Brazil’s main sugar harvest season began on April 1 with the world awash in sugar and world sugar prices depressed. How much sugar Brazil produces this year will be the key determinant of world excess supply and price. Weather conditions in the country’s main-producing center-south region will be a key driver of the sugar crop’s size. Precipitation and temperature during the harvest season affect the sugar yield from crushing cane, and Gro’s data, particularly normalized difference vegetation index (NDVI), proves to be a strong predictor of sugar concentrations. Gro has created a model to project recoverable sugar levels during the Sao Paulo state crushing season.
NDVI is a satellite-derived measure of vegetative health and abundance in an area. It gives you an indication of crop stress. In the case of sugarcane during the harvest period, moderate crop stress is beneficial. Reduced precipitation limits the growth of the cane’s stem and leaves, but it does not unduly inhibit photosynthesis. This results in the sugarcane plant converting carbon dioxide and water into sucrose which is stored in the stalk. Crop stress indicated by NDVI can be an important basis on which to model recoverable sucrose concentrations in cane. Combining this yield with mill run rates determines how much sugar is produced.
The Brazil Sugarcane Industry Association (UNICA) provides bi-monthly reports of cane crushing activity in the center-south region. Levels of cane crushing, sugar and ethanol production, and total recoverable sugar (TRS) are released twice a month. During the main harvest period, the reports are an important factor in market forecasts and often impact the price of sugar in world markets. Using UNICA’s historic TRS data for Sao Paulo state, Gro was able to build a predictive model for TRS levels during the semi-monthly crushing period. NDVI was the most predictive factor, and temperature, precipitation and potential evapotranspiration contributed as well. The sensitivity to the respective variables shifted within the season, and linear regressions were created for each distinct period. During the July to September period, the model explained 71% of the variation in TRS.
The chart on the left shows actual total recoverable sugar levels from UNICA in Brazil’s Sao Paulo state and those forecast by Gro’s model over the past eleven years. The chart on the right compares Gro’s model output with actual UNICA values during the July to September period.