Our Research

Gravitational Microlensing

The galactic microlensing group at Jodrell Bank is one of the most active in Europe and undertakes both theoretical work as well as analysis of microlensing datasets. We have close links both with the main microlensing survey groups (OGLE and MOA), as well as with follow-up groups who are using the technique to look for extra-solar planets. Staff here have pioneered both the theory behind extra-solar planet detection with microlensing, as well as the application of the technique beyond the Milky Way. Read on for a brief overview of microlensing as well as a summary of projects being undertaken here.

What is microlensing?

Microlensing event in POINT-AGAPE data A microlensing event shown in false colour. The event is the red flash above the bright foreground star. It was observed in the Andromeda Galaxy in 1999 by the POINT-AGAPE dark matter microlensing survey. Click on the image for a movie of the event.

Gravitational lensing describes the deflection and distortion of light by intervening mass distributions. The size of the effect depends upon the geometry of the observer, lensing mass and background source, as well as on the mass, size and shape of the lensing object. On large distance and mass scales, foreground clusters of galaxies can produce multiple distorted images of background galaxies, sometimes with spectacular results that are akin to a "hall of mirrors" effect. Many of these lenses are studied by the Gravitational Lensing Group here at Jodrell. On much smaller scales, such as within galaxies, gravitational lensing signatures can be caused by the close passage of stars or planets across the line of sight to more distant stars. In these cases multiple images are also produced but are typically separated by around a milli-arcsecond and so are too close together on the sky to be resolved. Instead, the combined brightness of the images is observed to vary with time in a very precise way which is wavelength independent. This regime of transient, unresolved gravitational lensing is known as microlensing.

Microlensing within galaxies is an extremely rare phenomenon. Even towards the most crowded stellar fields in the bulge of the Milky Way there are typically one or two detectable microlensing events ongoing at any given time for every million stars observed. This led Einstein largely to dismiss the possibility of being able to detect microlensing events when he wrote about the subject in 1936. Worse still, there are approximately 1000 intrinsically variable stars for every detectable microlensing event. Fortunately, the precise transient behaviour of microlensing events allows them to be reliably distinguished from variable stars, and advances in detector technology and computing power allow today's astronomers to routinely monitor and analyse tens of millions of stars each night. To date the efforts of several microlensing surveys of the Milky Way, such as EROS, MACHO, MOA and OGLE have yielded several thousand microlensing events, almost all towards the Galactic bulge. Members of the Jodrell Microlensing Group have been involved in both of the currently ongoing main microlensing surveys, OGLE and MOA, which are responsible for the large majority of events detected to date. Einstein would probably have been surprised to learn that, seventy years on, microlensing is by far the most commonly observed form of gravitational lensing.

Microlensing samples have been used to measure the density of compact (Macho) dark matter candidates in the halo of the Milky Way and the Andromeda Galaxy, to probe the three-dimensional structure of the Galactic bulge, to identify black hole candidates, to resolve the photospheres of distant stars, and to detect extrasolar planets. Astronomers at Jodrell have been and continue to be at the forefront of these efforts.

Microlensing at Jodrell

Extra-solar planets

Artists impression of the 5 Earth mass planet
An artists impression of the 5-Earth mass planet discovered recently through the microlensing technique.

(Mao & Rattenbury)

In 1991 a seminal paper by Mao & Paczynski explored the possibility of detecting planets using the microlensing effect. The idea they put forward was that the presence of a planet orbiting the lensing star could, under a favourable alignment, induce a detectable perturbation on the microlensing lightcurve. This theoretical possibility became an observational reality 13 years later with the first detection of a microlensing planetary system.

Recent discoveries of low-mass planetary companions using microlensing have been achieved due to real-time alerts issued by the OGLE and MOA survey teams. The alerts are monitored intensively by follow-up networks such as PLANET, RoboNET and MicroFUN. Jodrell astronomers are members of both the main survey teams as well as the RoboNET robotic follow-up network. Results from these surveys indicate that planets in the Earth to Neptune mass range may be much more common than predicted by theoretical models of planet formation. Microlensing is currently the most sensitive technique for the discovery of such planets, having recently detected a planet with a mass around 5 Earth masses.

HST follow-up studies of microlensing events

Image of resolved microlensing event
HST images taken 3.7 years (left) and 8.9 years (right) after an observed microlensing event. The lens and source are clearly resolved in the later image (components A and B).

