Skip to main content
ARS Home » Research » Publications at this Location » Publication #314602
Title: Simulating maize yield and biomass with spatial variability of soil field capacity

Author
item Ma, Liwang
item Ahuja, Lajpat
item Trout, Thomas
item NOLAN, B - Us Geological Survey (USGS)
item Malone, Robert - Rob

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/27/2015
Publication Date: 10/27/2015
Citation: Ma, L., Ahuja, L.R., Trout, T.J., Nolan, B.T., Malone, R.W. 2015. Simulating maize yield and biomass with spatial variability of soil field capacity. Agronomy Journal. 108:171-184.

Interpretive Summary: Spatial variability in field soil water and other properties is a challenge for system modelers who use only representative values for model inputs, rather than their distributions. In this study, we compared simulation results from a calibrated model with spatial variability of soil field capacity (FC) expressed in different ways, against four years of maize experimental data for six irrigation levels in northern Colorado, USA. The spatial variability of FC in the field was estimated based on numerous measurements of soil water content about two days after irrigation in all plots over four years. Formulations of the spatial variability included the use of measured spatial distribution applied to all plots and years, random selection of FC for a plot from a log-normal distribution within a given range, use of field estimated FC for each plot, and the use of overall average measured value of FC for all plots and years. The experimental data were used to calibrate the six CERES-Maize crop parameters, along with FC in each soil layer, in the Root Zone Water Quality Model (RZWQM2). The calibrated model was then used to simulate the effects of spatial variability in FC on corn production. For this study, an average FC from all the plots was adequate in model simulation. However, the variability in FC did not fully explain the variability observed in yield and biomass in the field.

Technical Abstract: Spatial variability in field soil water and other properties is a challenge for system modelers who use only representative values for model inputs, rather than their distributions. In this study, we compared simulation results from a calibrated model with spatial variability of soil field capacity (FC), against four years of maize experimental data for six irrigation levels in northern Colorado, USA. The experimental data were used to calibrate the six CERES-Maize crop parameters, along with FC in each soil layer, in the Root Zone Water Quality Model (RZWQM2), with overall root mean squared errors (RMSE) of 354 kg/ha and 1202 kg/ha for yield and biomass. The spatial variability of FC in the field was estimated based on soil water content about one day after irrigation in the 96 plots. Keep the same plant parameters as calibrated, best results were obtained when the model was run with a distributed FC input for all the 96 sets of measured FC, rather than with a distributed FC input (100 sampled FC sets) from a long-normal distribution within the range given in the measured FC. We further showed that the use of average set of FC from measured FC values of soil layer produced as good results as the distributed FC input with the 96 sets of measured FC. When we run the model for each replicate (4 replicates each treatment) with its respectively measured FC and then averaged the results for each treatment, the obtained RMSEs were the worse (564 kg/ha for yield and 1867 kg/ha for biomass), which could be due to larger variation in FC within plots than between plots. However, the variability in FC did not fully explain the variability observed in yield and biomass in the field. Nonetheless, the results showed that field variability was best taken into account through distributed model inputs, and a model could be adequately calibrated using an average set of FC from a large numbers of FC measurements in a field.