Padrão de umedecimento e movimento da água no solo sob uma fonte pontual de irrigação no gotejamento subsuperficial

Abstract

The use of subsurface drip irrigation (SDI) has been increasing as a mean to deliver water, fertilizers and pesticides to plants, being more effective. Increasing the water use efficiency is one of the goals of irrigation management, since the water demand by human use and industry is expected to grow significantly in the near future. To achieve all the subsurface drip potential, some operational parameters optimization is required such as frequency and irrigation time, dripper flow, installation depth and spacing, as well as knowledge of the water distribution pattern. The distribution pattern and water movement can be accessed by direct measurement or by modeling. In this way, the objective of this study was to measure and simulate the water movement in a sandy loam soil, with two flows, in drip emitters installed at different depths, of a subsurface drip irrigation system. Two experiments were carried out at the experimental laboratory, located on Departamento de Engenharia Rural of Universidade Federal de Santa Maria, during 2017 and 2018 years. It was used polyethylene containers with a 380 mm diameter by a 630 mm height, filled with soil of a sandy loam texture (Rhodic Paleudalf). The sources of variation consisted of : the drip emitters installation depth (12, 24 and 36 cm deep), the irrigation management (8 hours of continuous irrigation and 12 hours intermittent irrigation), the emitters flow (0.9 and 1.8 liters hour-1). The Irrigation was performed through 16 mm self-compensating drippers, with 20 cm spacing between emitters. A set of FDR sensors, model CS616, was used to measure soil water content. The sensors were installed in the depths of 8, 18, 28, 38, 48 and 58 cm, inside the containers with soil. The Hydrus-2D numerical model was used to analyze the observed water content data and to simulate the wetting front under each emitter. The observed soil water content data were compared with those simulated using the root mean square error (RMSE), linear regression coefficient forced to the origin (bo), determination coefficient (R2) and modeling efficiency (EF) as statistical indices. The RMSE, for the emitter different depths, ranged from 0.01 to 0.06 cm3 cm-3, indicating good to very good agreement between the data observed and simulated by the model. Irrigation time influenced the infiltration and formation of the wetting front more than the emitter flow. The results simulated by the Hydrus-2D model demonstrated a linear relationship of more than 65% with the observed data, making possible its use to model water movement in subsurface drip irrigation. The observed differences between the observed and simulated data, although not significant, occurred during the first two hours of infiltration and are probably due to the effect of hysteresis or imperfect measures in the soil hydraulic conductivity.