My Advice to NASAPosted in: Technology By Tracy Tjaden September 1 2013
A few months ago, Dr. Alex Melnitchouck was summoned to a meeting with NASA specialists. The Calgary, Alberta-based senior technical advisor to AGRI-TREND Geo Solutions was among an elite group of the world’s leading experts in satellite sensors invited to the workshop for NASA. U.S. space agency officials wanted to pick their brains.
Melnitchouck, whose area of expertise encompasses various aspects of spatial data management and analysis in agriculture and environment, studies how satellite sensors can be used in agriculture. Some of the most advanced work in remote sensing has been in agriculture. It has become a powerful tool for crop monitoring over large areas and single fields alike — an efficient instrument for studying field variability, water management and land assessment. Satellite imaging arms farmers with detailed information on the precise make-up of their land. Huge cost savings, better decision making — not to mention environmental gains — can result from applying inputs based on what the land needs, rather than ‘blanket’ farming.
NASA had some questions for the top brains in this specialty science, so officials from Purdue University assembled industry experts in West Lafayette, Indiana in June. Without sharing proprietary information, Melnitchouck and his fellow advisors discussed the potential of utilizing hyperspectral imagery in agriculture.
“Everybody wants to know what the future holds, what’s the next step in this area,” he says.
NASA officials wanted to know what the experts thought about using hyperspectral sensors to gather more informative images. Compared to natural color imagery that displays only blue, green and red bands, or multispectral imagery still showing only several bands from visible to long-wave infrared, hyperspectral imagery gives complete spectral signature of objects with dozens or even hundreds of spectral bands. It can provide greater accuracy in identifying field crops and various types of stress.
“We said you’ll get more accurate information, but the industry still doesn’t have the processing capacity,” he notes. “Every hyperspectral image is so huge — we don’t have enough storage or fast enough computers to make it worthwhile for an average user.” Scientists don’t yet have reliable algorithms for correlating spectral signatures of crop reflection with different types of crop stress, such as plant diseases or nutrient deficiencies. “In most cases, the information collected from satellites still requires ground truthing.”
Another challenge would be scalability. “If you don’t have global coverage there is no business model.” Melnitchouck told the NASA team he thought the science was at least 10 years away from wide application of hyperspectral imagery in agriculture.
It may not have been what they wanted to hear, but Melnitchouck says, “The overall message was that despite some technological difficulties, groundbreaking science will continue to change the face of agriculture, steadily increasing efficiency of crop production and farmers’ profit, and lessening environmental footprint of the industry.”
comments powered by Disqus