All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. S ugarcane is of major social and economic importance in Brazil. It is one of the most important commodities in Brazilian agribusiness, contributing to the energy and food security of the country, as sugar, ethanol, and biomass for energy are produced from sugarcane (Goldemberg, 2007). Brazilian production of sugarcane has been expanding since the early 2000s, now reaching areas where it has never been planted, mainly driven by the rise in ethanol consumption in the internal market. The coefficient of variation of annual national production was 7.2% from 1990 to 2008, ranging from 5 to 12% among different regions. The new production areas, mainly located in Central and Northeast regions are more often subject to higher risk. Crop simulation models may contribute to improved crop monitoring and yield forecasting, while enhancing our understanding of sugarcane growth and yield. Worldwide, there are several models dedicated to sugarcane crop simulation: AUSCANE (

No abstract

No abstract

Resumo -O objetivo deste trabalho foi parametrizar e avaliar o modelo DSSAT/Canegro para cinco variedades brasileiras de cana-de-açúcar. A parametrização foi realizada a partir do uso de dados biométricos e de crescimento das variedades CTC 4, CTC 7, CTC 20, RB 86-7515 e RB 83-5486, obtidos em cinco localidades brasileiras. Foi realizada análise de sensibilidade local para os principais parâmetros. A parametrização do modelo foi feita por meio da técnica de estimativa da incerteza de probabilidade generalizada ("generalized likelihood uncertainty estimation", Glue). Para a avaliação das predições, foram utilizados, como indicadores estatísticos, o coeficiente de determinação (R 2 ), o índice D de Willmott e a raiz quadrada do erro-médio (RMSE). As variedades CTC apresentaram índice D entre 0,870 e 0,944, para índice de área foliar, altura de colmo, perfilhamento e teor de sacarose. A variedade RB 83-5486 apresentou resultados similares para teor de sacarose e massa de matéria fresca do colmo, enquanto a variedade RB 86-7515 apresentou valores entre 0,665 e 0,873, para as variáveis avaliadas.Termos para indexação: Saccharum, biometria, calibração, modelagem, validação. Parameterization and evaluation of the DSSAT/Canegro model for Brazilian sugarcane varietiesAbstract -The objective of this work was to parameterize and to evaluate the DSSAT/Canegro model for five Brazilian sugarcane varieties. The parameterization was done using biometric and growth data from the varieties CTC 4, CTC 7, CTC 20, RB 86-7515, and RB 83-5486, obtained in five Brazilian locations. Local sensitivity analysis was performed for the main parameters. Model parameterization was done using the generalized likelihood uncertainty estimation (Glue) method. The predictions were evaluated using the coefficient of determination (R²), Willmott's D-index, and the root mean square error (RMSE) as statistical indexes. CTC varieties showed D values ranging from 0.870 to 0.944 for leaf area index, stalk height, tillering, and sucrose content. The RB 83-5486 variety showed similar results for sucrose content and fresh matter mass of stalks, whereas the RB 86-7515 variety showed values ranging from 0.665 to 0.873 for all variables analyzed.