The numerical analysis is focused on a specific location in the Texas High Plains, Moore County, which accounts for a substantial proportion of the total irrigated corn area in the region (U.S. Department of Agriculture, 2012). Figure 2.7 illustrates the location of Moore County within the state of Texas.
Figure 2.7: Map showing locations of Moore County, the weather station at Amarillo (Potter County), and the agricultural research station at Bushland (Randall County) in Texas.
data recorded at Amarillo in neighbouring Potter County (Figure 2.7), obtained from the Na- tional Oceanic and Atmospheric Association Global Summary of the Day data set (U.S. National Climatic Data Center, 2014), therefore are used as the basis for generating the stochastic intrasea- sonal crop-water production functions. The weather station at Amarillo provides daily values of maximum and minimum temperature, dew point temperature, total precipitation, and average wind speed over the period 1943-2013. Global Summary of the Day data undergoes a range of quality control measures (Durre et al., 2010). However despite these, 16 of the 71 record years (1943-1946, 1965-1973, 1981, and 1992-1993) are not used in the generation of production functions as in these years greater than 10% of days contain missing data for one or more vari- able. For the 55 years that are retained, the final weather input variable required by AquaCrop, reference evapotranspiration, is computed using the standardised American Society of Civil En- gineers (ASCE) Penman-Monteith equation (Allen et al., 2005). At a daily time step the ASCE Penman-Monteith formula is identical to the FAO-56 Penman-Monteith equation, and is the rec-
ommended method for estimating reference evapotranspiration (Allen et al., 1998, 2005). The ASCE Penman-Monteith equation has also been shown to perform better than alternative equa- tions (e.g., Hargreaves, Penman, Kimberley-Penman etc.) in the High Plains region of the United States where high winds and vapour pressure deficits have large impacts on reference evapotran- spiration rates (Itenfisu et al., 2003). Furthermore, the ASCE Penman-Monteith equation was used as the basis for estimating reference evapotranspiration in the calibration of AquaCrop for corn at Bushland in Texas (Figure 2.7) thus providing consistency with the methods used to develop the crop parameter set for use in AquaCrop in this chapter (Heng et al., 2009).
In addition to the specification of weather inputs, soil, crop, and management parameters have to be defined in AquaCrop. Soil type is defined to represent a Sherm silty clay loam soil. This soil is the most commonly cropped soil type for corn production in Moore County, as identified by a comparison of the spatial distribution of soils in the county given in the SSURGO data set (U.S. Department of Agriculture Natural Resources Conservation Service, 2014) with the historic (2009-2013) distribution of corn production areas in the county (U.S. Department of Agriculture National Agricultural Statistics Service, 2014) (Figure 2.8). Based on this selected soil type, soil textural properties of 23 % sand, 46 % clay, 31 % silt, and an organic matter content of 0.66 % are extracted from the SSURGO dataset. These values subsequently are used to calculate soil hydraulic properties, which are required as inputs to AquaCrop, using the pedotransfer function model of Saxton and Rawls (2006) (Table 2.1).
Table 2.1: Soil hydraulic properties for AquaCrop in the Texas High Plains.
Parameter Value
Water content at permanent wilting point (m3 m-3) 0.274
Water content at field capacity (m3 m-3) 0.406
Water content at saturation (m3 m-3) 0.483
Saturated hydraulic conductivity (mm day-1) 27
In AquaCrop, the complete set of parameters used to simulate crop growth and yield devel- opment can be divided into conservative and user-specific parameters (Raes et al., 2012; Steduto et al., 2012). Conservative parameters are those that are believed to be constant spatially and temporally, and are expected to be unaffected by management practices or climate. Conservative crop parameters include, but are not limited to, the soil water thresholds at which different crop growth processes are inhibited by water stress, the temperature thresholds used to determine
Figure 2.8: Map showing the overlap between historic (2009-2013) corn production areas (yellow) (U.S. Department of Agriculture National Agricultural Statistics Service, 2014) and the extent of Sherm silty clay loam soil type (brown) (U.S. Department of Agriculture Natural Resources Conservation Service, 2014) in Moore County, Texas.
crop growing degree days, and the crop water productivity parameter that defines the amount of aboveground biomass that is accumulated per unit of transpiration. Contrastingly, user-specific parameters may vary depending on the specific crop cultivar that is planted or due to local man- agement practices and climatic conditions. Examples of user-specific parameters are the crop phenological calendar defining the number of growing degree days in each growth stage, and the maximum effective rooting depth of the crop. In this analysis, the crop parameter set is defined according to a previous validation of AquaCrop in the Texas High Plains at Bushland, Pot- ter County (Heng et al., 2009), which demonstrated that AquaCrop was able to reproduce corn growth and yield adequately under full and deficit irrigation conditions. Heng et al. (2009) define the values of conservative crop parameters according to the default values for corn reported in Hsiao et al. (2009) and Raes et al. (2012). User-specific parameters are estimated by Heng et al. (2009) based on observations of climate and crop growth during field experiments conducted at the United States Department of Agriculture Agricultural Research Service Conservation and Production Research Laboratory (USDA-ARS CPRL) in Bushland, Texas, and these parameter
values are summarised in Table 2.2. In addition, it is assumed in the calculation of the crop-water production functions that the corn crop is planted on May 1 each year with a planting density
of 74,132 plants ha-1 (30,000 plants ac-1) that is characteristic of typical agronomic practices in
the Texas High Plains region (U.S. Department of Agriculture National Agricultural Statistics Service, 2010), and that soil moisture levels at the start of the growing season are equal uniformly to 80% of soil water holding capacity throughout the soil profile.
Table 2.2: User-specific crop parameters for AquaCrop in the Texas High Plains.
Parameter Value
Time from planting to 90% canopy emergence (GDD*) 100
Time from planting to start of flowering (GDD) 900
Duration of flowering (GDD) 190
Time from planting to start of canopy senescence (GDD) 1600
Time from planting to physiological maturity (GDD) 1850
Maximum effective rooting depth (m) 1.6
*GDD: growing degree day
Using the parameters and weather inputs described above, the Matlab-AquaCrop model is applied to simulate crop yield and total irrigation requirements for a range of soil moisture targets using the procedure described in Section 2.2.2. The soil moisture target is varied from 0, representing permanent wilting point, to 1, indicative of field capacity, in increments of 0.05 to capture the complete range of potential soil moisture management strategies that are available
to a farmer. When irrigation is triggered water is considered to be applied uniformly over
the full irrigated area during a single day in order to bring the soil water content back to field capacity. The maximum daily irrigation rate is, however, limited by the instantaneous application constraint imposed by well yield and irrigated area (Equation 2.4). Stochastic intraseasonal crop- water production functions are generated for maximum daily irrigation rates that vary from 0.1
mm day-1 to 20 mm day-1 in increments of 0.1 mm day-1 to capture a range of conditions from
effectively no irrigation potential to an irrigation application level necessary to satisfy full crop water requirements on every day of all growing seasons. Additionally, it is assumed that 10% of the irrigation water applied does not reach the crop root zone, which is consistent with typical application losses from a centre-pivot system operating in the region (Wagner, 2012).