Skip to main content
ARS Home » Pacific West Area » Maricopa, Arizona » U.S. Arid Land Agricultural Research Center » Water Management and Conservation Research » Research » Research Project #441595
Research Project: Improving Water Management for Arid Irrigated Agroecosystems

Location: Water Management and Conservation Research

2023 Annual Report


Objectives
In Arizona and the Southwest U.S., irrigation is essential for field crop production. However, long-term drought and increasing urban development have decreased water availability. Historically, surface irrigation has been the main irrigation practice. However, with current water limitations, development of modern irrigation systems, sensing technologies, irrigation management tools, and crop water use estimates are now critical. The overall goals of this project are to improve knowledge of water, nutrient, and crop growth processes in arid agroecosystems and to develop sensing, computing, and decision support technologies that improve water and nutrient use efficiency for crop production. Objective 1 details the main field component with field experiments, primarily in cotton and guayule, that test different models for irrigation scheduling with feedback from soil water content sensors and imagery from small unmanned aircraft systems (sUAS). Objective 2 focuses on use of satellite remote sensing to monitor and forecast regional crop evapotranspiration (ET). Objective 3 continues the development of tools for improved management of surface irrigation with a special focus on modeling field-scale infiltration processes. The over-arching philosophy of the research is to develop knowledge and tools based on the integration of proximal and remote observations with physical process-based and artificial intelligence-based models. Objective 1: Integrate sensor data and simulation models to improve irrigation and fertilization decision support for irrigated cropping systems. Sub-objective 1A: Develop remote, proximal, and in-situ sensing technologies for estimating crop, water, and nutrient status of irrigated agroecosystems. Sub-objective 1B: Develop and evaluate simulation models, machine learning algorithms, and data integration strategies that better inform crop management decisions. Sub-objective 1C: Develop and field-test decision support tools that integrate data and models for improving in-season crop management. Sub-objective 1D: Develop irrigation guidelines, tools, and models for direct-seeded guayule. Objective 2: Create and evaluate suites of satellite-based hydrology models that enable accurate monitoring and forecasting of evapotranspiration and other soil water balance components over irrigated agriculture, leading to improved irrigation scheduling. Sub-objective 2A: Develop and test crop coefficient models driven by remote sensing data. Sub-objective 2B: Develop and test algorithms that use remote sensing to track water budgets across multiple cropping seasons. Objective 3: Design, test and/or improve sensors and technologies for optimizing surface irrigation systems. Sub-objective 3A. Evaluate and improve infiltration modeling approaches for irrigation design and management, tied to the Natural Resources Conservation Service (NRCS) soils database. Sub-objective 3B. Develop design and management strategies that account for the spatial and temporal variability of conditions, including infiltration, hydraulic resistance, and flow rate.


Approach
Objective 1 Goal 1A: Develop novel sensing approaches and data pipelines for timely collection and processing of in-season agroecosystem data that can be immediately used for crop and soil management. Tools to prepare sensor data for immediate integration with irrigation scheduling algorithms are necessary because the data will be incorporated with decision models (Sub-objective 1.B) and used to inform irrigation management decisions for field studies (Sub-objective 1.C). Objective 1 Goal 1B: Develop simulation models or machine learning algorithms as tools to synthesize in-season field data and provide reliable recommendations for real-time or near-term crop management. Data collected during previous field studies will be used to evaluate model responses to experimental conditions. Objective 1 Goal 1C: Field-test decision support tools and methods for irrigation management with focus on identifying approaches that improve crop yield and water use efficiency. Irrigation management experiments will be continuously conducted for summer cotton crops and winter cover crops or small grains for the duration of the project. Objective 1 Goal 1D: Determine irrigation scheduling and timed water stress strategies for optimum rubber yields and water use efficiencies. Develop crop coefficient models and determine remote sensing indices for real-time Kcb and plant growth estimation. Develop a customized soil water balance (SWB) irrigation model that provides decision support for growers in the region. Objective 2 Goal 2A: Develop crop coefficients for all economically significant crops grown in Central Arizona. Priority crops will be cotton, alfalfa, potato, sorghum, barley, and corn. Years of evaluation are 2016 to current. Objective 2 Goal 2B: Develop and test remote sensing-based surface energy balance algorithms and incorporate them into an app tool for irrigation decision support. While research under Sub-objective 2.A focuses on answering questions about irrigation management for specific crops grown in Arizona, Sub-objective 2.B addresses research to improve management skill across entire districts and spanning multiple years. Objective 3 Goal 3A: Provide WinSRFR with additional infiltration modeling capabilities, namely an alternative to the NRCS furrow infiltration families, and Green-Ampt based soil models for two soil layers, for one- and two-dimensional infiltration. Two groups of activities will be undertaken as part of this subobjective: 1) new infiltration modeling options will be developed and added to the software and 2) studies will be conducted to further validate procedures for the estimation of GA and WGA infiltration parameters from irrigation evaluation data. Objective 3 Goal 3B: Provide WinSRFR users with additional capabilities for examining the uncertainty of model outputs as a function of uncertainty of variable inputs. The first part of the proposed work involves adding new options for examining the sensitivity of outputs visually. The second part will use data analysis tools to conduct uncertainty and sensitivity studies to develop quantitative measures for various synthetic scenarios.


