Determining water requirements and scheduling irrigation of apple trees using soil-based, plant-based and weather-based methods

The goal of this work was to estimate water requirements, and to develop precision methods for automating the irrigation of apples. Two models based on the energy balance of a single leaf and infrared thermometry (IRT) were developed to calculate potential (Tp) and actual (T) transpiration from the whole tree. Tp and T were compared with ETr and ETc, respectively. The models were evaluated using the canopy temperature (Tc) and air temperature ( Ta) data collected in a well-watered orchard and weather data from a nearby weather station during the 2007, 2008 and 2013 growing seasons. In addition, the microclimate of the orchard was investigated using a suite of sensors. Moreover, a wireless data collection network and scheduling algorithms were developed to create a site-specific irrigation control system. The precision methods were compared based on the total irrigation water requirements and water use in 2013. The Tp model was able to reflect the high degree of coupling between the apple trees and the humidity of the surrounding air during cold and humid periods. Both T and Tp were better correlated with ETc on warm and dry days than during cold and humid periods. Similar results in all of the three growing seasons indicated that DeltaT (Tc-- Ta) could be linearly related to T. The results showed that the transpiration of the trees was intense late in the morning and afternoon. A high correlation and small difference between daily mean canopy and trunk surface temperatures suggested the potential to use trunk temperature as an alternative for traditional IR measurements. Because of the high discrepancies between the Ta measurements in the orchard and the weather station, it was concluded that Ta should be measured in the vicinity of the IRTs. Within-canopy wind velocities were about 0.1 times the surface wind speeds. In general, the daily means of the measurements in the orchard and weather station were highly correlated while they were not well related at solar noon. The total irrigation water applied by the conventional irrigation (CNTRL) was significantly higher than all of other methods.