Introduction

Gro’s Yield Forecast Models use a suite of machine-learning models to estimate in-season yields at the district, province, and/or national levels on a daily basis. Gro’s Brazil Corn Yield Forecast Model is updated daily, providing in-season yields for corn at the district, province, and national levels throughout the growing season.

Customers Use the Model to:

  • Track corn production in the world’s third-largest corn-producing country
  • Monitor crop production, the most significant part of the balance sheet, by tracking its most variable component, which is in-season yield
  • Gauge crop availability and crop prices
  • Understand how weather impacts yields in microclimates
  • Inform other models focused on damage caused by pests and diseases

Why It Matters

Brazil’s second crop corn area, or safrinha, has grown significantly on the back of soybean expansion. Currently, Brazil is the world’s third-largest corn producer and the second-largest exporter of corn in the world. Crucially, the country’s main export window overlaps with the US harvest, pushing the main US export window into Q1. Corn production is concentrated in four of Brazil’s 26 states: Mato Grosso, Minas Gerais, Goias, and Parana. When combined with our other Corn Yield Forecast Models, for China, the United States, and Argentina, Gro is able to forecast more than 70% of global corn production in a given year, providing users with a highly accurate, real-time window into global corn supply.

Methodology

Modeling corn yield in Brazil presents particular challenges, primarily because there are two corn-growing seasons per year. Further, the breakdown of the two annual crops has changed: the second-season crop has become more dominant, as soybeans have taken priority over the first corn crop during the summer season. In addition, significant differences in weather, soil conditions, and technology adoption around the country results in high spatial variation of crop yields. Gro’s Brazil Corn Yield Forecast Model uses the following variables:

  • Latitude and longitude
  • Gro Crop Cover
  • Yield, area and production history in Brazil
  • Eight-day normalized difference vegetation index (NDVI)
  • Land surface temperature
  • Rainfall
  • Weather forecasts of air temperature and precipitation
  • Evapotranspiration difference from 10-yr median (2003-2013)
  • Drought Severity and Coverage Index
  • Soil characteristics
  • Latitude and longitude
  • 30-meter resolution cropland masks

Related Analytics