Progress Report
This report documents progress in fiscal year (FY) 2023 for project 2020-13660-009-000D, “Improving Water Management for Arid Irrigated Agroecosystems, which began in January 2022. In support of Sub-objective 1A, ARS researchers in Maricopa, Arizona, continued analyzing the development of hyperspectral sensing techniques to inform status of chlorophyll content in a cotton breeding population. A field spectroradiometer was used to measure cotton leaf reflectance with concurrent collection and subsequent laboratory analysis of leaf punches for chlorophyll content. Machine learning techniques were developed using high-performance computing resources to relate leaf hyperspectral reflectance with leaf chlorophyll measurements. Estimates of leaf chlorophyll from these phenotyping methodologies will be included in genetic analyses to identify genes responsible for chlorophyll content in cotton. For Sub-objective 1B, research continued on the organization and curation of cotton data sets collected by ARS scientists at the Maricopa Agricultural Center (MAC) over the past 25 years. A data dictionary is being developed to codify terminology and units, and Python scripts are being developed to codify the required data reduction steps and workflows, for example the process of reducing raw seed cotton weights from harvest bags to cotton fiber and cottonseed yield on an area basis. The overall goal is to automate the organization of agronomic data in a standardized format, which will facilitate data integration with databases, machine learning algorithms, and other modern data science tools. Progress on Sub-objective 1C, included ARS researchers initiating a cotton field experiment in the 2023 growing season to compare agronomic outcomes (stand density, crop cover, yield and soil water status) for cotton grown with both conventional and conservation tillage. The treatments include 1) Cotton grown on conventional raised beds with full tillage, 2) Cotton grown on flat ground with full tillage, 3) Cotton grown with strip tillage in a terminated barley cover crop, and 4) Cotton grown with no tillage in a terminated barely cover crop. The experiment will demonstrate cotton responses to pre-plant field treatments, which from past experience has been known to affect growth and yield of sprinkler-irrigated cotton in Maricopa, Arizona. For Sub-objective 1D, field data for the ratooned, guayule regrowth studies in Maricopa, Arizona, and Eloy, Nevada, (completed in 2022) were analyzed by ARS researchers, and a manuscript was written. The studies indicate that rubber yield may be substantially greater for the ratooned crop when regrown for two years than that from an initial harvest after two years following seeding, Also, the substantial field operations and irrigation water use needed in establishing and growing the seeded crop for the first two years will be eliminated and reduced, respectively, for a ratooned guayule crop. Thus, ratooning guayule should provide growers with significant additional yield returns that would offset the initial crop establishment investment costs. In support of Sub-objective 2A, significant progress has been made developing crop coefficient curves for the economically important crops grown in Yuma Valley. Evaluation of crop evapotranspiration using eddy covariance and Sentinel 2 remote sensing of vegetation indices was completed for Fall-Winter-Spring vegetable crops: broccoli, cauliflower, celery, iceberg, romaine, spinach, and spring mix; and for Spring-Summer Rotational systems: cantaloupe, cotton, durum wheat, Sudan grass, and watermelon. Crop evapotranspiration studies for leaf lettuce, alfalfa, and citrus are underway. ARS researchers also acheived progress for Sub-objective 2B, in which a smart phone application,'DesertAgWise', was developed for individual growers to use to manage water budgets and near-surface salt levels over multiple years. The application has been developed for cotton, wheat, and lettuce and uses a salinity check book system to help track water needed for salt leaching events. Related research is underway in collaboration with ARS researcers in Riverside, California and scientists at the University of California, Riverside, to develop an artificial intelligence system that will use soil and water status across the southwestern United States to help manage any irrigated crop. For Sub-objective 3A, substantial progress was made in the development of alternatives for modeling furrow infiltration with the NRCS infiltration families. The infiltration families are generic infiltration vs time relationships developed by NRCS to model and categorize infiltration and used to support analyses of surface irrigation systems. The infiltration family analysis was conducted in comparison with predictions based on porous media flow theory under a wide range of soil textural conditions. A fundamental assumption of the original method is that two- dimensional infiltration can be derived directly from the one-dimensional families. The one-dimensional families do not represent specific soil textures, but rather, soils with similar infiltration relationships. Simulation tests have confirmed that soils with dissimilar textural characteristics can produce very similar infiltration relationships when considering one-dimensional flow only. However, two-dimensional infiltration relationships can be very different due to substantial differences in sorptivity, and thus, lateral flow. Since the Natural Resource Conservation Service (NRCS) method assumes that lateral flow effects are the same for any soil, but proportional to one-dimensional infiltration, it can substantially overestimate or underestimate infiltration in comparison with porous media flow theory. Relationships are being developed for developing texture-specific sorptivity estimates that can be incorporated into the Web Soil Survey. In support of Sub-objective 3B, ARS researchers are adding new analytical capabilities to the WinSRFR surface irrigation software package. The objective is to provide users with the capability of developing sensitivity contours based on two user-selected variables. Users will most likely use this capability to examine the sensitivity of performance as a function of varying infiltration parameters, within user-specified ranges for those parameters. However, the procedures will also allow users to test the effect of other input variables. The framework for computing irrigation performance data tables as a function of the two input variables has been completed. Development of the code that will convert the tabular results into graphical products was delayed due to other programming efforts. Additionally, for Objective 3, in response to error reports for the WinFlume 2 software package (design of long-throated flumes for open channels, released in 2020), a substantial effort was undertaken to correcting those problems. Previous tests failed to detect defects in the implementation of new features that were introduced in WinFlume 2. A systematic testing campaign was conducted to include a much wider range of configuration options than done during the initial development. The extensive testing revealed additional problems introduced with WinFlume 2, some of which were difficult to resolve due to constraints imposed by the underlying code for the hydraulic computations which was developed more than 20 years ago. The testing identified computational problems in the original code. Per the request of our collaborators, a new feature was added for the upcoming WinFlume 2.1 version. Researchers are currently resolving problems related to the development of the installer, but still expect to release the new software in quarter 4 of FY 23 or quarter 1 of FY 24.


