Andrew M. Davidson - Associate Director - Agriculture Canada's AgroClimate, Geomatics and EO, Talk: Space for Agriculture: Challenges & Opportunities to supporting a Competitive Agricultural sector in Canada.

Remote sensing is the most cost-effective means for gathering timely detailed and reliable information over large areas with high revisit frequency. The integration of remote sensing data with national statistics, field observations and secondary data show great potential for mapping crop acreages. This requires the selection of appropriate satellite data, the collection of quality ground information, the application of suitable pre- and post-processing methods and the implementation of a robust classifier. However, this is a non-trivial task because cropping systems are often diverse and complex, and remote sensing approaches need to be adapted to local cropping systems and environmental conditions.

Agriculture and Agri-Food Canada has taken advantage of recent advances in satellite and sensor engineering and the proliferation of space-based EO missions to map agricultural land use and its change. This includes image data acquired by a multitude of satellites/sensors that span the optical and microwave regions of the electromagnetic spectrum and a range of spatial resolutions. In the future, AAFC’s ability to meet the informational needs of the sector is expected to improve as new satellite sensors are launched, brought online and made available (e.g. RADARSAT-Constellation). The EO approach will be critical to the development of AAFC’s next generation of useful and authoritative informational products.

One of the most valuable of the EO-based information products produced operationally by AAFC is the Annual Space-Based Crop Inventory for Canada (ACI). The ACI – updated annually and available free of charge to the public via the Government of Canada`s Open Data Portal – comprises a gridded map of the agricultural land use and non-agricultural land cover found within Canada’s agricultural extent. By providing highly accurate field-level information on detailed crop types, pasture and grassland, the ACI is an important foundational dataset for supporting the development of programs and policy at AAFC. Beyond AAFC, the ACI provides important land use information for other Government of Canada departments, various Provincial Governments, producers, non-governmental organizations, universities and colleges, the private sector and the public.

Despite the demonstrated success of the ACI, there are areas of development that must be addressed if it is going to be able to meet the future needs of AAFC and its clients in a cost-effective and computationally efficient manner. Of particular interest is the production of within-season crop acreage estimates that would be released as precursors to the final ACI product release. Achieving this objective will require: (a) negotiations for earlier access to the ground observations provided by provincial agencies that are used to train and validate the ACI; (b) the further optimization of classification routines to reduce the computational overhead that makes within season-estimates impossible; and (c) the incorporation of new and future data streams. In addition, transitions from the research to operational domains must be facilitated. Operational systems need to be dynamic, and require ongoing improvements based on changing needs, emerging research and ever-improving and changing satellite data streams and in situ data networks.