Data analytics professionals have a wealth of Earth observation data to deal with, so several representatives of industry suggested ways to manage the overwhelming flow of information at a Canadian Space Agency (CSA) conference.
Around 150 delegates attended the National Forum on Earth Observation from Space at CSA headquarters in Longueuil, Que. (which SpaceQ attended by teleconference.) This included not only representatives from Canada, but also from other countries who could point the way to ideas that our country could adopt.
One of those international representatives was Ian Jarvis, the program director of GEO Global Agriculture Monitoring Initiative (GEOGLAM). While it’s a program that few may know of, Jarvis told delegates, it is a program borne out of G20 discussions regarding the importance of monitoring food price volatility and food security.
GEOGLAM produces a monthly forecast of crop yields in four key agricultural sectors worldwide: maize, rice, wheat and soybean. “We believe it has a significant impact on market volatility, in a positive way, over time,” he said. The product has spawned a related forecast for food security, which includes an early-warning system so that governments can take action when problems begin to arise.
Earth observation data can also be adopted for other purposes, said François Bellavance, a professor in HEC Montreal’s department of decision sciences. While there are a wealth of products that could be produced from Earth observation data, he urged the delegates to look to those with the best business cases.
Examples of possible business cases could include products for the insurance industry (such as having an almost real-time monitoring of snow accumulation on rooftops, which would allow customers to clear their roofs preventatively) or for officials concerned about evacuation procedures at large outdoor events, since the crowds could be monitored by satellite.
Trevor Dhu, the program director for the Digital Earth Australia Program, provided an overview of Australian data products and urged the CSA and other government departments to remember to make their data products user-friendly. “The government needs to collect the data, they need to calibrate it, and to make it easy for others to consume it,” he said.
And Arnold Hougham, vice-president of sales and marketing at PCI Geomatics, provided a historical perspective of data analytics. He said that practitioners in past decades had to engage in “pixel torture” or “beating the crap out of an image” in an attempt to get useful information, but today’s high-resolution products provide the opposite problem – trying to decide what information to extract first.