Accuracy of Reference Evapotranspiration Estimation by Two Irrigation Controllers in a Humid Climate

Increasing municipal water demand has led to new innovations in water-saving technology. Evapotranspiration-based irrigation controllers, also known as weather-based irrigation controllers, are a technology that use evapotranspiration (ET) estimates to schedule irrigation. Although real-time on-site data collection could help improve data quality, limited data parameters collected or simplified reference evapotranspiration (${\mathrm{ET}}_{\mathrm{o}}$) estimation methods from an on-site sensor could potentially have an impact on estimated ${\mathrm{ET}}_{\mathrm{o}}$ values for stand-alone controllers. Conversely, collection of data from off-site weather stations that are not representative of the local conditions could lead to inaccurate ${\mathrm{ET}}_{\mathrm{o}}$ estimations for local irrigation sites. The Weathermatic SL1600 and Toro Intelli-sense controllers were selected for testing because of the different methods of data collection and ${\mathrm{ET}}_{\mathrm{o}}$ estimation. This study was conducted in two locations: Wimauma and Gainesville, Florida. In Wimauma the Weathermatic controller significantly overestimated ${\mathrm{ET}}_{\mathrm{o}}$ compared to on-site ${\mathrm{ET}}_{\mathrm{o}}$ estimates using the ASCE-EWRI standardized method (${\mathrm{ET}}_{\mathrm{o}}$) by 9–15% for the duration of the study, while ${\mathrm{ET}}_{\mathrm{o}}$ sent to the Toro controllers was similar to on-site ${\mathrm{ET}}_{\mathrm{o}}$ estimations (1–3%) for the duration of the study. However, the Weathermatic controller overestimated ${\mathrm{ET}}_{\mathrm{o}}$ for every season (fall: 5–8%; winter: 6–8%; spring: 11–16%; summer: 22–25%), and the Toro controller overestimated ${\mathrm{ET}}_{\mathrm{o}}$ in the summer seasons (12%) and underestimated ${\mathrm{ET}}_o$ in the winter seasons (17%). In Gainesville, the Weathermatic and Toro controllers both overestimated ${\mathrm{ET}}_{\mathrm{o}}$ compared to on-site ${\mathrm{ET}}_{\mathrm{os}}$ (30 and 10%, respectively). However, when compared to ${\mathrm{ET}}_{\mathrm{os}}$ using data from the Gainesville National Oceanic and Atmospheric Administration (NOAA) weather station, the ${\mathrm{ET}}_{\mathrm{o}}$ of the Toro controller was consistent with the equation and location of the weather data that were likely used by the weather service to calculate ${\mathrm{ET}}_{\mathrm{o}}$, indicating that this was likely the data source for the ${\mathrm{ET}}_{\mathrm{o}}$ sent to the Toro controller. Overall, ${\mathrm{ET}}_{\mathrm{o}}$ sent to and used by the Toro Intelli-sense controllers was closer to on-site ${\mathrm{ET}}_{\mathrm{o}}$ estimations than that which was estimated by the Weathermatic SL1600 controllers on site.