Piercing the heart of a globular star cluster with its needle-sharp
vision, NASA’s Hubble Space Telescope has uncovered tantalizing clues
to what could potentially be a strange and unexpected population of wandering,
planet-sized objects.

In results published this week in NATURE, the international
science journal, Kailash Sahu (Space Telescope Science Institute, Baltimore,
MD) and colleagues report six unusual microlensing events inside the globular
cluster M22.

Microlensing occurs when a background star brightens momentarily
as a foreground object drifts by. The unusual objects thought to cause
these events are far too dim to be seen directly, but instead were detected
by the way their gravitational field amplifies light from a distant background
star in the huge central bulge of our galaxy. Microlensing has been used
before to search for low-mass objects in the disk and halo of our galaxy,
but Hubble’s sharp vision is essential to probe the interiors of globular
clusters further.

From February 22 to June 15, 1999, Sahu and colleagues monitored
83,000 stars, detecting one clear microlensing event caused by a normal
dwarf star in the cluster (about one tenth the mass of our Sun). As a
result of gravitational lensing, the background star appeared to grow
10 times brighter and then returned to its normal brightness over a period
of 18 days.

In addition to the microlensing event caused by the dwarf
star, Sahu and his team recorded six even more interesting, unexpectedly
brief events where a background star jumped in brightness by as much as
a factor of two for less than 20 hours before dropping back to normal
brightness. This means that the microlensing object must have been much
smaller than a normal star.

These microlensing events were unusually brief, indicating
that the mass of the intervening object could be as little as 80 times
that of Earth. Objects this small have never before been detected by microlensing
observations. If these results are confirmed by follow-up Hubble observations,
the bodies would be the smallest celestial objects ever seen that are
not orbiting any star.

So what are they? Theoretically they might be planets that
were gravitationally torn away from parent stars in the cluster. However,
they are estimated to make up as much as 10 percent of the cluster’s mass
— too numerous to be wandering, “orphaned” planets.

The results are so surprising, the astronomers caution that
these preliminary observations must be confirmed by follow-up Hubble observations.
If verified, these dark denizens could yield new insights about how stars
and planets formed in the early universe.

“Hubble’s excellent sharpness allowed us to make this remarkable
new type of observation, successfully demonstrating our ability to see
very small objects,” says Sahu. “This holds tremendous potential for further
searches for dark, low-mass objects.”

“Since we know that globular clusters like M22 are very
old, this result opens new and exciting opportunities for the discovery
and study of planet-like objects that formed in the early universe,” adds
co-investigator Nino Panagia (European Space Agency and Space Telescope
Science Institute).

“This initial observation shows that our microlensing method
works beautifully,” states co-investigator Mario Livio (Space Telescope
Science Institute).

As microlensing events are brief, unpredictable and rare,
astronomers improve their chances of observing one by looking at many
stars at once — much like a person buying several lottery tickets at
once. Most microlensing searches have been aimed at the central bulge
of our galaxy or out towards the Magellanic Clouds — the densest observable
regions of stars in the sky. In general these surveys cover areas of sky
larger than the full Moon and look for foreground objects lying somewhere
between us and the background population of stars.

Sahu and his team took advantage of Hubble’s superb resolution
and narrow field of view to aim the telescope directly through the center
of a globular star cluster lying between Earth and the galactic bulge.
This gave the team a very dense stellar region to probe for drifting low-mass
foreground objects and a very rich background field of stars to be lensed.
Only Hubble’s resolution is sharp enough to actually peer through the
crowded center of the cluster and see the far more distant stars in the
galactic bulge. As the lensing objects were part of the cluster, the astronomers
also had an accurate distance (8,500 light-years) and velocity for these
objects.

In a normal lensing event, a background star brightens and
dims for a length of time depending on the mass of the lensing body. The
short, “spurious” events seen by the team are shorter than the interval
between the Hubble observations, leading to an upper estimate for the
mass of an object of one quarter Jupiter’s mass.

