Sensitivity analysis and parameterization of two agricultural models in cauliflower crops

Aim of study: The development of a procedure to calibrate the LEACHM and EU-Rotate_N models for simulating water and nitrogen dynamics in cauliflower crops. Area of study: Calibration was performed using experimental data obtained from measurements in a cauliflower crop sited in Valencia (Spain) re...

Full description

Bibliographic Details
Main Authors: Antonio Lidón, Damián Ginestar, Sofía Carlos, Carlos Sánchez de Oleo, Claudia Jaramillo, Carlos Ramos
Format: Article
Language:English
Published: Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria 2020-02-01
Series:Spanish Journal of Agricultural Research
Subjects:
Online Access:http://revistas.inia.es/index.php/sjar/article/view/15314
Description
Summary:Aim of study: The development of a procedure to calibrate the LEACHM and EU-Rotate_N models for simulating water and nitrogen dynamics in cauliflower crops. Area of study: Calibration was performed using experimental data obtained from measurements in a cauliflower crop sited in Valencia (Spain) region. Material and methods: A procedure based on generalized sensitivity indices for time-dependent outputs was used to determine the most influencing model parameters, in order to reduce the number of parameters to be calibrated and to avoid overparameterization. The most influencing parameters were introduced in an optimization process that uses the experimental measurements of soil water and nitrate content to determine its optimal value and obtain calibrated models. Main results: After this analysis, the most important hydraulic parameters found were the coefficients of Campbell’s equation for the LEACHM model and the soil water content at field capacity and drainage coefficient for the EU-Rotate_N model. For the N cycle, the most influencing parameters were those related with the nitrification, humus mineralization rate and residue decomposition for both models. Both calibrated models provided good simulation of soil water content with an error between 5-7%. However, larger errors in soil-nitrate content simulation were found, mainly in the period corresponding to the crop residues incorporation. The prediction of the calibrated models in a different plot gave error values of about 7-9% for soil water content, but for soil nitrate content errors computed were 34% and 58%. Research highlights: After calibration, both models can be used to optimize the farmer water management and fertilization practices in horticultural crops, although in the N case further studies should be performed.
ISSN:2171-9292