Soil water status is a state variable that is often proposed as a key input to irrigation management decision support systems (DSS). Both soil matric potential and volumetric water content (VWC) have been used (Evett et al. 2008). Decisions about irrigation initiation and quantity are typically based on comparison of measured or sensed soil water status to some threshold value (Evett 2007). When VWC is used, then the threshold is often the water content at field capacity or a value called the management allowed depletion (MAD) that is some fraction of the difference between the soil water contents at field capacity (FC) and at permanent wilting point (WP). The FC is understood to be the soil water content after a thoroughly wetted soil has drained for some period of time, typically 24 h. In the laboratory, it is often taken to be the water content of a soil core after a pressure of 0.33 kPa has been applied to a saturated soil core. Irrigating to replenish the soil water to field capacity after a crop has taken up water from the soil is considered a best practice, because irrigating more than that would cause the soil to quickly lose water through internal drainage. Irrigating before the VWC declines to the MAD value is considered a best practice to avoid yield loss due to plant water stress. These concepts are appropriate for soils that are relatively uniform with depth and freely draining. However, the water content at field capacity (FC, m3 m−3) is rarely determined in the field, but rather is determined using a pressure plate apparatus operated to place a core sample under 0.33 kPa pressure. Moreover, many agricultural soils are not uniform with depth and may include horizons that restrict internal drainage.
At the USDA ARS Coastal Plain Soil, Water and Plant Conservation Research Center, Florence, SC, research on variable rate center pivot irrigation was implemented in the mid-1990s (Camp and Sadler 1994). Irrigation management was accomplished using a network of up to 45 tensiometers placed in a 6-ha field under a three-span center pivot (Sadler et al. 1996, 2002). The variable rate system was purpose built by research center personnel (Omary et al. 1997). In 2015, a new Cooperative Research and Development Agreement (CRADA) was reached between ARS and Valmont Industries, Inc. to outfit the center pivot with the Valley variable rate irrigation (VRI) system and to conduct beta tests of an Irrigation Scheduling Supervisory Control and Data Acquisition (ISSCADA) system developed and patented by ARS (Evett et al. 2014). ARS at Florence cooperated in the testing with ARS scientists and engineers at Bushland, TX; Portageville, MO; and Stoneville, MS. The ISSCADA system utilizes wireless infrared thermometers to sense crop canopy temperature, soil water sensors to determine soil water status, and weather sensors to determine solar irradiance, wind speed, humidity, and air temperature.
The ISSCADA software, named ARSPivot (Andrade et al. 2017), runs on an embedded computer at the pivot point, automatically collecting data from the wireless-sensing sources, applying algorithms to determine when, where, and how much to irrigate, and developing a VRI prescription map that is sent to the pivot control panel where it is executed. The algorithms used involve computing an integrated crop water stress index (iCWSI) and comparing it to built-in threshold values. A hybrid algorithm uses the outcome in conjunction with soil water depletion (SWD) data to develop the prescription map. The percent of SWD is calculated using the soil VWC, FC, and WP at each sensor depth. If SWD < 10%, then no irrigation is applied; if SWD > 50%, then irrigation is prescribed at the maximum depth; and if 10% < SWD < 50%, the iCWSI thresholds are used to determine the irrigation depth. When requested by the user, the software displays water content for every site and for every depth at that site and plots daily, 3-day, week-long, or season-long VWC along with FC and WP values for a specific user-chosen depth (Fig. 1). Note that, in Fig. 1, the water contents for the 15.7-inch (40-cm) depth were always larger than the FC value plotted for that depth. The sensors that were used (model TDR-315L, Acclima, Inc., Meridian, ID) were found to be accurate in the soils at Bushland (e.g., Schwartz et al. 2016). Therefore, the fact that VWC was larger than FC was attributed to water perched above and in the slowly permeable B2t horizon and the inhibition of soil water flux by the abrupt change with depth of pore size between the small pores in the B2t and the larger pores in the calcic Btka horizon underneath.
Research to beta test the ISSCADA center pivot variable rate irrigation (VRI) decision support system (O’Shaughnessy et al. 2015, 2016, 2018) at Florence was conducted in 2017 and 2018. Sensed water contents at given depths were often, even commonly, larger than field capacity values (Fig. 2). The FC values were previously determined by applying tension table and pressure plate methods to soil samples from those depths (Peele et al. 1970). Since the soil surface horizons were predominantly sandy (e.g., 15-cm depth in Fig. 2) and the FC values determined by Peele et al. (1970) seemed reasonable for those textures, this discrepancy called into question the accuracy of the sensors.
A field investigation was conducted at Florence to determine if sensor inaccuracy was causing sensed soil water content to be larger than FC and WP values, to investigate the vadose zone hydrologic factors that might cause sensor readings to deviate from expectations, and to determine if other ways of analyzing and viewing the soil water data might prove more useful for irrigation management.