Research Project: Impacts of Variable Land Management and Climate on Water and Soil Resources

Location: Agroclimate and Hydraulics Research Unit

2024 Annual Report

Objectives
Objective 1: Evaluate and compare performance of alternative and business as usual agricultural production systems with respect to water availability, selected environmental indicators, and productivity. Sub-objective 1.A: In accordance with the LTAR Common Experiment G x E x M framework, identify, quantify, and elucidate differences between BAU and ASP systems in terms of concentrations of N and P in surface water runoff, water use efficiencies (WUE), and pre-planting soil water content. Sub-objective 1.B: For the BAU and ASP systems, investigate impact of agricultural system (M and G) and climate (E), on soil microbial community activity and structure and subsequent changes in fractions of soil C and N, soil water content, and quality of surface water runoff. Objective 2: Develop hydrologic modeling tools to improve evaluation of the effects of land management, conservation practices, and climate variability on water and soil resources for agricultural practices in the Southern Plains. Sub-objective 2.A: Develop process-based distributed hydrologic and transport models to evaluate changes in soil and water quality under different management practices. Sub-objective 2.B: Evaluate and compare performance of RUSLE2 and WEPP models to predict soil loss under different land management systems. Sub-objective 2.C: Quantify the impacts of static vs dynamic land use on hydrologic model simulation performance. Sub-objective 2.D: Develop, incorporate, and evaluate new irrigation algorithm in SWAT to improve water budget predictions for increasing accuracy of water quantity and quality simulations. Objective 3: Develop, implement, or evaluate artificial intelligence, remote sensing, and spatial analysis tools for quantifying water, soil, and plant variables. Sub-objective 3.A: Develop a predictive and eXplainable artificial intelligence (XAI) framework to quantify climate-related risks for water resources and crop production and to assess adaptation pathways in the Southern Great Plains. Sub-objective 3.B: Use field-based radiometry to develop an in-the-field technique that would facilitate application in field research, increase the timeliness of results, reduce laboratory chemical wastes, reduce costs, and increase sample analysis throughput of soil and plant samples. Sub-objective 3.C: Use a GIS-based Revised Universal Soil Loss Equation (RUSLE) linked to a sedimentation module to predict (estimate) reservoir sedimentation.

Approach
This research is guided primarily by two USDA national research initiatives, the Conservation Effects and Assessment Program (CEAP) and the Long-Term Agroecosystem Research (LTAR) network and builds upon the prior 5-year project. The project is structured around three interrelated research objectives that: 1) improve the understanding of the impact of agricultural production systems on water availability, selected environmental indicators, and productivity, 2) develop hydrologic modeling tools to improve evaluation of the effects of land management, conservation practices, and climate variability on water and soil resources, and 3) develop artificial intelligence, remote sensing, and spatial analysis tools for quantifying water, soil, and plant variables. Objective 1 primarily deals with LTAR research goals. Objectives 2 and 3, though both tool-based, differ in their research thrust. Objective 2 is driven by the CEAP program goals and objectives, and addresses hydrologic model improvement, assessment, and environmental applications. Objective 3 is not limited to hydrologic models but seeks to develop, apply, or assess remote sensing, geospatial (e.g., Geographical Information Systems), machine learning, and data driven methods to address agricultural and natural resources problems. Our long-term goal is to elucidate system-wide performance indicators of the impacts of land management and climate variability on water, soil, productivity, and other ecosystem services at farm, watershed, and regional scales using long-term field, farm, and watershed research sites as the primary outdoor laboratories. Research approaches include field studies, remote sensing analyses, mathematical and statistical assessment of climate, and plot-to-watershed scale modeling. This research will assist agricultural producers, landowners, and governmental action agencies to contribute towards adopting more resilient and sustainable agricultural production systems by providing knowledge and tools that help them evaluate and optimize multiple management objectives for mixed-enterprise agricultural systems.

