Location: Watershed Physical Processes Research
2025 Annual Report
Objectives
1. Develop and implement acoustic based methods for measurement and interpretation of sediment transport in streams.
1.A. Develop novel methods for data analysis and visualization to aid interpretation of acoustic sedimentation data while continuing to develop sediment transport measurement technologies.
1.B. Engage in CEAP/LTAR research by implementing existing acoustic sediment monitoring technology in CEAP/LTAR watershed.
2. Develop and adapt acoustic and geophysical methods for characterizing soils and monitoring processes within the agricultural watershed.
2.A. Develop an acoustic based soil water status assessment tool for improving water management and irrigation and rain fed decision support systems.
2.B. Develop rapid and noninvasive agrogeophysical methods for mapping and monitoring erosional processes (e.g., soil pipes in relation to gully erosion) in agricultural landscapes.
2.C. Application of geophysical measurements for estimating groundwater flow, aquifer parameters, and aquifer thickness.
2.D. Development of relationships between soil properties and geophysical attributes using machine learning.
Approach
Development of acoustic and geophysics technology addressing gaps in the USDA's suite of tools to map and monitor hydraulic processes over a range of time and space scales will consist of: theoretical and modeling efforts, controlled laboratory experiments, and field measurements using the newly developed hardware and techniques. Theoretical and modeling efforts establish the feasibility and sensitive of acoustic attributes to soil processes and sediment transport. Laboratory measurements help to better understand the physics of soils and its interaction with water and determine optimal sensor configurations, data quality requirements, and data processing schemes. Field measurements provide the final proof of concept design and incorporation into USDA applications. The first objective relates to the development of novel methods for data analysis and visualization to aid interpretation of acoustic sedimentation data while continuing to develop sediment transport measurement technologies. The second objective relates to acoustic and geophysical methods that can monitor and evaluate the performance of agricultural irrigation, drainage, and rain-fed systems, improve technology for studying soil pipe development, and to assess suitable sites and monitor the efficacy of surface water to groundwater interaction. Both parties are actively engaged in independent research projects related to the development and use of acoustic/ seismic technology for water resources applications. The parties agree that meeting the objectives of this project will expand the suite of tools, technology, and sensors for acquiring data to support science-based decision support systems.
Progress Report
Progress on Objective 1A,laboratory experiments were conducted at the National Center for Physical Acoustics to isolate the effect of algal concentration on the Single Frequency Acoustic Attenuation System measurement since acoustic observations at the University of Mississippi Biological Field Station limno-corrals and the Rio Grande deployment in San Acacian, New Mexico, indicate real variations in acoustic signals. Analysis of the acoustic data alongside the available ancillary data indicated multiple parameters that may affect the acoustic signal, primarily light intensity, temperature, and dissolved oxygen. Algae concentrations were cultivated by the University of Mississippi Department of Biology and placed in a 10-gallon fish tank with a dissolved oxygen sensor, thermometer, and the Single Frequency Acoustic Attenuation System. Aquarium lights were placed on timers to simulate the daily solar clock, and heaters were used to regulate water temperature. Physical samples were extracted and placed in a Turner Designs Trilogy Laboratory Fluorometer to measure the relative fluorescence units in relation to the co-located acoustic signals. From previous laboratory tests, it was determined that neither light intensity nor dissolved oxygen has an independent influence on the acoustic signal. It was demonstrated that the acoustic signals were correlated with changes in algal populations, and a correction fit was created to enhance the interpretation of acoustic signals when algal populations are present. This work was presented at the Mississippi Water Resources Conference in Flowood, Mississippi, on 10/10/2024.
Progress on Objective 1B: Following instrumentation maintenance and laboratory calibration, the Single Frequency Acoustic Attenuation System was redeployed in the Goodwin Creek Experimental Watershed at Stations 1 and 2. These sites are instrumented with numerous measurement systems and will provide more parameters with which the acoustic signal can be compared. As of redeployment in November 2024, the two systems have collected acoustic data over twelve high-flow events spanning approximately twenty-two days. Researchers are awaiting the 2024-2025 data from the physical samples collected by partners at the National Sedimentation Laboratory in Oxford, Mississippi. The raw acoustic data have been processed, time-stamped, and uploaded to a database, allowing for comparison with parameters such as flow discharge, stage height, and measured sediment concentration. This compiled 3-year database of field data has been uploaded to a research data catalog and repository for public access. A conference presentation was presented at the Acoustical Society of America in New Orleans, Louisiana, on May 21, 2025. A journal article on this work was published in the International Journal of Sediment Research on 6/9/2025.
