Your Reading List

Crop forecasting in Canada

Stats Can’s crop production estimates can have a big impact on prices. 
Find out how they come up with these numbers

Statistics Canada (StatsCan), a federal government department, conducts the farm surveys that obtain information about crop production, seeded areas, harvested areas, expected yields and production in Canada.

Stats Can surveys

The surveys are fielded in a series of six data collection periods and the information is released at pre-scheduled, strategic times during the calendar year. The intention behind the timing of the surveys and the reporting periods is to provide publicly available and accurate estimates of seeding intentions, seeded and harvested areas, production, yield and farm stocks of the principal crops in Canada at a provincial level. StatsCan gathers information on wheat, oats, barley, rye, flaxseed, canola, corn for grain, soybeans, sunflower seed, dry beans, dry field peas, lentils, mustard seed, canary seed and chickpeas.

Area, yield and production data are released in March, June, July, September and November. Stocks estimates are available at the end of December, March and July/August. The collection period for each survey starts about four weeks prior to the release of the results and may last from six to 10 days depending on the size of the survey. The survey size differs by month, as shown in the table.

The survey is mandatory. Generally by the end of the collection period 80 per cent of the questionnaires are fully completed. Estimates are available on Statistics Canada’s key socioeconomic database (CANSIM) that also contains a large range of other publicly available statistics the agency collects.

Satellite imaging

StatsCan also hosts a free web mapping application that provides weekly cropland and pasture condition reports across Canada and the northern U.S. called the Crop Condition Assessment Program (CCAP).

Satellites carrying Advanced Very High Resolution Radiometer (AVHRR) record images of the entire Earth’s surface twice a day at a one kilometre resolution. This detector has been found useful for vegetation monitoring to produce Normalized Difference Vegetation Index (NDVI) images. NDVI images work on the theory that healthy, thick vegetation will absorb more visible light than bare soils.

Since 2010 there has been a further enhancement using the Moderate Resolution Imaging Spectroradiometre (MODIS). MODIS operates at a spatial resolution of 250 metres, a significant enhancement. The MODIS sensor can to show vegetation conditions (NDVI) at a higher resolution with more accuracy than the AVHRR imagery.

The Remote Sensing and Geospatial Analysis Section (RSGA) within the Agriculture Division of StatsCan is responsible for this data collection and has built an online interactive mapping interface. Image products show vegetation conditions on a pixel by pixel basis, while map products illustrate the predominant vegetation condition by regions down to areas as small as townships.

This type of data collection allows more frequent updates of the crop condition and potential yield, well in advance of StatsCan traditional crop surveys.

There was talk over the summer about the possibility of StatsCan changing the way it develops its estimates. Jeffrey Smith, director of Statistics Canada’s Agriculture Division says StatsCan is comparing survey results with estimates derived from satellite imagery combined with administrative and meteorological data.

Smith says, “Any change would only be entertained for the September survey, an early harvest survey, and while we might reduce costs by making this change, the bigger impetus to consider a change is to reduce the reporting burden on farmers while they are busy with harvest.”

If a change is to be made, the decision will be made in March, 2014 and would impact the September, 2014 survey. “Satellite imaging data alone would not be sufficient to make estimates,” says Smith. “The use of other sources of data is required.” †

About the author

Andrea Hilderman's recent articles



Stories from our other publications