitrate removal on each path depends on the length of the path—and thus travel time—and depth because of variations in denitrification potential with depth.
(4) The travel time can be estimated well by slightly modifying a one-dimensional approximation.
Progress is being made on modeling nitrate fate and transport, with the help of a new student. The transport model MT3D to compute fate and transport is being applied in an unconfined aquifer of dimensions typical of SRBs, and the results from the study of the groundwater flow will be used to determine conditions likely to maximize the removal of nitrate.
Outreach included a presentation at the 2020 Soil and Water Conservation Society International Conference and publication of a journal article, “Improving the Effectiveness of Saturated Riparian Buffers for Removing Nitrate from Subsurface Drainage” in the Journal of Environmental Quality. Graduate students Andrea McEachran and Loulou Dickey also defended related master's theses on saturated riparian buffers.
The research team’s findings show that for saturated riparian buffers (SRBs) that allow overflow to the stream (as in the six sites investigated in this project), a wider buffer is not necessarily more effective. The optimal width is determined by balancing the need to allow time for nitrate removal with the need to divert as much of the tile flow as possible to the SRB. We find that the optimal width can be smaller than the width specified in USDA-NRCS Code 604 (see figure).
Progress has also been made on the groundwater flow modeling aspect of the project. A paper on this topic is being developed that will describe how groundwater flows in a saturated buffer and the implications for nitrate removal. Although three-dimensional flow occurs in the SRB, much of the flow is one-dimensional, and a simple expression for the time for water to travel through the buffer applies well. This latter point is important for designing SRBs and estimating nitrate removal. The modeling also reveals the paths that water exiting the distribution pipe takes, which allows investigation of tradeoffs between residence time and depth of organic carbon in nitrate removal.
Two graduate students presented findings from the project at the International Erosion Control Association Annual Conference in Raleigh, North Carolina, February 26th, 2020.
Progress has been made in many aspects of the project. Students have traveled to each of the sites to continue gathering data, which primarily consists of downloading water-level data from monitoring wells to calibrate the groundwater model. Significant progress has been made in the groundwater modeling during this reporting period. The model setup has progressed to better represent the sites being modeled with increasing complexity over time. The final models will be used to understand how groundwater and nitrate travels in a saturated buffer. Papers will reflect insights gained from the groundwater modeling, including what it reveals about nitrate flows in a saturated buffer and how the information can be used to most effectively design saturated buffers for nitrate removal (such as tradeoffs between the benefits of a wide buffer and a narrow buffer).
During this period, progress was made in obtaining field data, writing a journal article for effectively designing saturated buffers to remove nitrate and reviewing relevant literature. Students visited five of the six sites during the spring. Completed field work includes obtaining water level data from transducers, gathering water samples, slug testing and surveying the buffers and stream. Drone photogrammetry and lidar scanning were conducted at the site with the most severe bank erosion. Soil cores were obtained from three sites, although the cores are of poor quality due to compaction caused by saturated field conditions in the early spring. Important field work remaining includes obtaining soil cores, extracting pressure transducer data, and performing slug tests at the rest of the sites.
Some progress has been made this quarter in using MODFLOW to model groundwater flow and nitrate transport. We are still obtaining the input values for the model from the sites and addressing challenges in the modeling.
Researchers worked through most of the calculations and are working on a paper describing how to effectively design saturated buffers for nitrate removal. A larger buffer width means there will be a longer residence time, so more of the nitrate entering the saturated buffer will be removed, while a shorter buffer width increases the flow rate entering the saturated buffer, therefore reducing nitrate-rich tile drainage that will overflow to the stream untreated. This tradeoff indicates that there is an optimal width for removing nitrate in saturated buffers, which is the focus of the paper.
Throughout the second quarter, researchers gathered existing data, collected new data from six sites with saturated buffers, became familiar with individual site conditions and worked on the models.
The two undergraduate students involved in this project have graduated and will begin working as full-time graduate students starting in January 2019. The students have visited all of the sites with Kent Heikens, an agricultural science research technician with the U.S. Department of Agriculture who has worked with co-PI Tom Isenhart and collaborator Dan Jaynes. Each site has monitoring wells, which are now fully instrumented with pressure transducers to record water level data to be used in developing the models. Topographic surveys have been conducted at each of the sites to relate the elevation of the water table to the readings from the pressure transducers and water samples collected regularly to determine nitrate concentrations. Several eight-foot cores were taken to determine soil properties, however, the soil has been too saturated at many sites to collect deep cores.
Researchers will wait until the soil is drier to retrieve cores that can be used to determine hydraulic conductivity with permeameter tests. Results will be compared with results from slug tests, which have been started at the Bear Creek site. Past data gathered from the saturated buffer sites includes flow rates from the tile drains, monitoring well dimensions, water levels, nitrate concentrations, soil type, soil moisture and more. The groundwater and nitrate models will rely largely on this data, but water levels, nitrate concentrations and slug test data being gathered will supplement this data set. Dr. Isenhart has provided electrical resistivity data for one of the buffers at Bear Creek and the buffer at the Maass Farm. This data provides insight into heterogeneity of the aquifer and the elevation of the bottom of the aquifer, which is important in developing the groundwater model.
Next steps include understanding how to use this data and how to extend the information to the other sites. We have started developing a set of groundwater models of increasing complexity for the saturated buffers that will incorporate travel time, water table elevation and contaminant transport through a saturated buffer. These models will serve as benchmarks for the numerical models that incorporate more complexity.
During this first quarter of the project, a project team was built, sites were identified , a sampling plan developed, data gathered, , and modeling worked on. Students were recruited to assist with the project. . Another undergraduate. Tyler Groh, who is a postdoctoral researcher with PI Isenhart, and Kent Heikens of the USDA are sharing their extensive experience with saturated buffers and helping to guide the project. Sites for field work were prepared for data collection. The sites include Maass Farm, Shearer Farm, Dysart Farm, two buffers near Bear Creek and a site at Hickory Grove. Three sites were visited and data gathered on water levels, flow, soil properties and nutrient concentrations and loads. Further measurements will augment the existing data set. As noted in the proposal, hydraulic conductivity will be measured with permeameter and slug tests. We have developed a method to calibrate the hydraulic conductivity and elevation of the aquifer bottom by using the water- level data. Data collection on nutrient concentrations will leverage the efforts of Mr. Heikens. He is training the assistants on collecting, handling, and analyzing samples and the three have developed a plan to sample at the sites. Modeling has started with an analytical model of saturated buffers and initial steps with the groundwater model MODFLOW. Although idealized, the analytical model of flow helps to identify the key parameters and build intuition; for example, it has highlighted the need to estimate the elevation of the aquifer bottom and led to the approach for calibrating aquifer properties from the water levels.