Modeling Soil Moisture, Growth, and Yield in Common Beans: Calibration and Validation of AquaCrop in Northeast Brazil

Agricultural models based on calibration and validation processes that consider meteorological conditions, soil and crop parameters and management practices are promising tools to support decision-making in agriculture. Thus, the objective of this study was to calibrate and validate the FAO AquaCrop model for common bean crop in the region of Rio Largo, Alagoas, Brazil, simulating soil moisture, total dry biomass (TDB), grain yield (GY), and canopy cover (CC). The model was calibrated with data from an experiment conducted from 11-17-2015 to 02-01-2016, changing the non-conservative parameters until obtaining the lowest percentage deviations between measured and simulated values of final TDB and GY. The model was validated using meteorological and soil data from another experiment, conducted from 12-16-2004 to 03-08-2005, and a previously calibrated crop file. The model performance in calibration and validation were evaluated based on the root mean square error (RMSE), Willmott’s index of agreement (d), Pearson’s correlation coefficient (r), standard deviation, and the confidence index (c). The model calibration model showed RMSE of 0.25 t ha^-1 (TDB) and 15.58% (CC), and a very good confidence index (c > 0.76) for these variables. The validation was very good for TDB (c = 0,81) and poor for CC (c ≤ 0.40). The crop yield simulations showed percentage deviations of 0.09% (calibration) and 11.06% (validation). AquaCrop model satisfactorily simulates soil moisture and canopy cover, with better performance for TDB and GY.