I often tell people that hyperspectral imaging is much like the mass spectrometer used by the crime lab technicians on the television show, CSI. You’ve probably seen the mass spectrometer in action. A lab tech takes a substance found at the crime scene – say a reddish fluid of some sort – and runs it through the device to determine whether the fluid is ketchup, red paint, or blood.
Although the physics behind hyperspectral imaging and mass spectrometry are different, the end objective is the same – categorizing some unknown substance by comparing it to a library of known physical signatures.
As applied to natural resource exploration, hyperspectral imaging can be used to identify substances on the Earth’s surface. Every substance – soils, vegetation, minerals, and hydrocarbons – has a unique hyperspectral signature associated with it (vegetation examples below).
We can use hyperspectral signatures to identify direct, surface-based indicators like trace quantities of gold or natural gas that have leached up a fault, or to spot indirect indicators like certain clays that are associated with subsurface deposits of precious metals.
The tool is so sensitive that it can detect subtle changes in the spectral signatures of plants that are growing in areas impacted by leaching hydrocarbons (imagine being able to detect a somewhat ‘wilted plant’ that was growing near a fault along which trace quantities of natural gas were migrating to the surface, surrounded by otherwise healthy and unaffected plants of the same family).
Hyperspectral imaging is based on the physical principle that different substances absorb and reflect both visible and invisible light in different ways. By measuring the relative variations of absorption and reflectance across more than 600 channels (bandwidths) from ultraviolet to thermal infrared, a scientist can characterize and identify a previously unknown substance.
Because hyperspectral data can be acquired with airborne platforms, large swaths of land can be analyzed quickly and efficiently, making the technology well suited to natural resource exploration.
We’re not the only ones intrigued by the technology. A 2010 article in Forbes indicated that supermajor Shell has been experimenting with remote hydrocarbon detection – what the Forbes author classed as “oil sniffing.” Interestingly, it was also used by SEAL Team 6 in pre-mission planning before the raid on Osama Bin Laden’s hideout in Pakistan.
Although hyperspectral imaging can be applied stand-alone, as it was on Operation GULFscan (to monitor the effects of the Macondo oil spill on Gulf Coast ecosystems), the datasets are normally integrated and interpreted with other geological and geophysical measurements. Knowing that there are trace quantities of gas on the surface is interesting. But if one can see that the gas is traveling along a known fault system that cuts the surface (as depicted on a gravity or seismic dataset) and that the fault system extends into a potential reservoir structure (that itself shows EM anomalies), then one has real insight and the ability to develop a holistic, risk-minimizing interpretation that honors all available data. And that would make Grissom proud!
To learn more about hyperspectral imaging, watch the narrated slideshow: