Statistics Canada released its first crop production estimates using a new model-based methodology on Thursday.
The model takes into account satellite data, as well as historical figures from the traditional StatsCan Field Crop Reporting Series and agroclimatic information. What wasn’t required in compiling the numbers was a survey of farmers.
“The traditional survey relies completely on interviewing farmers. The model-based estimates require no contact at all. We’re relying exclusively on satellite data, climatic data and the historical database,” explains Frederic Bedard, a senior analyst in remote sensing and geomatics application with StatsCan, in the interview above.
He says the agency’s goal is to replace the survey-based September report in the Field Crop Reporting Series.
“That will eliminate a few thousand phone calls that we run every year. It’s a burden that will be less on the farmers, as well it will reduce costs and make results earlier as well,” notes Bedard.
For 2015, the model-based report was released as a supplemental dataset in advance of the September survey results (scheduled to be published on October 2nd.)
Other countries are still experimenting with using satellite-based models for official estimates, says Bedard.
“This new approach has been tested in other countries, but I believe Canada is the first to officially adopt this method for estimating crop yields,” he says. “We’re pretty proud of this accomplishment.”
How do the model-based estimates compare?