Simulação da extração de água do solo pela cultura do milho (Zea mays L.) utilizando modelo HYDRUS-1D

Abstract

The understanding of the process of water extraction by the plant assists in irrigation management practices and contributes to a better management of applied water resources. Such understanding requires practical field studies, which require time and high financial resources. Mathematical models can simulate these processes satisfactorily and can be used in the most different environmental conditions. The objective of this study was to use the HYDRUS-1D model to simulate the soil water content, pressure head and water extraction by the roots in a soil cultivated with maize under different irrigation regimes. The experimental area used belongs to the Federal Institute of Ceará Iguatu campus. AG-1051 hybrid corn was cultivated and climate irrigation management was performed, and the data obtained from the automatic meteorological station of the National Meteorological Institute, located near the experimental area. The treatments were divided into five irrigation slides, which differed in 50, 75, 100, 125 and 150% of the crop evapotranspiration. The corn was harvested in its green form and analyzed data regarding weight, length and ear diameter, and the treatments were compared through the Tukey test at 5% probability. For statistical comparison between simulated and measured tensiometer batteries were installed at depths of 10 and 30 cm in each treatment to obtain matric potential and soil moisture. The simulation was performed using the HYDRUS-1D model, which uses the Feddes macroscopic model in the simulation of water extraction by the roots. The results of the simulation were compared to the measured data using the statistical parameters Willmott index, coefficient of efficiency, square root of the mean error, mean error and absolute maximum error. The highest production data were from the treatment with 150% of crop evapotranspiration. The simulated data that presented the best similarity with the data measured were those obtained for the 30 cm layer in the treatments with greater than 100% of the crop evapotranspiration. The Feddes macroscopic model did not adequately simulate the analyzed parameters, especially in the 10 cm layer, being this one of greater importance due to the higher density of roots. These results point out that it would be inappropriate to use simulated data with this model for agricultural management.