Progress on Objective 2A
Soil water status measurements using the high-frequency multichannel analysis of surface waves soil profiler system were conducted continuously on the campus of the University of Mississippi, before, during, and after irrigation under different initial soil conditions. Temporal variations in instantaneous shear wave velocity profiles, extending up to 2 meters below the surface, are utilized to evaluate the initial soil condition, the ongoing soil response to irrigation, and drainage processes. Results from the study revealed significant changes in shear wave velocity, depending on initial soil conditions, irrigation intensity, and depth. Additionally, to validate the results from the surface wave method across different seasons and soil conditions, a systematic comparison was conducted between the temporal variations in shear wave velocity and the variations in moisture content at different depths during irrigation and rainfall. It is found that, at deeper depths, variations in moisture content lag those of shear wave velocity due to the longer time required for water to penetrate deeper. Moisture content is not a suitable parameter for validating dynamic changes in shear wave velocity, as evidenced by a low correlation coefficient of 0.2. This has led to a notable observation that the fluctuation in shear wave velocity could be governed by variations in water potential rather than moisture content. A study is currently ongoing using multiple tensiometers buried at different depths to monitor changes in water potential. Preliminary results show that variations in water potential show the same temporal variations as the shear wave velocity. While the system was successfully tested at a farmland irrigation site in Holly Springs, MS, plans are in place to include water potential measurements in upcoming field tests. A database of the test results of irrigation and rainfall has been established, and a journal manuscript of this work is currently in preparation.
Progress on Objective 2B,high-resolution drone elevation data were collected at Goodwin Creek Experimental Watershed. This area encompasses 54 acres of pastureland and 17 acres of farm field, featuring documented soil pipe formations and surface collapse features, including flute holes, sinkholes, and gullies. Previous studies indicated that the distribution of soil pipe collapses does not consistently agree with the flow paths related to surface topography. The initial analysis of the digital elevation model and the orthomosaic of the site at a 2-centimeter ground sampling resolution showed no consistent indication of a correlation between surface topography and the location of collapse features. This suggests that water-restrictive layers in the subsurface, such as fragipans, govern the formation of soil pipes more than topographic layout. These restrictive layers promote the lateral flow of water percolating from the surface, gradually creating internal erosion. A study is ongoing to further evaluate correlations between previously collected electromagnetic surveys at the site and surface topography. A possible correlation could lead to an improved zoning map using geophysical methods. A peer-reviewed paper on integrating multiple geophysical methods for identifying soil pipes was published in the Journal of Environmental and Engineering Geophysics, and a paper on a novel approach for detecting soil pipe networks using acoustic excitation is submitted for publication.
Progress on Objective 2C, research was conducted to evaluate the effectiveness of expedient electromagnetic induction techniques for estimating infiltration rates. Infiltration rate refers to the speed at which water enters the soil and moves downward through the soil profile. It is a crucial parameter in hydrology, agriculture, and environmental management. High infiltration rates regulate groundwater recharge, minimize surface runoff, reduce soil erosion, and enhance agricultural productivity by increasing soil water availability for crop growth. High-resolution electrical conductivity measurements were performed using two systems with different depths of investigation at a stormwater detention basin located at the University of Mississippi. A limited number of soil samples and double-ring infiltration tests were collected based on variations in electrical conductivity to estimate saturated hydraulic conductivities. Correlation between collocated electrical conductivity and saturated hydraulic conductivity was then established. This correlation transformed the high-resolution electrical conductivity data to high-resolution saturated hydraulic conductivity. Finally, a geophysics-aided hydrological model was used to convert the high-resolution saturated hydraulic conductivity map into an infiltration rate map, taking into consideration the site’s physical properties. A water-balance study using two rain events validated the geophysics-aided basin-scale infiltration values. This work was presented at the Symposium on the Application of Geophysics to Engineering and Environmental Problems in Denver, Colorado, on April 15, 2025, and at the Geological Society of America conference in Anaheim, California, on September 25, 2024. A paper on integrating electrical resistivity tomography and self-potential techniques for aquifer characterization was published in Groundwater.
Progress on Objective 2D,research was conducted to develop a machine learning-based pedotransfer function for predicting soil saturated hydraulic conductivity. Pedotransfer functions are equations that use basic soil properties, which are easier to measure, to predict soil properties that are harder to determine. A database of over 20,000 soils, including soil properties such as texture, bulk density, and organic carbon content, was generated from existing literature. Three machine learning techniques — artificial neural networks, boosted regression trees, and random forests — were compared to develop a new pedotransfer function for predicting saturated hydraulic conductivity. The models based on these machine learning techniques were evaluated, and the relative importance of different predictor variables in estimating saturated hydraulic conductivity was assessed. Among the evaluated models, decision tree-based ensemble models, particularly boosted regression trees, demonstrated superior performance, achieving a low root mean square error of 0.34 and a coefficient of determination of 0.87. Clay content emerged as the most significant predictor, followed by nearly equal contributions from sand content and bulk density. The best-performing machine learning-based pedotransfer function model, based on the boosted regression trees algorithm, was then applied to a gridded dataset of soil properties to produce a high-resolution map of saturated hydraulic conductivity for the Mississippi Alluvial Plain. This work was presented at the Geological Society of America conference in Anaheim, California, on 9/25/2024, and at the Mississippi Water Resources Conference in Flowood, Mississippi, on 10/10/2024.
