Análise de Sensibilidade dos Coeficientes Genéticos do Modelo Canegro/DSSAT

Abstract

The sensitivity analysis of genetic coefficients used in the productivity simulation of sugarcane by Canegro/DSSAT model (CD) was performed in order to identify which parameters have most relevance in the model calibration. The CD uses 20 parameters that aim to capture the differences between the varieties of sugarcane. The sensitivity analysis has been developed varying one parameter at a time within a certain range and keeping the others constant. Were simulated the stalk dry weight, leaf area index of green leaves and sucrose dry weight at the end of each crop cycle in a period of twenty years, between 1995 and 2014, considering three planting dates, August 01, September 01 and 01 October. The simulations were made to the city of Pelotas - RS and the results were evaluated by the standard deviation (D). The results show that genetic coefficients with great sensitivity in the simulation of the stalk dry weight were PARCEmax (D ≈ 6 t.ha-1) and STKPFmax (D ≈ 5 t.ha-1). In the sucrose dry weight simulation, the genetic coefficients with higher sensitivity were PARCEmax (D ≈ 3,5 t.ha-1) and STKPFmax (D ≈ 3 t.ha-1). In the simulation of the maximum leaf area index, the more sensitive genetic coefficients were LFMAX (D ≈ 2 cm2 .cm-2), TT_POPGROWTH (D ≈ 1,4 cm2 .cm-2) and Mxlfarea (D ≈ 1 cm2 .cm-2). The results also show that differences in the planting date have influence in the sensitivity of genetic coefficients

Keywords

• Sensitivity analysis
• Genetic coefficents
• Canegro
• Sugarcane