Analysis and simulation of the water soil atmosphere continuum in relation to the growth and yield of annual crops

The aim of the present work is the analysis and simulation of the water-soil-atmosphere system, in relation to growth and yield of annual crops. Detailed datasets were constructed for gun-sprinkler irrigated experimental fields on loamy soil, cropped with maize and cotton during the seasons 1998 and 1999 in central Greece. Soil water content was measured using TDR instrumentation and crop’s LAI by the SunScan System. Irrigation water depths and yields were also measured. The effective crop rooting depth, its growth over time and the maximum water uptake rate by roots from each soil layer were determined after soil water content measurements. The WAVE 2.1 model fed with the experimental and meteorological data was used to simulate the water-soil-atmosphere continuum. The maximum water uptake distribution with depth approached by three functions (linear, exponential and sigmoidal) and the van Genuchten and Gardner functions of hydraulic conductivity were used to run the model. The success of the model was tested by comparing its output with the experimental data such as soil water content, evapotranspiration and yield. The results were evaluated using simulation graphics and specific modelling evaluation statistics. Results showed that the model is able to predict the fate of water in irrigated soils subjected to the semi-arid Mediterranean environment and agronomical conditions of central Greece. Reliable predictions of the seasonal evolution of the crops evapotranspiration were obtained by the model. Soil water content was well predicted when considering total moisture storage in the root zone. However, the deviations between the predicted and observed moisture values increased when considering each soil layer separately, especially in the top one. The maximum water uptake describing functions generated simulations of equal accuracy, while van Genuchten’s function describes the soil water content more precisely than the Gardner’s one. As far as the predicted maize dry matter and seed yield, it was found that they are underestimated by the model about 10% and 20% respectively. Sensitivity analysis revealed that the parameters of the maximum water uptake function have a significant influence on model predictions when the change of parameters is towards the reduction of its value, while in the opposite case the effect is very small. It has been also shown that the characteristic curve parameters, mainly its slope, influence the model results