Non Technical Summary
Irrigated agricultural systems in the U.S. are among the most productive in the world. However, diminishing water supplies and nutrient pollution threaten the sustainability of more than 50 million acres of irrigated farmland in the U.S. The proposed integrated research, education, and extension project develops, pilots, and broadly disseminates an integration software technology to help producers make smart irrigation water and nutrient management choices to improve yield, conserve water, and improve water quality. While in-situ, remotely sensed, and drone data products are increasingly available, their use remains siloed due to lack of integration capacity. The use of partial information often leads to incomplete assessments and potentially erroneous decisions. The proposed software assimilates interconnected soil moisture, evapotranspiration, and canopy measurements using science-guided machine learning and process-based modeling to create insight that supports robust decisions. Producers often reconcile multiple factors that influence short-term profits (like yield and costs) with those that support long-term resilience (like soil health and water quality).The proposed software enables producers to explore tradeoffs among these factors, thus fostering smart, data-driven, and integrated irrigation and nutrient management solutions. Specific decisions that are supported by the software include variable rate irrigation application, full and deficit irrigation strategies, and fertilizer application. While complexity and uncertainty are comprehensively represented in the software, its development follows co-design activities with key stakeholders in Colorado and New Mexico to ensure that its uses are impactful for the target audiences. Extensive extension and education efforts promote the software in the study areas and nationally.

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
111	0199	2050	20%
102	0199	1070	20%
402	5360	2020	60%

Knowledge Area
102 - Soil, Plant, Water, Nutrient Relationships; 111 - Conservation and Efficient Use of Water; 402 - Engineering Systems and Equipment;

Subject Of Investigation
0199 - Soil and land, general; 5360 - Drainage and irrigation facilities and systems;

Field Of Science
2020 - Engineering; 2050 - Hydrology; 1070 - Ecology;

Goals / Objectives
This project aims to develop and pilot an integrated software platform to support informed irrigation and nutrient management in cropland agriculture. The software synergizes new algorithms and methods for assimilation of smart in-situ, remote sensing, and unmanned aerial vehicle (UAV) data with model-based predictions to inform near- to long-term management decisions. Specifically, the technology provides features under the following capabilities:Application Amount Advisory: The software provides robust recommendations for irrigation and nutrient application amounts. Near-term (i.e., within-season) forecasts support variable rate irrigation (VRI) decisions and seasonal full and deficit irrigation strategies by optimizing water use while minimizing effects on yield and soil health. Long-term recommendations inform irrigation water and nutrient application strategies across multiple seasons considering effects on yield, financial risks to the producer, and the environment.Robust Irrigation and Fertilization Decisions: The software provides a thorough characterization of tradeoffs between crop production, life cycle (nutrients, water, and energy) costs, water pollution, and soil health for informed irrigation and nutrient management decisions.Data Acquisition Network Design: The software informs location-specific optimal types and placement of in-situ soil moisture sensors by minimizing data acquisition costs and maximizing their information content for variable rate, full, and deficit irrigation strategies.The long-term goal of the project is to improve water management in irrigated cropland by enabling producers to optimize crop production, costs, water conservation, and environmental stewardship. The integrated research, extension, and education objectives are to:Obj 1) [Extension] Engage with key stakeholders in co-design activities throughout the project to identify priority capabilities, features, and deployment strategies for software.Obj 2) [Research] Develop science-guided machine learning algorithms to coherently assimilate related smart sensing and model-based information for combined irrigation and nutrient management recommendations.Obj 3) [Research] Develop methods for characterization and effective communication of tradeoffs between crop production, costs, and conservation effects of irrigation water and nutrient management solutions (i.e., technologies and strategies).Obj 4) [Research and Extension] Determine and document, through carefully designed pilot studies at research and producer farms, the effectiveness of the software for improving irrigation water and nutrient application efficiency and conservation.Obj 5) [Research and Education] Create a service-oriented architecture for desktop, mobile, and web-based deployment of the software with educational materials that enable other researchers and technology developers to add new data and modeling algorithms.Obj 6) [Extension] Promote widespread adoption of the smart irrigation and nutrient management software by stakeholders in the project areas and across the U.S.Obj 7) [Education] Enable access to the software by creating and streamlining effective training and learning materials in partnership with existing irrigation education organizations, including the Master Irrigator and national extension programs.