Accomplishments
1. Rapid geophysical methods were used to find optimal locations for groundwater recharge methods. Agricultural irrigation accounts for the largest share of groundwater use in the United States. According to the National Agricultural Statistics Service, almost two-thirds of irrigated agricultural acreage in the United States relies on groundwater as a primary or secondary source. However, groundwater is a limited resource, and its depletion due to irrigation increases the need to identify optimal locations for groundwater replenishment and management approaches. Researchers at the University of Mississippi and ARS in Oxford, Mississippi, successfully applied electromagnetic induction methods to generate high-resolution spatial maps of infiltration rates, which will help identify optimal locations for groundwater recharge techniques. Areas with higher infiltration rates allow more water to reach the groundwater aquifers, and spatial maps showing optimal locations will lead to more accurate estimates of irrigation storage capacity and better design and placement of infiltration basins for aquifer recharge.
2. Machine learning utilized to generate maps of the ability of water to move through soils. The rate that water can move through soil, which is called hydraulic conductivity, is an essential parameter for understanding how water behaves after irrigation of crops or rainfall on watersheds, since it affects the amount of water that is absorbed by soil and the amount that is runoff. It is a fundamental input for modeling runoff, drainage, and movement of materials dissolved in water through soils. However, measuring the rate of water movement through soils is time-consuming and labor-intensive, making it expensive and impractical for large-scale agricultural fields. Researchers at the University of Mississippi and ARS in Oxford, Mississippi, successfully demonstrated that machine learning can be used to provide a method for generating high-resolution maps of saturated hydraulic conductivity using readily available soil data. Accurate prediction of saturated hydraulic conductivity using machine learning techniques will lead to more informed decisions for soil, crop, and irrigation management by farmers, as well as more accurate regional hydrogeological computer models.
3. Rapid geophysical methods used to map internal soil pipe networks in agricultural fields. Loss of fertile soil due to erosion is one of the most significant challenges to sustained agricultural food production and effective soil management. Although soil erosion due to surface processes has been studied widely, the contribution of internal soil pipes is often overlooked due to the difficulty in detecting them. University of Mississippi and ARS researchers at Oxford, Mississippi successfully implemented rapid, non-invasive, high-resolution geophysical methods to detect the subsurface network of internal soil pipes on agricultural fields. Additionally, analysis of high-resolution digital elevation models revealed that the locations of soil pipes were not always near where water flows on the ground surface. Results from geophysical surveys can be used to locate soil pipes in agricultural fields to help prioritize conservation efforts to reduce soil and nutrient losses from agricultural fields.
4. Signal correction for algae when using acoustic methods for measuring sediment in water. Sand and soil transported in rivers affect water quality, ecological processes, and navigation of channels. Methods for measuring sediment transported in rivers are needed, and acoustic systems can be an effective approach; however, it was found that high algae populations in some water bodies interfered with the acoustic measurements. To address these effects and further improve the accuracy of sediment concentration measurements, laboratory experiments were conducted at the University of Mississippi to quantify the effect of algae on acoustic signals and to find a way to correct the data so that more accurate measurements of sediment could be made. The work resulted in a relationship between acoustic signal loss and algae concentration that can be used to improve acoustic sediment measurement in the presence of algae.
5. Continued deployment of single frequency acoustic attenuation system at goodwin creek. Rivers that flow through agricultural lands carry sediment particles that affect flooding, restoration efforts, and ecological processes. It is difficult to measure sediments in rivers, and automated systems that can function without personnel on site are needed. A system that uses sound to measure sediment concentration was developed, and field testing has continued with two units installed in the Goodwin Creek Experimental Watershed near Batesville, Mississippi, USA. The systems collect data automatically and at a much higher rate than traditional manual methods, improving the quality of the sediment transport data. This enables researchers, scientists, and government agencies to have access to improved measurements of sediment transport that can be used to more effectively manage streams and rivers.
Review Publications
Mamud, M.L., Holt, R.M., Hickey, C.J., O’Reilly, A.M., Wodajo, L.T., Rad, P.B., and Samad A., 2024. Integrating ERT and SP techniques for characterizing aquifers and surface-groundwater interactions. Groundwater 63:265–279. https://doi.org/10.1111/gwat.13444
Carpenter, W.O., Goodwiller, B.T., Wren, D.G. 2025. Multi-year deployment of a single frequency high-frequency acoustic attenuation system for measuring fine suspended sediments in stream channels. International Journal of Sediment Research. https://doi.org/10.1016/j.ijsrc.2025.06.005.