To confirm these extraordinary, but tentative results, Sahu
and colleagues next plan to monitor the center of the globular cluster
continuously over a seven-day interval. They expect to detect 10 to 25
short-duration microlensing events, which will be well-sampled enough
to yield direct measurements of the true masses of the small bodies.

This release is issued jointly by NASA
and ESA.

CONTACT:

Ray Villard
Space Telescope Science Institute, Baltimore, MD
(Phone: 410-338-4514, E-mail: villard@stsci.edu)

Kailash Sahu
Space Telescope Science Institute, Baltimore, MD
(Phone: 410-338-4930; E-mail: ksahu@stsci.edu)

Nino Panagia
European Space Agency/Space Telescope Science Institute, Baltimore, MD
(Phone: 410-338- 4916; E-mail: panagia@stsci.edu)

Lars Lindberg Christensen
Hubble European Space Agency Information Center, Garching, Germany
(Phone: +49-(0)89-3200-6306; Cellular-24 hr: +49-(0)173-38-72-621;
E-mail: lars@eso.org)

Background Information

Gravitational Microlensing: Looking for Twinkles
in the Dark

More than 60 years ago, Albert Einstein calculated
that the gravity of a celestial body could act as a giant magnifying
glass by bending the light of a more distant object behind it.

But he dismissed the idea as a theoretical exercise,
saying there was “no hope of observing such a phenomenon directly,”
since the probability of observing such an effect within our Milky
Way Galaxy is generally less than one in a million.

What a difference a few decades make. With the advent
of powerful telescopes, scientists in the late 1980’s began capitalizing
on this phenomenon, known as gravitational microlensing. Astronomers
employ the microlensing detection technique to collect clues on
things they can’t observe directly, using this natural phenomenon
to hunt for a whole range of unseen celestial material and objects,
from dark matter, to extrasolar planets, to wandering stellar-mass
black holes.

Now, for the first time, a telescope has penetrated
the jam-packed core of a cluster of 10 million stars to search for
microlensing events. Astronomers used the Hubble Space Telescope
to search for unseen lightweight bodies – planets or “failed stars”
called brown dwarfs – in the core of the globular cluster M22 by
looking for their gravitational effects on the light from distant
stars behind them. (A brown dwarf is 80 times more massive than
Jupiter. It’s called a failed star because it doesn’t have enough
hydrogen in its core to shine as a star.)

Here’s how microlensing works: As an unseen body
floats across the face of a background star, it acts like a powerful
lens by gravitationally bending the starlight and thus creating
two separate images of the faraway star. Even Hubble can’t resolve
these images, because the bending angle is about 100 times smaller
than the telescope’s angular resolution. But the object’s gravity
also amplifies the starlight, causing it to brighten as the body
passes in front of the star.

In the Hubble telescope observation, most of the
background stars reside in our Milky Way’s central bulge. Hubble
monitored 83,000 stars every three days for nearly four months.
The thousands of stars near the cluster’s core are so tightly packed
together that only Hubble’s sharp eyes can resolve them. Monitoring
the stars in the central bulge, the orbiting observatory detected
six “mystery objects” in the cluster that eclipsed the light from
those background stars. In each case, a background star jumped in
brightness for less than 20 hours before dropping back to normal
brightness. These short eclipses mean that the objects must be much
smaller than a normal star, perhaps as small as 80 times Earth’s
mass. Objects this small have never been detected through microlensing
observations.

Astronomers determined that the mystery objects are
adrift in the cluster, meaning that they’re not orbiting stars.
Otherwise, these dark denizens would have been buried in the glow
of their parent stars and wouldn’t have been found.

Hubble also discovered another microlensing event
in which a dwarf star in the cluster transited the face of a background
star in 18 days.

Through these microlensing events, astronomers can
estimate a mass for an unseen body based on the duration of the
eclipse and the amount of brightening of background starlight. For
example, a brown dwarf 80 times heftier than Jupiter would transit
a background star in about 15 days; a Jupiter-mass object, about
1.5 days.