Hydroponic systems can drain nutrient-rich waste into the environment. Increasing irrigation efficiency would decrease effluent and improve cost efficiency for growers. However, current methods accessible to small- and mid-sized growers to determine moisture content in growth media are often imprecise. Simplified transpiration models could inform irrigation needs. This study aimed to improve transpiration estimates using vapor pressure deficit (VPD) and solar radiation. We compared our model to an existing transpiration model. Three years of transpiration and environmental data from tomato production were used to calibrate (year 2) and validate (years 1 and 3) the model. Randomly chosen subsets from all years of data were also used. The new model (TVPD) predicted the observed values more closely than the previous model (PG) in year 1 (TVPD: RMSE = 0.1570 mm, r2 = 0.95; PG: RMSE = 0.5594 to 0.6875 mm, r2 = 0.27 to 0.78) but not in year 3 (TVPD: RMSE = 0.5430 mm, r2 = 0.44; PG: RMSE = 0.1873 to 0.2065 mm, r2 = 0.95). TVPD calibrated using random subsets of the combined data improved consistency and predictive capacity (RMSE = 0.2387 to 0.2419 mm, r2 = 0.87 to 0.91). TVPD is a simpler alternative to complex models and to those focusing on solar radiation alone. TVPD is less reliable under low solar radiation (year 3); however, reliability could be improved by calibration across a broader environmental range. TVPD also allows for exploration of the relative influences of low VPD and high solar radiation on evapotranspiration found in greenhouse settings.
Dunn et al. (Fri,) studied this question.