Progress Report
Objective 1: Evaluate and compare performance of aspirational and business-as-usual agricultural production systems with respect to water availability, selected environmental indicators, and productivity. We installed automated soil moisture sensors and an additional eddy covariance system to measure water, energy, and carbon flux among plant, soil, and atmosphere on the Grazinglands Research agroEcosystems and the ENvironment (GREEN) Farm. We conducted first year monitoring of field soil water content and plant biomass (or yield) from business-as-usual (BAU) and aspirational (ASP) agricultural systems in the Water Resources and Erosion (WRE) unit watersheds and on the GREEN farm. We continued to collect and process precipitation, soil water content, and microbial community samples (prepped for microbial extraction of DNA). We collected, processed, and ran soil samples on soil carbon (C) and nitrogen (N) fractions to quantify effects of soil microbial community structure and activity and available water holding capacity (AWC) in BAU and ASP agricultural systems. The water potentials between wilting point and field capacity as well as AWC vary with the effects of management and differences in soil C fractions. Part of the carbon data is included in a cross Long-Term Agroecosystem Research (LTAR) network C budget project. Sub-objective 2A.1: Evaluation of MIKE-SHE hydrological model. Research continued the feasibility of enhancing livestock production while mitigating its impact on grassland ecosystems and surface water quality. Nitrogen transport was simulated after development of a coupled physically based distributed model, namely MIKE SHE. The model was calibrated and validated using measured data from the WRE watersheds in El Reno, Oklahoma. Strategic matching of grazing activities with prevailing weather patterns shows the potential to increase livestock production while promoting environmental sustainability in pasture management. Sub-objective 2B,2C, 2D: Compilation of measured data for model calibration and evaluation (RUSLE, WEPP, SWAT models). Crop management practices for WRE watersheds, soil properties including soil texture, density, plant water available storage, and water potentials, precipitation breakpoint data (time-depth pairs), and measured surface runoff and soil loss data during 1980-2010 were compiled for running Revised Universal Soil Loss Equation (RUSLE) and Water Erosion Prediction Project (WEPP) model. Nine land use maps for the Fort Cobb Reservoir Experimental Watershed (FCREW) were developed using data from the Multi-Resolution Land Characteristics Consortium company to be used for Soil and Water Assessment Tool (SWAT) model research to determine the impact of dynamic land use on model performance and outputs. All model input and model evaluation data including measured irrigation data for the improvement of the irrigation process in the SWAT model were collected and processed. Great progress has been made in the development and incorporation of this new irrigation method into SWAT. Research collaboration with ARS in Kimberly, Idaho, and University of Idaho to test current and improved irrigation routines in SWAT continues. Sub-objective 3A.1,2,3: Developing the Artificial Intelligence (AI) pipeline, collecting input data, and acquiring climate projections from the Coupled Model Intercomparison Project Phase 5 (CMIP 5). Available soil moisture and available crop data from the National Agricultural Statistics Service (NASS) were collected and used in the development of a predictive artificial intelligence tool to quantify climate-related risks for water resources and crop production in the FCREW. In addition, available climate and flow data were collected and used to examine and assess historical droughts across the United States of America. Peer-reviewed manuscript has been prepared. The findings will be used to inform the impact of drought on agricultural production. Sub-Objective 3A.2: (1) Generating rolling monthly climate forecasts. Progress continues in developing and running continuous forecasts of precipitation using the K-Nearest Neighbor (KNN) seasonal forecast tool. Rolling monthly forecasts for precipitation have been completed since September 2023 using a dataset compiled from the Fort Reno Cooperative Observer Network (COOP) dataset and the El Reno Mesonet dataset (for more recent years) and gap filled using nearby station data as necessary. Instead of focusing on a single method of using the KNN forecast tool, it was determined to use three separate calibrations of the model to produce a range of forecasts, then identify the unique forecast values to develop a range of forecasts for each month. This was done to incorporate a broad range of forecasts into the rolling monthly forecasts instead of focusing only on a single iteration of the KNN model. Outside of the extreme precipitation totals that occurred in mid to late 2023 (June, July, and December 2023), this system has correctly identified (while Climate Prediction Center (CPC) was incorrect) the hydrometeorological class (above, below, or close to average precipitation totals) or matched the CPC forecast for the El Reno Mesonet location in all months. So far, in 2024, the KNN forecast tool has beaten climatology in all months and produced a near-perfect hydrometeorological class forecast for all months. While more calibration of the final forecast tool appears necessary (extremes are still quite difficult to catch with the current system) it appears the tool is showing skill in reproducing the hydrometeorological conditions for each month and thus will be useful when incorporated with the Decision Support Systems for Agrotechnology Transfer (DSSAT) wheat growth model later this year. Sub-Objective 3A.2: (2) Executing wheat grazing model for decision support using monthly forecasts. The KNN forecast tool will be used to create a suite of 30-day precipitation forecasts with a lead time of one month. The forecasted analogue years (i.e. dates) will be used to predict future agricultural support decision processes within the GREEN farm agricultural test site at El Reno, Oklahoma, starting with the 2024 winter wheat growing season. Sub-objective 3A.3: Assessment of climate change impacts under various cropping and tillage systems. A total of 29 combinations of cropping and tillage management systems were compiled for simulating the impacts of climate changes on surface runoff, soil erosion, and crop production under various Oklahoma common management practices using the WEPP model. Four tillage systems and six crops (five annual and one perennial) either in monoculture or rotation were combined to form 29 different cropping and tillage systems. Twenty-five Global Climate Models (GCMs) projections were selected from the CMIP5 projections based on their skills in simulating extreme precipitation events for use in simulating climate change impact with WEPP. Two computer programs (i.e., two weather generators) were used to downscale 25 GCM projections at a large scale to the Weatherford station for the two future periods (2021-2050 and 2051- 2080) under moderate and high warming conditions. In total, 200 climate scenarios (25 GCMs×2 periods×2 generators× 2 greenhouse gas levels) were used to run the WEPP model for all 29 cropping and tillage systems. The results showed that annual precipitation during 2021-2080 would decline by several percent points in central Oklahoma in both moderate and high warming conditions, leading to a decline in all crop yields except for cotton. However, total annual runoff and soi loss would tend to increase in future, compared to the same cropping and tillage systems at present, due to increases in extreme precipitation. No-till systems were most effective in reducing soil erosion among all tillage systems under climate change. Small grains were better than row crops in controlling erosion in future. Annual crops in a rotation with alfalfa could lower soil erosion significantly. The findings are useful in selecting effective conservation systems under future climates for the study region. Sub-objective 3B1,2,3: Measurements of C and N contents using hyperspectral reflectance. Quantification of soil organic C (SOC) is required for evaluation of soil erosion, assessment of land management practices, soil C model development and evaluation, and assessment of soil health. In a previous study, ARS researchers at El Reno, Oklahoma, evaluated four proximal sensors for their ability to estimate SOC from soil reflectance in the 400-1000 nanometer region. For oven-dried, ground, and sieved soils the calibration r2, reflecting the variance explained by the regression model, ranged from 0.73 to 0.93 for 3 of the radiometers (r2 equals 1 for a perfect model). The measurement technique is being further evaluated for consistency by considering the impact of soil sample treatment on the accuracy of SOC estimation using the 400-1000 nanometer wavelength range. Root samples have been collected, washed and are being stored for hyperspectral reflectance analyses after root samples are run for carbon and nitrogen via dry combustion.

