Articles
DETERMINATION OF CROP COEFFICIENTS FOR HORTICULTURAL CROPS IN CUBA THROUGH FIELD EXPERIMENTS AND WATER BALANCE SIMULATION
Article number
889_60
Pages
475 – 482
Language
English
Abstract
A proper soil water balance model provides support to improved irrigation management.
With this objective, the water balance simulation model ISAREG was calibrated and validated for micro-sprinkler irrigated sweet pepper, garlic, onion, cabbage and carrots, using field observations performed in the Irrigation Station of Alquízar, south of La Havana.
Model calibration and validation were performed using two independent data sets for each crop.
The calibration refers to the crop coefficients (Kc) and the soil water depletion fraction for no stress (p), and to the soil hydraulic properties of a Red Ferralitic compacted soil.
The calibration procedure consisted of adjusting first the soil properties and then, through an iterative procedure, to determine the Kc and p values that minimize the differences between observed and simulated soil water content along the crop season.
The validation was performed using the calibrated Kc and p with a different climate and crop data sets.
The following indicators of goodness of fitting were used to assess model calibration and validation: regression coefficient forced to the origin (b), determination coefficient (R2), root mean square error (RMSE) and average absolute error (AAE). Results show a good agreement between field observations and model predictions for all the five crops, with b close to 1.0, R2 ranging 0.84-0.95 for the calibration and 0.78-0.90 for the validation.
The RMSE ranged between 0.97 and 2.08 mm for the calibration and between 1.07 and 2.82 mm for the validation.
The Kc and p values are in the range suggested in FAO 56 manual and other literature papers.
Results allow further using the model to develop alternative irrigation schedules aimed at improved water use and productivity in Cuba.
With this objective, the water balance simulation model ISAREG was calibrated and validated for micro-sprinkler irrigated sweet pepper, garlic, onion, cabbage and carrots, using field observations performed in the Irrigation Station of Alquízar, south of La Havana.
Model calibration and validation were performed using two independent data sets for each crop.
The calibration refers to the crop coefficients (Kc) and the soil water depletion fraction for no stress (p), and to the soil hydraulic properties of a Red Ferralitic compacted soil.
The calibration procedure consisted of adjusting first the soil properties and then, through an iterative procedure, to determine the Kc and p values that minimize the differences between observed and simulated soil water content along the crop season.
The validation was performed using the calibrated Kc and p with a different climate and crop data sets.
The following indicators of goodness of fitting were used to assess model calibration and validation: regression coefficient forced to the origin (b), determination coefficient (R2), root mean square error (RMSE) and average absolute error (AAE). Results show a good agreement between field observations and model predictions for all the five crops, with b close to 1.0, R2 ranging 0.84-0.95 for the calibration and 0.78-0.90 for the validation.
The RMSE ranged between 0.97 and 2.08 mm for the calibration and between 1.07 and 2.82 mm for the validation.
The Kc and p values are in the range suggested in FAO 56 manual and other literature papers.
Results allow further using the model to develop alternative irrigation schedules aimed at improved water use and productivity in Cuba.
Authors
Y. Chaterlan, M. León, C. Duarte, T. López, P. Paredes, L.S. Pereira
Keywords
soil water balance, simulation model, crop coefficients, garlic, onion, cabbage, carrots, sweet pepper
Online Articles (79)