Agrometeorological models for forecasting yield and quality of sugarcane

Abstract

Climate is an important factor in sugarcane production, and its study is fundamental for understanding the climatic requirements of the crop. We developed regional agro-meteorological models to forecast monthly yields in tonnes of sugarcane per hectare (TCH) and quality of the total recoverable sugar (ATR). We used monthly climatological data (air temperature, precipitation, water deficiency and surplus, potential and actual evapotranspiration, soil-water storage, and global solar irradiation) of the previous year to forecast TCH and ATR for the next year using multiple linear regression. The parameters of monthly climatological data were chosen for their small mean absolute percentage errors (MAPEs) and p < 0.05 and ability to model longer periods of prediction. Data for Jaboticabal, a major area of sugarcane production in the state of São Paulo, Brazil, from 2002-2009 were used for calibration, and data from 2010-2013 were used for validation. All calibrated models were significant (p < 0.05) and accurate, with a MAPE of 4.06% for the forecast of TCH in the ambient C for July. The model calibrated for November had variable water deficits in all environments, showing the importance of this variable to the crops. The monthly models tested performed well. For example, the forecast by TCHMAY in the AB environment (MAPE = 1.89% and adjusted coefficient of determination = 0.90) overestimated the average yield of 90.6 t ha-1 by only 1.7 t ha-1. The predictive period for forecasting TCHMAY was eight months when the last climatological parameter used in the model was DEFSEP.