Water and solutes contents in soil: performance and sensitivity of Hydrus-1D model

Author(s)

Roque Emmanuel da Costa de Pinho

• Document type: Master's Dissertation
• Press: Piracicaba.
• Institution: Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC)
• Defense date: 2010-01-15

Examining board members

Jarbas Honorio de Miranda; Alberto Colombo; Quirijn de Jong van Lier

• Advisor: Jarbas Honorio de Miranda

Abstract

The concern about the fate of chemical products and water, applied to the soil, has been motivating several researchers to develop and apply theoretical models, aiming to describe the physical processes involved in the transport of those products in soil profile. The present research had as objective the application of the model Hydrus- 1D for water and solute (nitrate and potassium) simulation profile, in laboratory conditions, using soil columns filled with two types of unsaturated soil (Oxisol (Haplustox) and Hapludox), as well to evaluate the acting and sensitivity of the model. The transport parameters were obtained for each solute by breakthrough curves (BTCs). The transport parameters obtained and the boundary conditions were inserted in the Hydrus-1D model to realize the simulations. The acting of the Hydrus-1D model was evaluated using the statistical indicators: maximum error, mean absolute error, normalized root mean-square error, coefficient of residual mass, determination coefficient, efficiency and Willmott concordance index. The model sensitivity was evaluated by the method proposed by McCuen and Snyder (1986) and applied to the parameters: input flow, soil saturated hydraulic conductivity, water content (saturation point), alpha and n (soil water retention curve parameters), distribution coefficient and dispersivity. The obtained results, experimentally, showed the nitrate displacement following the wetting front and, in relation to the potassium ion, a larger retention was observed at the superficial soil columns layers. Therefore, was possible to conclude that Hydrus-1D model was efficient for both water and potassium displacement simulations, for both studied soils and by the balance model for solute transport in soil, the Hydrus- 1D model was not efficient to simulate the nitrate displacement in both soils. The experimental procedures to estimate transport parameters, by the BTCs, were enough to describe the potassium movement in soil, generating input information necessary to the model simulation. To the simulation of the nitrate displacement, was possible to conclude that the model Hydrus-1D requests a more detailed experimental process, being necessary the estimate of a larger number of transport parameters. In relation to the sensitivity analysis of the Hydrus-1D model, when simulated the content of potassium and water at the soil columns, was observed that the model shows more sensitivity about to the parameters: water content (saturation point) and input flow. This parameters need to be estimated with more precision. There was a low sensitivity to the parameters soil saturated hydraulic conductivity and dispersivity for both studied soils. (AU)

• FAPESP's process: 08/02054-2 - Model HYDRUS-1D applied to the simulation of soil moisture and solute profile in soils
• Grantee: Roque Emmanuel da Costa de Pinho
• Support Opportunities: Scholarships in Brazil - Master