Accomplishments
1. The “pyfao56” software package for Pytho. The “pyfao56” software package is a Python-based implementation of 1) the American Society of Civil Engineers (ASCE) Standardized Reference Evapotranspiration Equation, and 2) the Food and Agricultural Organization of the United Nations (FAO) Irrigation and Drainage Paper No. 56 (FAO-56) dual crop coefficient methodology. The goals of pyfao56 development were to codify the standardized ET algorithms in a modern programming language, release the software to open-source software repositories, and make the standardized ET algorithms available to a new generation of scientists with aspiration for modern software functionality and accessibility. Efforts by ARS researchers in Maricopa, Arizona, to generalize and modularize the software design have increased its applicability and relevance for scientific studies on crop evapotranspiration and irrigation management worldwide.

2. Guayule deficit irrigation strategies. Tire companies in the United States seek to commercialize guayule to provide domestic supplies for its natural rubber. However, severe water shortages in the region require guayule irrigation strategies that can maintain rubber yields, while also greatly reducing irrigation water use. ARS scientists in Maricopa, Arizona, explored several different deficit irrigation strategies to see if such approaches could attain high rubber yields comparable to a fully irrigated treatment. The research found that guayule plant size increased at the full irrigation amount, however, the rubber content in plants generally increased in deficit irrigation treatments. Notably, one deficit irrigation strategy was able to attain a rubber yield like that of the fully irrigated treatment and reduce irrigation amount by 36%. The research provides new knowledge on deficit irrigation management to sustain guayule rubber yields with significant water savings. The research will be of interest to the United States rubber industry, including tire manufacturers, irrigation consultants, water district water managers, and other research investigators of guayule.

3. Remote sensing model for crop water use. A satellite-based vegetation index model that tracks daily crop growth and evapotranspiration (ETc) was developed, tested, and validated by ARS researchers in Maricopa, Arizona, over irrigated farms in Yuma irrigation districts of Arizona and in California. Model inputs are remotely sensed normalized difference vegetation index (NDVI) images, crop type maps, and local weather. The model helps solve the problem of inaccurate crop water use estimates provided by the U.S. Bureau of Reclamation’s evapotranspiration modeling system. All Colorado River stakeholders scrutinize annual water use reports provided by Reclamation to ensure compliance with entitlements. If implemented, this model will improve accounting accuracy and reporting credibility.

4. Analysis of interference effects in furrow infiltration. Hydraulic analyses of furrow irrigation systems currently ignore potential infiltration interference effects from the merging of the infiltration bulbs of neighboring furrows. The analysis quantified the magnitude of this effect under a range of soil textures, which can reduce infiltration rates by as much as 30% under typical furrow geometric configurations and soil initial and boundary conditions. Surface irrigation analysists need to be at least aware of this effect to better interpret furrow infiltration data used for hydraulic analyses of irrigation systems. ARS researchers in Maricopa, Arizona, modified the existing infiltration model to account for the interference effect, and these modifications are expected to be added to the WinSRFR software package and made available to stakeholders at the Natural Resource Conservation Service before the project ends.


