SAR Research and Operational Use at Agriculture and Agri-Food Canada Copy
Space-based Earth Observation is one of the few reliable methods to get detailed information describing the changing state of Canada’s agricultural landscapes from coast to coast. Dr. Andrew Davidson familiarizes you with how the federal department of Agriculture and Agri-Food Canada (AAFC) uses SAR data operationally to monitor Canada’s vast agricultural landscapes and provide an annual inventory of major crops.
The following video introduces you to various lesson components and the Earth observation operations at AAFC.
Agriculture and Earth Observation (EO)
Did you know that AAFC has taken advantage of recent advances in EO data acquired by a multitude of satellites. Their sensors are including the optical and microwave regions of the electromagnetic spectrum and a range of spatial resolutions.
Did you know that new SAR sensors such as the Canadian RADARSAT Constellation Mission and the European Sentinel-1 are critical sources of data for the development of reliable information products at Agriculture and Agri-Food Canada? Watch the following video for details.
Canada’s Annual Crop Inventory (ACI)
One of the most valuable EO-based information products produced operationally by AAFC is the Annual Crop Inventory. The ACI is updated annually. It is available free of charge to the public via the Government of Canada’s Open Data Portal.
The product comprises a gridded map of agricultural land use and non-agricultural land cover of Canada. In the following video you will gain insight into Canada’s Annual Crop Inventory.
Some Notes on Producing the ACI
Canada’s agricultural landscapes are extensive and complex. Earth-orbiting satellite sensors are an essential data source for acquiring timely and relevant information on Canadian agricultural production trends.
In 2009, AAFC took its first steps towards the operational delivery of this information by developing a software system for mapping crop types using satellite observations. This system is based on two decades of remote sensing research and applications development at AAFC and elsewhere; it has been used for the Prairie Provinces of Alberta, Saskatchewan and Manitoba during 2009 and 2010, and it has been extended for all Provinces from 2011 onwards.
To create the digital crop inventory, a Decision Tree (DT) methodology is applied to optical (Landsat-8, Sentinel-2) and radar (RCM, Radarsat-2) satellite images. The DT algorithm uses the known crop types of certain locations on the ground to spectrally differentiate each of the crop types being mapped. These relationships are then applied to the satellite image data to identify the most likely crop type of each field in the study area.
More than 2000 satellite images – each linked to thousands of ground data points – are required to map Canada’s entire agricultural extent annually and validate the resulting product. Hundreds of hours of computer processing time are required to produce a final high-quality classification. So far, AAFC can consistently deliver a crop inventory that meets the overall target accuracy of at least 85%. The annual crop inventory maps have already been applied to address many needs for the sector.
Recent Changes in Soybean Acreages
A changing climate is expected to create shifts in where and how crops are grown. In Canada, climate change could open up new acreages to production and change cropping patterns.
Here is an example:
Soybeans are a long-season, heat-loving crop. Over the past decade, the Province of Manitoba is experiencing fewer killing frosts during the early fall season. This development has encouraged growers to plant soybeans. Soybean acreages have increased significantly.
You can follow the expansion of soybean fields in the Province of Manitoba (MB) from 2009 to 2015 in the following animation. The distribution maps are based on SAR image analysis and classification.
Pre-Processing of Satellite Data
Our research suggest that optical and SAR data are required in order to characterize key crop-growing stages with high accuracy. The combination of optical and SAR data provides unique and valuable information relating to plant type and growth.
Before using optical and SAR data as input to generate the Annual Crop Inventory, some pre-processing of these data is required. The following video introduces you to the various pre-processing steps.
Further references and links related to this sections are located here.
Crop Classification and Monitoring
This topic has introduced you to how AAFC uses radar remote sensing for mapping and monitoring of Canada’s agricultural landscape. Next, you will learn about radar backscatter mechanisms that are relevant for imaging and mapping agricultural fields.
Please complete the quiz below to advance to the next topic.
Click the ‘Submit’ Button to rate this topic.