Accomplishments
1. Revised Universal Soil Loss Equation (RUSLE)-Geographical Information System (GIS) predicts reservoir sedimentation. Watersheds in Oklahoma and the Southern Great Plains are prone to drought and periodic flooding. In response, the United States Department of Agriculture (USDA) Natural Resources Conservation Service (NRCS) aids in constructing flood control reservoirs. Current estimates of soil erosion and sediment yields indicate that ~50% of the reservoirs could be near the end of their projected service life due to sedimentation. Directly surveying the amount of sediment in a reservoir is the most common means used to determine whether rehabilitation of a structure is necessary but is costly and labor intensive. Modeling is a good choice for conducting preliminary assessments, especially in the context of prioritizing watersheds for follow-up investigations. The empirical Revised Universal Soil Loss Equation (RUSLE) model has been adapted to Geographical Information System (GIS) frameworks to study the spatial variability of soil erosion resulting from an assortment of land uses and landscapes and is used to estimate reservoir sedimentation. ARS Researchers in El Reno, Oklahoma, assessed the value of adjusting RUSLE-GIS model by using a regional representative proportion of gully/stream bank erosion contributions and by including a sediment delivery ratio (defined as percentage of the eroded materials that has been transported out of a watershed) in 12 watersheds and their accompanying reservoirs across Southwestern Oklahoma. The results showed that the modification of the RUSLE-GIS model has potential for improving estimation of reservoir sedimentation. NRCS staff, land managers, agricultural engineers, and soil conservationists may use this modeling approach to estimate reservoir sedimentation and prioritize reservoirs assessments of function and safety.

2. Simulated the impact of climate change and agricultural management systems on surface runoff, soil moisture, and soil erosion. The Long-Term Agroecosystem Research Network (LTAR) was established to develop a national roadmap to enhance or maintain agricultural productivity while achieving desirable environmental goals in the face of a changing and variable climate. Weather variability poses unprecedented challenges due to the increased frequency and duration of extreme events that alter patterns of precipitation and air temperature. The climate of the Southern Plains (SP) region and LTAR site are defined by droughts, intermittent flooding, heavy rainfall and wide temperature shifts that pose a threat to rainfed production systems in the region. ARS researchers in El Reno, Oklahoma, used computer models (Soil and Water Assessment Tool and a weather generator) to investigate the effect of projected climate on surface runoff, soil moisture, and erosion in three land use and management systems: continuous winter wheat under conventional tillage (baseline system), continuous winter wheat under no-till, and cool and warm season forage cover crop mixes under no-till. General Circulation Model (GCM)-projected future climate indicated an increase in average daily temperature (1.7-2.1 degrees Celsius) and potential evapotranspiration (6-7%) and a reduction in precipitation (3-10%) in the SP region. The projected future climate resulted in lowered soil moisture and greater erosion in the tilled continuous winter wheat system. In contrast, the no-till winter wheat and no-till forage cover crop systems significantly reduced surface runoff and erosion while conserving soil moisture. Incorporating cover crops significantly reduced surface runoff (57-73%) and soil erosion (87-91%) while preserving soil moisture. These findings may encourage producers to implement cover crops and/or no-tillage in their current management systems thereby reducing soil loss, increasing sub-surface soil moisture conditions and contribute to sustainable agriculture in the coming decades under projected climate conditions.

