Modelagem espacial no sistema de irrigação por gotejamento

Abstract

Localized drip irrigation has been gaining space in agriculture, especially in high economic value crops, because of its adaptation to different types of soils and topographies and the reduction of water and fertilizer costs. Therefore, knowing the behavior of the attributes to be applied to the soil is essential for planning and administering water resources, decision-making in practices of use, management, and reducing environmental impacts. One way to monitor the behavior is using geostatistics, which allows refining the analysis of these attributes applied to the soil because it adopts statistical methods considered for situations whose data are related to spatially distributed natural phenomena, allowing knowing their variability and continuity within an area under study. To this end, the research aims to evaluate and model the spatial variability of the topographical meshes of water flow and Krista and peat fertilizers in the drip irrigation system. The experiment was conducted in the Irrigation and Fertigation Laboratory of UNIOESTE university, at the Cascavel campus, on a workbench, creating a regular grid area of 15m², with 80 drippers spaced at 0.50 meters between them. Data on the flow rate of the artesian well water and the fertilizers peat, in liquid form, and Krista, in solid form diluted in water, were collected at each topographic slope of the irrigation system, which were processed in R software. The estimation methods adopted were ordinary least square (OLS) and maximum likelihood (ML). To select the adjusted models, the Akaike and Schwarz criteria of cross-validation were used since, from the parameters of these models, the spatial dependence indices SDI and the measure of spatial dependence SDM were determined for each specific semivariogram model, classifying the spatial variability in terms of weak, moderate and strong dependence. The results obtained from ordinary kriging made it possible to characterize the spatial variability through the generated maps. Thus, concerning the OLS estimation method, the models for water and the fertilizers Krista and peat were obtained in the leveled topography (0% slope), with moderate dependence for the SDI index and weak dependence for the SDM. In the ML method, weak dependence for SDI and SDM was obtained. The water, Krista fertilizer, and peat models were obtained with weak dependence on SDI and SDM indices on the rising slope topography (2% slope) for the OLS and ML estimation methods. On the falling slope topography (2% slope), OLS and ML methods were obtained with SDI and SDM indices; regarding the water flow and Krista and peat fertilizer models, classified in weak spatial dependence, only differing from the SDI index, were classified in moderate dependence to the Krista fertilizer model and OLS method. The maps generated in the leveled topography for water and Krista fertilizers reached the largest irrigated area with dripper flow rates between (3.08|- 3.15L/h) and, for peat, the flow rate (3.04|- 3.15L/h). Regarding their Christiansen uniformity coefficients (CUC), ≥ 97.47%, and uniformity distribution coefficient (UDC), ≥ 96.52%, there was excellent uniformity for water flow, Krista, and peat in the leveled topography of the irrigation system. On the rising slope topography, the most irrigated area was achieved for the water with flow rate drippers between (2.12|- 4.17L/h), for the Krista fertilizer (2.68|-5.32L/h), and with the peat fertilizer (3.01|-3.07L/h). This results in CUC = 98.08% and UDC = 96.48%, with excellent uniformity for the peat fertilizer; for the Krista fertilizer, CUC = 64.00% and UDC = 68.42% obtained poor uniformity; for the water, CUC = 55.00% and UDC = 45.34% resulted in unacceptable uniformity. On the falling slope topography, when water was used, drippers achieved the highest irrigated area with flows between (1.64 |- 3.26L/h) and, for Krista fertilizers, drip flows (3.11|- 3.17L/h) and peat (2.91|-3.17L/h). Regarding the coefficients CUC ≥ 95.99% and UDC ≥ 92.31%, excellent uniformity was obtained for the Krista and peat fertilizers, so it differed for the water CUC = 60% and UDC = 32.03%, obtaining a poor and unacceptable uniformity. The pressure uniformity coefficient PUC ≥ 97.57% obtained an excellent system rating for all three topographies when using water and the Krista and peat fertilizers. Thus, the analysis of the SDI and SDM indexes improved the description of the structure of spatial variability of the flows in the topographic meshes. This evaluation of the spatial variability of the flow in each topographic slope is of utmost importance because it makes it possible to define regions with critical levels of clogging, helping in the decision-making for the management at the site. Hence, this study increases the power of decision about the degree of spatial variability of agricultural attributes in the soil and irrigation, aiming at greater agricultural production and contributing to future studies