Applying apps in irrigation scheduling

Advances in irrigation technologies have made it possible to achieve high application and water use efficiencies. Yet, realizing the achievable benefits of these advances depends on effective management of the technologies for the crops and production conditions. Research-based recommendations and crop models can be very helpful in irrigation decisions of if, when and how much to irrigate, which can be especially valuable in managing limited irrigation resources. Proliferation of sensors and associated telemetry, network and gateway communications options and a wide range of remote data resources provide ready access to weather and climate information, soil and plant conditions. Technological advances, evolving research-based knowledge and increasing volumes of data can be very useful, or even overwhelming. One challenge in irrigation management is to integrate data and relevant background knowledge into timely recommendations to support “optimal” irrigation management decisions. There’s an app for that.

Actually, there are several apps for that. Irrigation scheduling software, or apps, can simplify integration of near real-time data (weather, soil water and crop status), research-based recommendations and crop water use models to inform irrigation management decisions.

Irrigation scheduling apps by public-sector and private-sector developers include a wide range of sophistication and ease of use. From basic “checkbook” soil water balance methods to complex tools integrating data from multiple sources with advanced mechanistic models, irrigation scheduling apps aim to inform and simplify irrigation management decisions.

While there are many advantages to these apps, there also are some common issues and concerns. It is helpful to understand how the apps work, quality of data required and the background assumptions incorporated in the models.

Weather data sources and quality

Evapotranspiration-based irrigation scheduling tools require weather data to calculate reference evapotranspiration (ET_ref) and crop coefficients (K_c) to convert ET_ref to crop ET (ET_c). Weather data can be obtained from local or regional weather station networks or on-site weather stations. While many weather station networks are designed for agricultural use and follow recommended standards for sensors, quality assurance/quality control and siting, others are not necessarily well sited or instrumented for ET_c purposes. Evapotranspiration networks, such as such as California Irrigation Management Information System (CIMIS, at cimis.water.ca.gov), Florida Automated Weather Network (FAWN, at fawn.ifas.ufl.edu/tools/et/), and Oklahoma Mesonet (at cimis.water.ca.gov), provide near real-time weather data needed to calculate ET_ref, as well as other agriculturally relevant data. Many sources of weather data may include stations sited for other primary purposes, so they may not be good representations of agricultural field environments. It is also notable that many agricultural areas are not covered by ET weather networks.

Weather data quality depends on instrumentation quality, calibration and maintenance, as well as siting. This applies to public or private networks, as well as on-site weather stations. Some weather parameters (such as solar radiation) are relatively stable with distance, but others (including rainfall) may vary widely over even relatively short distances. In many instances, other factors (topography, elevation and other local conditions) may indicate that the nearest weather station in a network may not necessarily be the most representative weather station for a given application. Fortunately, many ET networks provide additional information on their websites about their stations, locations and QA/QC, and many irrigation scheduling apps allow the user to select an applicable weather station or import their own data.

Reference ET crops and crop coefficients

Beyond considerations of weather data availability and quality, application of ET data to irrigation involves calculations of reference ET and crop ET. These steps offer opportunities for errors in application. For instance, there are different reference crops (tall crop or alfalfa compared to short crop or cool season grass model crops), with different ET_ref equations and different ET_ref values. This means that the crop coefficients to convert ET_ref to ET_c are specific to the reference ET equation used. Crop coefficients are crop-specific and growth stage specific, and available crop coefficient curves are limited with regard to other crop factors. Hence, ET_c is an estimate of crop water demand, generally assuming ideal crop conditions. A common mistake is that model users and developers may use ET_ref in lieu of ET_c, which can result in significant overestimate or underestimate of crop water demand, depending on the crop growth stage and reference ET equation used.

Ease of use and data input

Irrigation scheduling apps range from simple and generic to complex and configurable for specific operations. There are tradeoffs, of course. Generic apps may be less precise in their modeling of the crop, but they can be relatively easy to use and widely applicable for a variety of crops and conditions. More configurable apps may provide more specific recommendations, and they require additional inputs. User friendliness depends on data entry (and availability of required data), as well as interpretation and presentation of outputs. Some tools guide data entry by providing default input values or embedding limits to prevent the user from entering unreasonable values (which could result in unreasonable output). Because input data may not be readily available or they may be presented in confusing formats or units, a common mistake is that users may depend too much on default values, limiting the accuracy and benefits of the apps.

Ground truthing and verification

Irrigation scheduling apps often use models that are based on limited available data and limited research. It is important to remember that while apps can be useful in integrating information, their outputs need to be verified (ground-truthed) against the crop in the field. If the app simulates the crop growth (based on days or heat units, for instance), does the simulated crop match the crop in the field? If the app provides soil water estimates, do the simulated values match the field-observed soil water? Do the model’s crop water use estimates or soil water estimates make sense, based on soil physical characteristics, soil depth and crop characteristics?  If not, can the discrepancy be attributed to field conditions or input data errors, or is the app not applicable to your field or crop conditions?

Technical assistance or user support

Ideally, irrigation scheduling apps aim to be intuitive, convenient and useful in irrigation management decisions. Sometimes users can benefit from technical support, training and other support references. Conversely, users can be helpful in identifying problems or providing feedback useful for app improvement. Technical support contacts familiar with the app design, functionality, background models and assumptions can be valuable. Active support and updates are key to longevity and applicability of irrigation scheduling apps.

Dana Porter, PhD, PE, is a professor, extension agricultural engineering specialist – agricultural water management and extension program leader in the department of biological and agricultural engineering, Texas A&M AgriLife Research and Extension Service at Lubbock, Texas. Her research and extension program emphasizes agricultural irrigation technologies and management for water-limited conditions.

Thomas Marek, MS, PE, senior research engineer, Texas A&M AgriLife Research, Amarillo/Bushland, Texas. He has extensive applied research experience in evapotranspiration, crop water use and crop water management.