Buzzwords, Standards, and Discovery
A buzzword with some real endurance in the earth sciences is“data fusion”. Mostly, buzzwords are harmless. They represent our aspirations to better express ourselves, if sometimes over-eagerly. The problem with “data fusion” is it often no longer means anything specific, trendy or not. That’s when people start misunderstanding each other. There are large user groups who use data fusion to mean “pan sharpening”, and only that. For others, it’s SAR and optical imagery. So now data fusion only means using 2 or more data sources.Which is not terribly remarkable. In fact, it would be a little weird to only use one information source these days. If jargon is supposed to help communicate more precisely, though sometimes more opaquely to lay audiences,then “data fusion” is a failure in terminology. But it’s still a useful tool.
Many, maybe most, other sciences don’t really have this problem. Astronomers were wise enough, at some point, to put together a board that has to clear all the names and categories, and you don’t get to opt out. Pluto is a fun Disney character, but the evidence piled up that it just isn’t a planet. Chemists have done the same thing. Discovered an element? Talk to IUPAC about naming it. It’s nice to have widely used standards, even if it ruffles some feathers. Imagine buying gasoline at a station that defined its own gallon or liter.
However, we earth scientists can’t agree on terminology in the least. In remote sensing, the analysis called MNF is either Minimum Noise Fraction or Maximum Noise Fraction, depending who you ask. How many file formats do we need? At conferences, we’re more likely to have pub crawls than meetings of names and standards committees. No, I am most certainly not against pub crawls, geologic or otherwise.
Usually, this lack of standard terms is more annoyance than actual problem, but it can lead to real issues. My colleague Peg Shippert’s article on DEMs, DTMs, DSMs, etc. (all various types of 3D models) is the most-read entry on Imagery Speaks by quite a margin. Confusion costs time and money.Peg’s done everyone a service to help clear things up.
Having naming standards sounds stodgy, but it doesn’t stifle discovery in the least. A new moon of Neptune was announced just this week. Considering we have so much yet to discover here on Earth(we’ve mapped only about 7% of the ocean bottom), maybe we should try some standards.
Some great new research was published this week about a previously unknown chain of ancient volcanoes on the ocean floor between South America and Antarctica. It is critical to our understanding of ocean currents and climate. Not to mention it was a whole bunch of volcanoes we didn’t know about. New maps of Antarctica, with and without ice, are also a big advance in understanding the South Pole and our planet as a whole. None of it would have happened without combining data from LiDAR, sonar, EO imagery, SAR, gravity surveys, drilling, dredging, and more.
Whether or not you call it data fusion, using any and all data that might help shed light on your problem is more important than ever before. Fortunately, it’s also easier than ever before with modern software and analytics. I’m working on a project in ENVI to combine Landsat data with LiDAR,sonar, and bathymetric data to better understand coastal regions. What data are you using? Have you looked at other fields to see what they use? It might be the big new advance you need in your work, and that’s when discovery can happen.