The ATLAS Project

Motivation

We have all worked for years on the Pan-STARRS project, which is also trying to find near-Earth objects (NEOs) among other things. Pan-STARRS is designed to discover NEOs throughout the Solar System, long before they might strike us. This is an extremely difficult job, and although the warning time provided by Pan-STARRS can be very long, Pan-STARRS will have a very poor chance of finding very small asteroids, even if they are going to impact the Earth. Pan-STARRS observes "deep but narrow", needing months to patrol the whole sky.

It occurred to us that a complementary job could be done by observing "shallow but wide". Any asteroid that is actually going to hit us will eventually get very close to the Earth on its final approach, and that will make it very bright. So if we can relieve ourselves of the need to look all the way across the Solar System we can afford to canvas the sky much more frequently.

ATLAS will not provide enough warning time to deflect the asteroid -- it's intended to provide warning of impending impact. As with the asteroid impact over Sudan in 2008, ATLAS will provide exquisite accuracy on the time and location of the impact.

The amount of warning time is proportional to how distant the asteroid is when we first spot it, and that depends on how big the asteroid is and on how faint our system can see things. We decided that a large explosion of a few Mton required a week's warning and a giant explosion of 100 Mton would need three weeks of warning in order for effective civil defense.

The trick to having a good chance of seeing an asteroid before it hits us is to look everywhere possible in the sky. There is a big portion of the sky towards the Sun, about one quarter of the total, that is impossible to look into from the ground without seeing daylight and losing all sensitivity. Another quarter of the sky in the south cannot be seen from the northern hemisphere. This leaves half of the entire sky that can in principle be seen over the course of a night from a single, north latitude site. We decided that we ought to try to cover that entire area twice a night. This gives us a decent chance of observing an incoming asteroid that is so small that we can only detect it on the very last night before it hits us. Of course this corresponds to a relatively harmless asteroid of only a few meters in diameter that probably would not survive passage through the atmosphere.

Technique

It's simple enough to survey the entire sky with a 35mm camera, but the sensitivity doesn't give us enough warning time. It's easy enough to survey portions of the sky to very good sensitivity (which is what Pan-STARRS and the other NEO surveys are doing), but the coverage is slow enough that a large fraction of impactors can sneak by when the system is looking elsewhere.

Proposed ATLAS design.

We realized that it's possible to use an array of 10-inch, commercial telescopes with some really hot CCD cameras that we know how to build to do this scan of the entire, visible sky twice a night. We also know how to write software that will automatically sift through 500MB a minute in real time right down to the photon limit, to discover everything that has changed or moved since the last time we looked. We're proposing to build a set of eight such telescopes and cameras on two mounts at two locations as illustrated.

The advantage of having two sites is that we can triangulate the 3D position of asteroids that are within a week of impact, and therefore know immediately whether or not it's something to worry about. ATLAS will issue alerts to the Minor Planet Center in real time of any new detections, and we are confident that anything threatening would immediately trigger many other professionals and amateurs to start collecting data that will verify the impending impact and refine the orbit so that we can predict the moment and place of impact accurately.

ATLAS can detect objects to magnitude 20, which is astronomer-speak for "respectably, but not extremely faint". It corresponds to a match flame in New York viewed from San Francisco (if somebody would move the intervening Earth out of the way!). When we run simulations of how well ATLAS will perform, allowing for weather and using a realistic distribution of approach directions of impactors, we find that ATLAS has a better than 50% chance of seeing a 50m (160 ft, few-Mton) impactor and a better than 60% chance of seeing a larger, 140m (500 ft, 100 Mton) asteroid, as shown on the left. Basically, ATLAS will find every object bigger than 50m diameter that is coming towards the Earth in the night sky, but there's a characteristic leveling-off of the probability right around 60% because 40% of the sky is in daylight and due to the southern blind spot. Much larger asteroids we have a better chance of seeing because we'll spot them many orbits before their final approach. The chances of seeing smaller, less dangerous asteroids is less because ATLAS cannot see them until they are much closer, so the probability of having at least one clear chance to observe it are lower.

When ATLAS does spot an incoming asteroid there is a distribution of warning times, depending on whether the incoming trajectory is favorable to make the asteroid bright and on the luck with the weather. The graph on the left shows that ATLAS typically provides a week's notice for incoming asteroids of size 50m and about 3 weeks notice for those 140m or larger.

Funding

We think that constructing and operating ATLAS is a difficult, challenging, but ultimately straightforward job. Experience building Pan-STARRS and other projects has taught us a lot about how to get things built and going. To the greatest extent possible we want to just buy things like telescopes, mounts, enclosures, and the like; we want to just recycle things like camera designs, and software; and we want to use CCD detectors that we already have available. We also want this system to operate robotically, requiring only very light human oversight to make sure that things are working properly.

We think we can build one system (eight telescopes at two sites) for about $3M, and we think we can operate it for about $0.5M per year. While this is a lot of money, it's quite cheap compared to the other surveys, and ATLAS has a unique mission that the other surveys cannot do nearly as well. Our best guess is that it will take us about 2 years to get everything built and running -- end of 2013 if we can start in early 2012. You can help too by making a tax-deductible donation * to the ATLAS Project.

We have writtten a proposal, 11-NEOO11-0017, to the NASA Near-Earth Object Observation (NEOO) program requesting funding for the ATLAS project, and right now we're waiting to hear from reviewers. (Last year's proposal, 10-NEOO10-0009, foundered because Congress did not provide NASA a budget.) The congressional mandate to discover NEOs is supposed to increase the NEOO program funding by a very substantial amount, so we are hopeful. We're proposing to build the eight telescope system, placing one set of telescopes on Haleakala and the other on the slopes of Mauna Kea, at the Hale Pohaku Visitor's Center. We could equally well deploy in California, at Mt. Palomar and Mt. Laguna, for example, or in Arizona on Kitt Peak, Mt. Hopkins, or Mt. Lemmon.

It has not escaped our attention that ATLAS is extremely modular, and that it would be relatively easy to build another unit, deploy it in Australia, and close the southern hole through which asteroids can approach unseen. Deploying more units in Hawaii and California and Arizona would help mitigate the blind spots from weather. And of course, distributing units in longitude would enable us to view the night sky 24/7. We are being careful to ensure that each unit has enough local computer power to operate by itself but has enough internet connectivity that we can network together as many units as we want. The initial ATLAS with the two sites for triangulation will be a proof of concept that we can effectively link multiple sites.

Other Cool Stuff

There's a lot of cool stuff that ATLAS can do besides watch for killer asteroids.

Look for denizens of the outer Solar System, such as dwarf planets like Pluto or Eris or a Nemesis star.
Detect gravitational lensing when nearby stars pass in front of distant ones.
Collect light curves of almost all of the variable stars in our Galaxy.
Detect thousands of Type Ia supernova explosions to a redshift of 0.1 within the visible sky.
See the flashes of light when a star is gobbled up by a super-massive black hole in a distant galaxy.
Nightly monitoring of the activity of 100,000 active galactic nuclei caused by black holes at the centers of galaxies.

Many more details about ATLAS, including all the other science that it's good for, can be found at astro-ph 1011.1028.

*This information is not intended, and should not be construed, as legal, tax or investment advice. You should always contact your tax advisor to help answer specific questions regarding charitable contributions and donations and the tax implications as they apply to you and/or your business.