The existence of ice on the moon was revealed with a bang last year, when kamikaze spacecraft crashed into a crater at the lunar south pole, kicking up enough water for researchers to finally detect.

Today scientists in six separate studies announced new findings from the Oct. 9, 2009,  LCROSS moon crash mission. They found, among other discoveries, substantial amounts of water ice at ground zero for the impact — water that could one day be key to the humanity's future in space.

But as successful as the mission proved in the end, the complicated affair was fraught with uncertainty and came perilously close to failure. Now scientists reveal the story of how they made their discovery and what challenges they faced along the way. [10 Coolest New Moon Discoveries]

A coordinated dance

In 2009, NASA sent two missions to the moon at the same time: the Lunar Reconnaissance Orbiter, which is mapping the moon's surface, and LCROSS, short for Lunar Crater Observation and Sensing Satellite. It was LCROSS that collided with the moon.

Scientists had speculated on the existence of water on the moon since the 1940s, but it took LCROSS to help kick up significant amounts of water to reveal evidence for it within the permanent shadows in a crater at the moon's south pole.

The scientists wanted as many eyes watching the impact as possible to squeeze out every drop of knowledge they could, which meant finding the right time of day so that as many observatories in space and spread across thousands of miles on the ground could face the moon during the impact.

In the end, in addition to LRO in orbit around the moon, 22 observatories on the ground and four in Earth orbit watched the lunar collision.

"It was a very coordinated dance between a lot of different parties," said NASA space scientist Tony Colaprete, who was principal investigator for the LCROSS mission.

The two-year campaign to coordinate these ground- and space-based assets spearheaded by Jennifer Heldmann and Diane Wooden at NASA in large part involved learning how to watch the moon in the first place.

"Normally, moonlight is the enemy, always contaminating our view of the rest of the sky," Colaprete explained. "And its brightness made it hard to track with the auto-guidance systems we had."

Instead of using "guide stars" as points of light to help them find their way in the night sky, the astronomers used "guide craters" as points of darkness to help them scan the moon. "It took a lot of practice to get it all to work right," Colaprete said.

It was tricky for ground-based observatories to watch the moon, since it moves fast through the sky relative to the stars they normally look at. It was complicated for assets in Earth orbit as well, as they are moving very quickly themselves and are not necessarily made to look at something so close to our planet.

"The impact had to be timed to within a few seconds of what the Hubble Space Telescope was doing to maneuver to look at it," Colaprete said.

Although LRO was blessed with the closest view of the impact, the researchers had to make sure it actually didn't get too close, arriving over the explosion 90 seconds after it happened, or else the debris from the explosion might have damaged it.

"We had to make sure all these pieces of the puzzle were at the right place at the right time, down to the second," Colaprete said.

Verge of failure

One crisis threatened to kill off the mission entirely. Slightly more than a month before the impact, a glitch on LCROSS caused the probe to burn more than half of its remaining propellant.

"The spacecraft did a little dance without our permission," Colaprete told "I remember getting a call at 4 a.m., and thinking, 'Oh, no.'"

"It was really scary," he added. "At the rate it was burning fuel, if we had found out an hour or two later, it would've been out of fuel and dead in space. Luckily, our stellar team figured out what the problem was and how to save fuel. We got there with fuel to space."

An uncertain target

It was also hard to figure out what the best target to collide into was. Originally the researchers planned on slamming into the crater named Cabeus A, but ultimately they changed their minds to nearby Cabeus crater a little more than a week before the impact.

"At first we picked the smaller Cabeus A crater as an impact site because it was really good for observing from Earth," Colaprete recalled. But as they got closer and closer and gathered more data from LRO on the way there, based on their scans of Cabeus A for hydrogen, a signature of water, "I wasn't able to convince myself there was hydrogen there, so I decided to go for a certain target instead," he said.

"It was an agonizing decision," Colaprete said. "Not everyone was happy, since a lot wanted the ground-based observatories to have the best views they could, but I didn't want to impact someplace irrelevant."

He knew he made the right decision after a 12-hour marathon session of combing the data after the impact. "It was one of the most fun, giddy experiences I ever had, and we knew at that point we had something really special — the data was there," he said.

Ultraviolet glows

The closest "eyes" on hand to watch LCROSS crash was its partner, LRO, which brought a range of sensors to probe the results.

To identify the elements and compounds in the debris, vapor and dust the impact kicked up, LRO employed the shoebox-size Lyman Alpha Mapping Project (LAMP) instrument, which looked at the crash's plume in the ultraviolet range.

"The ultraviolet range is where a lot of materials emit and absorb, and LAMP can therefore very, very sensitively detect tiny amounts of gases," said LAMP acting principal investigator Randall Gladstone, a planetary scientist at the Southwest Research Institute in San Antonio, Texas.

LAMP normally observes the night-side lunar surface using light from nearby space and stars, which bathes all bodies in space in a soft radiance known as the Lyman-alpha glow. Since LRO was flying over the sunlit part of the moon, which is way too bright for LAMP, the spacecraft rolled over to point the instrument at the dark sky at the edge of the visible surface of the moon instead.

LAMP detected molecular hydrogen, carbon monoxide, calcium, magnesium — and, oddly, mercury.

"When one of our team suggested it was mercury when we were just throwing out ideas, well, I said, 'What a stupid idea, how can there be mercury there,'" Gladstone recalled. "I had to eat my words later."

The rock samples the Apollo missions brought to Earth showed heavier levels of mercury deeper down, which led one retired scientist, George Reed, to propose that sunlight might have baked the mercury out from the soil that then got trapped in craters at the south pole.

"He did a simple back-of-the-envelope calculation and nailed it," Gladstone told "If I made a great prediction like that, I'd be tickled that I was right."

"The detection of mercury in the soil was the biggest surprise, especially that it's in about the same abundance as the water detected by LCROSS," said LAMP team member Kurt Retherford at the Southwest Research Institute. "Its toxicity could present a challenge for human exploration."

These findings support the idea that the freezing cold of the permanently shadowed regions of the moon can trap vaporized compounds that wafted in from deep space or other areas of the moon and preserved them there for eons. "They're almost literally gold mines for science," Gladstone said.

Heat of the impact

To see how the energy of the impact dissipated, the Diviner Lunar Radiometer measured its thermal signature. They found the collision heated the crater floor from about minus 387 degrees F (minus 233 degrees Celsius) to more than 1,250 degrees F (675 degrees Celsius), enough to turn about 660 pounds (300 kilograms) of water ice directly to vapor.

"The cooling behavior we saw at the impact site can tell you about its physical properties," said Diviner Lunar Radiometer team member Paul Hayne, a planetary scientist at the California Institute of Technology. "What we saw was consistent with lunar soil containing the amount of ice that LCROSS found. The site LCROSS hit also wasn't a skating rink — it wasn't a solid block of ice. We saw thermal emissions from there up to four hours later, and ice just doesn't do that, so ice there must have been mixed with lunar soil."

The Diviner instrument had challenges of its own, Hayne recalled.

"We had never used Diviner to target a specific site on the moon," Hayne said. "Just a day beforehand, we did a trial run to see if we could accurately target the impact site, and when we got our data back, we saw we were way off, so we had to scramble at the last minute to figure out what we did wrong. Thankfully we fixed the problem and hit it spot on. It worked out great."

Colaprete, Gladstone, Hayne and their colleagues detailed their latest findings in the Oct. 22 issue of the journal Science.

This article was reprinted with permission from

Related on