(Kozlowski, Mao & Wood)

Whilst the microlensing effect can produce strong changes in the brightness of the background source, the foreground lensing star need not be visible at all. Nonetheless some events should, in principle, involve detectable lensing stars. However, the difficulty in identifying the lensing star is that its close proximity on the sky to the source star during the event prevents it from being directly resolved. Nonetheless, microlensing surveys have been monitoring the brightness of millions of stars in the Galactic bulge since the early 1990s and so differences in proper motion should allow the stars to be resolved after several years.

Astronomers at Jodrell have recently used the Hubble Space Telescope (HST) to directly identify the lensing star involved in a microlensing event which occurred nine years previously. This is the first time a microlensing star has been directly identified. The importance of this identification is that it allows the nature of the lens to be established. Furthermore, it permits a consistency check on the microlensing hypothesis by back-tracking the observed lens and source trajectories and checking for consistency with the lensing geometry inferred from the original event data.

Galactic structure from microlensing datasets

A simulated map of optical depth
A simulation of the predicted distribution of optical microlensing events across the Galactic bulge. This is the first such simulation to incorporate a realistic 3D dust model.

(Kerins, Mao & Rattenbury)

Microlensing is sensitive to the line-of-sight geometry of stars as well as their transverse kinematics. This make microlensing datasets potentially powerful probes of Galactic structure.

At Jodrell we are using microlensing datasets to probe the 3D structure of the inner Galaxy. Using state-of-the-art synthetic stellar population models we are constructing detailed theoretical maps of the distribution of microlensing events. Our latest maps include 3D models for interstellar absorption.

Microlensing surveys also provide accurate photometry for large samples of standard candle stars such as red clump giants. Their magnitude distribution has been used recently to constrain the orientation and physical parameters of the Galactic bar.

At Jodrell we are also investigating the possibility of infrared microlensing surveys. To date all microlensing surveys have been conducted at optical wavelengths. Unfortunately our location within the disk plane means that large areas of the Galactic bulge are obscured at visible wavelengths by intervening dust. This means that microlensing surveys have hitherto been confined to regions of the bulge where the dust column density is relatively small. Theoretical models developed at Jodrell show that microlensing surveys can trace the underlying Galactic structure much better at near-infrared wavelengths, such as in the K band (centred at 2.2μm).

Using K-band data obtained from the WFCAM array on UKIRT, currently the World's largest infrared telescope, we have demonstrated that near-infrared monitoring provides a powerful step forward for galactic structure studies with microlensing. The next step forward is a major large-scale near-infrared variability survey of the Galactic bulge and disk, dubbed VVV, which is to be undertaken from 2009 on the new VISTA infrared telescope in Chile. VVV will be able to detect microlensing right across the bulge. Microlensing samples from VVV will trace the bulge structure far better than can be achieved with current optical surveys.

Extra-galactic microlensing

The Andromeda Galaxy (M31)
Our birds-eye view of the Andromeda Galaxy, our nearest giant galactic neighbour, makes it a good choice to look for microlensing events. Its triaxial bulge is the focus for the Jodrell-led Angstrom Project (NOAO image).


Microlensing events are also detectable in galaxies other than our own, however the challenges of detection are more difficult. In general the source stars to extragalactic microlensing events are unresolved in the absence of lensing. As a result we see only the peak of the microlensing event when it is highly magnified above the galactic background starlight. In order to obtain stable photometry of the event, image enhancement techniques must be employed to correct for changing atmospheric conditions. This kind of unresolved microlensing is sometimes referred to as pixel lensing.

The Angstrom Project is a Jodrell-led microlensing survey of the bulge region of the Andromeda Galaxy (M31). This unique international collaborative project employs a distributed network of five 2-metre-class telescopes to monitor the M31 bulge 24 hours per day. The intensive monitoring is necessary to see the relatively short-lived peaks of ongoing microlensing events. Two of the telescopes in the network are fully robotic and the data from them are analysed by a fully-automated data reduction pipeline. The Angstrom Project Alert System analyzes tens of thousands of lightcurves daily to look for signals of ongoing microlensing candidates or other transient objects such as classical novae. The Angstrom Project is the first survey ever to detect microlensing events without any human intervention from the point of observation to the point at which a candidate is flagged up. It is also the first extragalactic microlensing survey to employ real-time data reduction. We aim to use the spatial and timescale distribution of M31 microlensing events to probe the 3D structure of the triaxial bulge in M31. Using real-time data reduction we also intend to issue alerts of ongoing events. Our aim is that some of the exciting science being performed by real-time microlensing discovery within our own Galaxy will also be possible with extragalactic microlensing systems, possibly including the discovery of extra-galactic planetary systems.