Evaluation of Testing Procedures for Weather-Based Irrigation Controllers

Abstract

HighlightsThe SWAT test was deficient in terms of independent reproducibility.Performance results increased for the virtual test but decreased when testing irrigation schedules from an actual landscape by switching to an hourly version of the daily soil water balance.There was lack of transferability in performance results as evidenced by the dichotomous model outcomes.Abstract. Weather-based irrigation controllers (WBICs) are modern technologies designed to schedule irrigation applications with automatic in-ground irrigation systems based on plant water requirements. The Irrigation Association developed the Smart Water Application Technologies (SWAT) testing protocol as a pseudo standard for evaluating the scheduling abilities of these WBICs. SWAT test scores provide an estimate of irrigation adequacy and scheduling efficiency as measures of under- and over-irrigation, respectively. The EPA WaterSense Program adopted the SWAT test as part of its criteria for identifying water-efficient devices for consumer labeling. The objectives of this study were to evaluate the rigor of the SWAT test in terms of (1) reproducibility of the evaluation, (2) transferability of performance results when tested in a different climate, and (3) whether an hourly soil water balance that better reflects natural soil water movement would significantly alter performance results and improve the relationship between performance scores and water conservation potential. Both hourly and daily soil water balances were evaluated using data from a virtual test, modeled after original SWAT testing conditions, and a field test conducted in southwest Florida that evaluated the same brands of WBICs to assess water conservation potential against time-based schedules. During the exploratory process, it was determined that the SWAT test was not independently reproducible due to deficiencies in protocol documentation, controller setup and programming procedures, order of operations, and weather data availability. In the virtual test, both irrigation adequacy and scheduling efficiency increased by as much as 11.8 and 9.6 percentage points, respectively, for the hourly model compared to the daily model for each WBIC. Scheduling efficiency scores from the field test declined by as much as 12.9 percentage points when switching to the hourly model, with almost perfect irrigation adequacy in every season. These results indicate over-irrigation by WBICs in practice, with few instances when scores met the minimum criteria for achieving a label by WaterSense. Over-irrigation by WBICs in the field test was partially attributed to the inability to effectively incorporate rainfall into scheduling decisions. The dichotomy in the results of these two datasets highlights deficiencies in the transferability of results. Intermittent rainfall created transitional periods, both before and after the rain event, when the controller must make smart decisions and should be included in any evaluation of a smart controller. Additionally, the connection between water conservation potential and performance results must be explored in more depth. Keywords: ET controllers, Smart Water Application Technologies, Soil water balance, SWAT, Testing standards.