Review Publications
Starks, P.J., Moriasi, D.N., Fortuna, A. 2023. GIS-based RUSLE reservoir sedimentation estimates: Temporally variable C-factors, sediment delivery ration, and adjustment for stream channel and bank sediment sources. Land. 12(10). Article 1913. https://doi.org/10.3390/land12101913.
Whitesel, D., Mahmood, R., Flanagan, P.X., Rappin, E., Udaysankar, N., Pielke Sr., R.A., Hayes, M. 2024. Impacts of irrigation on a precipitation event during GRAINEX in the high plains aquifer region. Agriculture Forest Meteorology. 345. Article 109854. https://doi.org/10.1016/j.agrformet.2023.109854.
Zhang, X.J., Busteed, P.R. 2024. Accuracy and sensitivity of soil erosion estimation using 137Cs technology: a statistical perspective. Geoderma. 444.Article 116863. https://doi.org/10.1016/j.geoderma.2024.116863.
Xu, W., Chen, J., Xu, C., Zhang, X.J., Xiong, L., Liu, D. 2024. Coupling deep learning and physically-based hydrological models for monthly streamflow predictions. Water Resources Research. 60. Article e2023WR035618. https://doi.org/10.1029/2023WR035618.
Zhang, X.J. 2023. Evaluating and improving 137cs technology for estimating soil erosion using measured soil loss during 1954-2015. Earth-Science Reviews. 247. Article 104619. https://doi.org/10.1016/j.earscirev.2023.104619.
Wang, K., Li, J., Zhou, Z., Zhang, X.J. 2023. Editorial: Soil degradation and restoration in arid and semi-arid regions. Frontiers in Environmental Science. 11. Article 13007500. https://doi.org/10.3389/fenvs.2023.1307500.
Shi, H., Zhao, Y., An, X., Zheng, F., Liu, G., Li, H., Zhang, X.J., Pan, X., Wu, B., Wang, X. 2023. Tracing soil erosion with Fe3O4 magnetic powder: Principle and application. International Soil and Water Conservation Research. 12:419-431. https://www.sciencedirect.com/science/article/pii/S2095633923000643.
Lee, S., Moriasi, D.N., Danandehmehr, A., Mirchi, A. 2024. Sensitivity analysis of standardized precipitation and evapotranspiration index (SPEI) to probability distributions and potential evapotranspiration methods. Journal of Hydrology. 53. Article 101761. https://doi.org/10.1016/j.ejrh.2024.101761.
Lee, S., Moriasi, D.N., Fortuna, A., Mirchi, A., Danandeh Mehr, A., Chu, M.L., Guzman, J.A., Starks, P. 2024. Modeling the impact of measured and projected climate and management systems on agricultural fields: Surface runoff, soil moisture, and soil erosion. Journal of Environmental Quality. https://doi.org/10.1002/jeq2.20565.
Nelson, A.M., Maskey, M.L., Northup, B.K., Moriasi, D.N. 2024. Calibrating Agro-Hydrological Model under Grazing Activities: Challenges and Implications. Journal of Hydrology. https://doi.org/10.3390/hydrology11040042.
Samimi, M., Mirchi, A., Taghvaeian, S., Moriasi, D.N., Sheng, Z., Gutzler, D., Alian, S., Hereema, R., Wagner, K., Hargrove, W. 2022. Adapting irrigated agriculture in the middle Rio Grande to a warm-dry future. Journal of Hydrology: Regional Studies. 45. Article 101307. https://doi.org/10.1016/j.ejrh.2022.101307.
Liu, J., Chen, J., Zhang, X.J. 2023. Reliability of simulating internal precipitation variability over multi-timescales using multiple global climate model large ensembles in China. International Journal of Climatology. 43(14):6383-6401. https://doi.org/10.1002/joc.8210.
Mehata, M., Datta, S., Taghvaeian, S., Mirchi, A., Moriasi, D.N. 2023. Effects of soil data accuracy on outputs of irrigation scheduling tools. Journal of the ASABE. 66(3):677-687. https://doi.org/10.13031/ja.15323.
Singh, A., Taghvaeian, S., Mirchi, A., Mansaray, A., Alderman, P.D., Moriasi, D.N. 2023. Analysis of climatic trends in climate divisions of Oklahoma, U.S. Applied Engineering in Agriculture. 39(2):167-177.
Singh, A., Taghvaeian, S., Mirchi, A., Moriasi, D.N. 2023. Station aridity in weather monitoring networks: Evidence from the Oklahoma Mesonet. Applied Engineering in Agriculture. 39(2):167-177.