Review Publications
Elshikha, D.M., Wang, G., Waller, P.M., Hunsaker, D.J., Dierig, D., Thorp, K.R., Thompson, A.L., Katterman, M.E., Herritt, M.T., Bautista, E., Ray, D.T., Wall, G.W. 2022. Guayule growth and yield responses to deficit irrigation strategies in the U.S. desert. Agricultural Water Management. 277. Article 108093. https://doi.org/10.1016/j.agwat.2022.108093.
 Maqsood, H., Hunsaker, D.J., Waller, P.M., Thorp, K.R., French, A.N., Elshikha, D.M., Loeffler, R. 2023. WINDS model demonstration with field data from a furrow-irrigated cotton experiment. Water. 15(8). Article 1544. https://doi.org/10.3390/w15081544.
 Brekel, J.J., Thorp, K.R., DeJonge, K.C., Trout, T.J. 2023. Version 1.1.0-pyfao56: FAO-56 evapotranspiration in Python. SoftwareX. 22. Article 101336. https://doi.org/10.1016/j.softx.2023.101336.
 Thorp, K.R. 2022. pyfao56: FAO-56 evapotranspiration in Python. SoftwareX. 19. Article 101208. https://doi.org/10.1016/j.softx.2022.101208.
 Kimball, B.A., Thorp, K.R., Barnes, E.M., Choi, C.Y., Clarke, T.R., Colaizzi, P.D., Fitzgerald, G.J., Haberland, J.A., Hendrey, G., Hunsaker, D.J., Kostrzewski, M.A., Lamorte, R.L., Leavitt, S.W., Lewin, K., Mauney, J.R., Nagy, J., Pinter, P.J., Waller, P.M. 2022. Cotton response to CO2, water, nitrogen and plant density - A repository of FACE, AgIIS and FISE experiment data. Open Data Journal for Agricultural Research. 8:1-5. https://doi.org/10.18174/odjar.v8i0.18152.
 Marek, G.W., Evett, S.R., Thorp, K.R., DeJonge, K.C., Marek, T.H., Brauer, D.K. 2023. Characterizing evaporative losses from sprinkler irrigation using large weighing lysimeters. Journal of the ASABE. 66(2):353-365. https://doi.org/10.13031/ja.15300.
 Chen, X., Hou, Y., Kastner, T., Liu, L., Zhang, Y., Yin, T., Li, M., Malik, A., Li, M., Thorp, K.R., Han, S., Liu, Y., Muhammad, T., Liu, J., Li, Y. 2023. Physical and virtual nutrient flows in global telecoupled agricultural trade networks. Nature Communications. 14. Article 2391. https://doi.org/10.1038/s41467-023-38094-4.
 Kamruzzaman, M., Wahid, S., Shahid, S., Alam, E., Mainuddin, M., Islam, H., Cho, J., Rahman, M., Biswas, J., Thorp, K.R. 2023. Predicted changes in future precipitation and air temperature across Bangladesh using CMIP6 GCMs. Heliyon. 9(5). Article e16274. https://doi.org/10.1016/j.heliyon.2023.e16274.
 Ayankojo, I.T., Thorp, K.R., Thompson, A.L. 2023. Advances in the application of small unoccupied aircraft systems (sUAS) for high-throughput plant phenotyping. Remote Sensing. 15(10). Article 2623. https://doi.org/10.3390/rs15102623.
 Kimball, B.A., Thorp, K.R., Boote, K.J., Stockle, C., Suyker, A.E., Evett, S.R., Brauer, D.K., Coyle, G.G., Copeland, K.S., Marek, G.W., Colaizzi, P.D., Acutis, M., Alimagham, S., Archontoulis, S., Babacar, F., Barcza, Z., Basso, B., Bertuzzi, P., Constantin, J., De Antoni Migliorati, M., Dumont, B., Durand, J., Fodor, N., Gaiser, T., Garofalo, P., Gayler, S., Giglio, L., Grant, R., Guan, K., Hoogenboom, G., Jiang, Q., Kim, S., Kisekka, I., Lizaso, J., Masia, S., Meng, H., Mereu, V., Mukhtar, A., Perego, A., Peng, B., Priesack, E., Qi, Z., Shelia, V., Snyder, R., Soltani, A., Spano, D., Srivastava, A., Thomson, A., Timlin, D.J., Trabucco, A., Webber, H., Weber, T., Willaume, M., Williams, K., van der Laan, M., Ventrella, D., Viswanathan, M., Xu, X., Zhou, W. 2023. Simulation of evapotranspiration and yield of maize: An inter-comparison among 41 maize models. Agricultural and Forest Meteorology. 333. Article 109396. https://doi.org/10.1016/j.agrformet.2023.109396.