Simulation-Optimization Modelling of Yield and Yield Components of Tomato Crop

Abstract

This study simulate and optimize the yield and yield parameters of tomato using AquaCrop model and genetic algorthm (GA) respectively. The AquaCrop model was firstly calibrated using the data obtained from the field and was later used to simulate the observed yield, water productivity and biomass of tomato. The Root Mean Square Error (RMSE), Coefficient of Residual Mass (CRM) Normalized Root Mean Square Error (NRMSE) and Modelling efficiency (EF) were used to compare the observed and simulated values. The governing equation of AquaCrop simulation software was then optimized using the evolutionary optimization method of GA with MATLAB programming software. All the statistical indices except CRM used in comparing the simulated and observed values indicated good agreement. The CRM values of -0.11, -0.06 and -0.20 were obtained for the yield, biomass and water productivity of tomato which indicated a very slight over-estimation of the observed results by the AquaCrop model. The optimization algorithm terminated when the optimal values of yield and biomass were 4.496 〖ton ha〗^(-1) and 4.90 〖ton ha〗^(-1) respectively. The GA revealed that the yield and biomass of tomato can be increased by 57% and 23% respectively if the optimized parameters were either attained on the field experiment or used during simulation. Thus, the study ascertained that crop simulation models such as AquaCrop and optimization algorithms can be used to identify optimal parameters that if maintained on the field could improve the yield of crops such as tomato.