A Spy Plane Hunts Critical Minerals

"There's four pilots in the world that fly this for NASA." NASA research pilot Dean Neeley told KKTV in Colorado Springs, describing the aircraft he flies. It is the ER-2, a civilian variant of the U-2 spy plane, and KKTV reported that its camera reads light reflecting off the ground from roughly 65,000 to 70,000 feet. At that altitude the pilot wears a pressure suit and the sky above him is nearly black.

What he is looking for is not a military target. It is rock.

A survey built on spectra

The USGS says the flights out of Colorado Springs began April 1 and were due to run until May 20, weather permitting. By the May 12, 2026 announcement, Andrea Travnicek, the assistant secretary of the Interior for Water and Science, said the survey had covered about 400,000 square miles. The USGS release called the effort an addition to "the world's largest airborne hyperspectral survey."

The motive is arithmetic of dependence. According to the USGS Mineral Commodity Summaries 2026, the United States relied on imports for all the titanium it used last year, and for at least 75% of its aluminum and 41% of its nickel. NASA's Jet Propulsion Laboratory adds that in 2023 the country imported its entire supply of 12 of the 50 critical mineral commodities, and at least 50% of its supply of another 29.

The USGS notes a wry detail: the ER-2 is itself built mainly from aluminum, nickel and titanium, all on the 2025 List of Critical Minerals. The plane is made of the stuff it is hunting.

How the plane finds a mineral

The instruments measure dozens to hundreds of wavelengths of light bouncing off the ground, including ultraviolet, short-wave infrared and thermal infrared. USGS scientists then read the "spectral fingerprints" to identify minerals. The newest sensor, AVIRIS-5, is about the size of a microwave oven, NASA says. JPL says its spatial resolution is finer than its predecessor's, resolving areas from under a foot to about 30 feet.

The technique has deep roots off Earth. JPL says imaging spectrometers have visited every major rocky body in the solar system from Mercury to Pluto. Robert Green of JPL said the Moon Mineralogy Mapper dataset "continues to drive our investigations as we look for in situ resources on the Moon" under the Artemis campaign. USGS Director Ned Mamula credited the "60-year partnership between the USGS and NASA" with benefits in both directions: mapping the Moon for Apollo and Artemis, and Landsat and these hyperspectral flights at home.

The ER-2 also works in tandem with orbit. JPL says the EMIT instrument on the International Space Station surveys Earth from about 250 miles up, while the plane's AVIRIS and MASTER instruments supply the fine detail that complements EMIT's broader view. The dry, treeless West suits the method, because vegetation hides rock from a spectrometer.

"We're not digging for gold. We're revealing what's hidden in plain sight." Green said of the approach.

Where the spectra stop

An anomaly is a hypothesis, not a deposit. In April, USGS geologist Erik Tharalson investigated one flagged on an earlier flight that could be tied to a porphyry copper deposit. Metal Tech News reported that he described it as a system "potentially enriched with other critical minerals." On the ground, the picture was less obliging. "It's a large outcrop of volcanic rocks, but the minerals in it would be easy to miss." Tharalson said.

That is why the last step is ground-truthing. The USGS says scientists walk prospective areas with backpack-mounted spectrometers that account for atmospheric conditions and confirm the identifications. The analysis itself takes patience. "After about three months, we can analyze the spectra, and we can detect the exact mineralogy." USGS geophysicist Raymond Kokaly told KKTV. He said the aim is to collect all the Western states, adding "it's still another four years to probably accomplish that goal."

New Space Economy is blunter about the limits. Ordinary optical and hyperspectral satellites cannot see through thick soil, vegetation, overburden or hundreds of meters of rock to image an orebody, it says, so satellites cannot replace drilling. For lithium, the method detects minerals and assemblages linked to the geology that concentrated it, not lithium atoms directly. The payoff is triage. Metal Tech News reported that hyperspectral surveys can narrow exploration targets and cut the time and cost of finding promising resources. That matters because New Space Economy cites the International Energy Agency as saying global mineral exploration spending fell by more than 10% in 2025, which leaves teams less room to waste field seasons.

Old mines, new ledger

Some of the most promising finds are in places already scarred. NASA says an early result is hectorite, a lithium-bearing clay, in the tailings of an abandoned California mine. NASA quoted remote sensing geologist Bernard Hubbard, who was speaking of old copper and silver workings, abandoned before anyone prized lithium or rare earths, when he said: "What has been a pollution source for communities could now be a resource." JPL's Dana Chadwick named two long-term aims: helping communities capture value from old prospects, and locating sources of acid mine drainage. The USGS says it is also working with academic and commercial groups on targeted surveys of legacy mine sites.

The flights feed a larger program. On September 29, 2026, the USGS released an updated database of known U.S. deposits of 59 critical minerals. Its staff compiled 1,271 known deposits in 41 states and Puerto Rico to prioritize Earth MRI surveys, and 412 of them have documented critical mineral resources and no production, which the USGS calls untapped potential. Earth MRI's re-mapping of the country supports assessments of domestic resource potential by 2031. Through the partnership, the USGS September release says, the program has amassed the "largest-ever trove of terrestrial hyperspectral data."

Orbit catches up slowly

Replacing a high-altitude plane with satellites is a longer project. New Space Economy says the thermal-infrared subsystem on the ASTER instrument was permanently shut down on January 16, 2026 because of power limits on Terra, and its shortwave-infrared subsystem has been unavailable for many years. Commercial options are arriving: Planet's Tanager captures about 426 spectral bands at 30-meter pixel spacing, and Pixxel plans to add shortwave-infrared capability with its Honeybee generation, scheduled for 2027 per company documentation. Landsat 10 is expected in 2031 with 26 spectral bands, against 11 on Landsat 8 and Landsat 9.

Europe's answer is CHIME. The Copernicus wiki, run by ESA, gives a working launch date of 2028 for CHIME-A and 2030 for CHIME-B. Its imager will capture more than 200 bands from 400 nm to 2500 nm at 30 m resolution over a 130 km swath, and raw materials exploration is among its primary drivers. Revisit is 22 days at the equator with one satellite and 11 days with two, in cloud-free conditions.

Until then, the work falls to a handful of pilots, an airframe made of the metals it is hunting, and geologists with backpacks. Kokaly's four-year clock is running. So is the shortage the clock is meant to answer.