Project Methods
The project activities hinge on an adaptive and fully integrated research, extension, and education framework to systematically enhance the design and usefulness of the proposedWISE Pro software for improved irrigation and nutrient management through continuous collections of perspectives, experiences, and feedback from stakeholders. Project activities are conducted in Colorado and New Mexico, whichare ideal locations for this project.Both areas have increasing competition for water demands and/or diminishing or highly variable surface and ground water supplies.The WISE Pro softwarebuilds on two existing tools - Water Irrigation Scheduling for Efficient Application (WISE)and Edge of Field Conservation Assessment Tool (EFCAT) - developed by the project team. These tools are deployed using the environmental Resources Assessment and Management System (eRAMS) cloud computing platform. The WISE tool was developed to create web and mobile tools for irrigation advisory using site specific climate, soils, and management information. The tool is currently used by various producers, associations, and other stakeholders in local and state agencies.The EFCAT tool was developed to enable producers to assess the water quality effects of agricultural conservation practices, including irrigation water management, nutrient management, and tillage and residue management.Hydrologic and water quality processes are simulated by EFCAT using the Soil and Water Assessment Tool (SWAT).Moreover, the project incorporates features from other existing irrigation advisory tools such as ARSPivotfor VRI and Lindsay Fieldnet software.In WISE Pro, the soil water balance underpins site-specific irrigation water demand estimation and scheduling. As the crop grows and extracts water from the soil to satisfy its ET requirement, the stored soil water is gradually depleted. The soil water deficit depth, which is the difference between field capacity and current soil water content in the root zone. Most existing irrigation advisory tools do not account for crop stress and are mostly applicable for full irrigation treatments. The project team recently developed the FAO56 Evapotranspiration tool in Python (pyFAO56) to encode algorithms for crop ET calculation in deficit irrigation systems with capacities to incorporate smart sensing data. This project will incorporate this open-source tool in WISE Pro for enhanced ET and irrigation calculations.We use a recursive Bayesian machine learning technique to assimilate data from various smart sensing systems and process-based modeling to improve estimation of water budgets, i.e., soil moisture, available water, ET, percolation, runoff, and water quality. Water balance principlesand crop ET relationships guide scientifically coherent and statistically rigorous assimilation of data from different sources with varying levels of accuracy. The technique is a sequential time-stepping procedure, in which a previous model forecast is compared with newly received observations, the model state variables are then updated to reflect the observations, and new forecasts are made.Stakeholders are at the core of the project activities to (1) co-develop and (2) broadly disseminate WISE Pro. Interwoven research and extension activities enable the co-design of the software in close collaboration with a stakeholder advisory group comprised of producers and other local and state stakeholders in Colorado and New Mexico. The advisory group includes crop producers, local and state water managers, crop consultants, irrigation manufacturers technical staff, and national leaders in developing smart agriculture technology and software. Stakeholder inputnot only guide product development and testing, but also inform the direction for all relevant outreach and extension activities.In addition to the stakeholder advisory group, feedback on project activities issought in years two and three from the wider stakeholder network known to project leadership. Interactions include individual conversations, presentations at crop clinics, regional conferences, and other opportunities as available. Alignment of the project with the extension mission of the partnering land-grant universities (CSU and NMSU) enhances the ability of the project team in reaching interested producers and stakeholders. The extension and educational activities focus on promoting WISE Pro for improved irrigation and nutrient management decisions in the project areas and nationally.Smart sensing and analysis systems for agriculture is a rapidly evolving field. We design and deploy the WISE Pro software using the eRAMS platform developed by the CSU project team with an architecture that enables other researchers and developers to integrate their analysis tools in the proposed software platform. The eRAMS technology is a cloud computing platform with a computationally scalable database, content management, GIS, computing, and web building blocks. New modules can be seamlessly added to the software ecosystem and tested with extensive project dataprovided in the platform. Specifically, the software includes Application Programming Interfaces (APIs) as well as desktop and mobile decision support interfacesto meet the needs and specifications of producers and other stakeholders. The service-oriented architecture of the software enables the creation of interfaces that are tailored for specific applications, code revisions and version control, software maintenance, and application lifecycle management.The WISE Pro software will be developed using extensive data from theLimited Irrigation Research Farm (LIRF).USDA-ARS Water Management and Systems Research Unit has conducted full and deficit irrigation experiments at the 50-acre farm and research facility since 2008. This unique infrastructure for field research at LIRF includes linear-move variable-rate irrigation (VRI) sprinkler, subsurface drip, surface drip, and furrow irrigation systems.The software will be piloted in two experimental farms managed by the CSU and NMSU team: CSU's Agricultural Research, Development, and Education (ARDEC) in the South Platte River Basin in Colorado and NMSU Science Farm in the Ogallala Aquifer system in New Mexico.The final deployment of the WISE Pro software is informed by its application and demonstration in a producer farm inColorado. The project provides the participating producer with smart sensing devices for in-situ soil moisture and UAV canopy monitoring. The participating producer isa part of the co-design process and is therefore well positioned to test the software in their operations. Primary field identification and technology deployment is executed with the full guidance and support of the producer.The project createsuser's guide and theoretical manuals for application of the software in different irrigated agricultural fields. The materials include several examples with associated datasets from the LIRF, CSU TAPS, and NMSU Science Farm. The materials are augmented with educational videos to comprehensively describe the various features of the software. Specifically, educational modules are developedfor irrigation water and nutrient management scenarios. The materials aremaintained at the project website for the WISE Pro software. The project updates the CSU Irrigation Handbook with the findings of the research activities at LIRF, CSU and NMSU research fields, and the participating producer's farm. A pamphlet for producers is created to support learning about WISE Pro capabilities and features, with QR code to download its online and mobile apps and/or navigate to the software website. The educational materials areused to develop online course modules for the Master Irrigator program curriculum.