MDM 2.4m:

MDM 1.3m:

INT 2.5m:



Microlensing primer.

Microlensing is an exciting new technique for probing the mass distribution of our galaxy and nearby galaxies. The basic idea is very simple: the propagation of light from background stars is effected by the gravitational field of intervening objects regardless of their nature. The diagram illustrates the essential geometry.

As the lens passes close to the source-observer line of sight, the light from the source is focused causing an apparent brightening of the source star. The microlensing tube is defined as the region of space within one Einstein radius of the line-of-sight. Inside the microlensing tube the lens will produce a source amplification above 1.34. A simple microlensing event produces a very clear signature: a non-repeating, achromatic, symmetric and substantial enhancement of luminousity. Only two quantities are measurable for such an event, the duration, defined as the time the lens spends within the microlensing tube, and the maximum amplitude. The maximum amplitude is dependent only on the distance of closest approach and is thus void of interesting information. The duration on the other hand is dependent on the mass, location and velocity of the lenses.

As a further statistical check on the presence of microlensing the candidate events should be equally distributed in the color-magnitude diagram.

Classical microlensing.

Classical microlensing can be summed up: "stare at a few million stars and wait..." More specifically, the classical microlensing pipeline works as follows. One very high quality, or a series of similar images, is used as a template. Stellar photometry is performed on this template locating a catalogue of photometry centers and measuring their reference magnitudes. We will call these photometry centers "objects". Each timeseries image is then processed to derive magnitude estimates for each "object" in the catalogue. Thus a collection of object lightcurves is built up over time. These individual light curves are the basic items of interest in classical microlensing. Various cut are applied to the light curves leaving only a handful of candidates. The diagram below illustrates the observational strategy.

reference image--> "star" catalogue
time series image--> "star" catalogue
--> "star" catalogue
--> "star" photometry
--> "star" lightcurve
--> candidate list

Present experimental efforts.

From its humble beginnings, in less than a decade microlensing has grown to a well developed observational sub-field of astronomy. Many collaborations, from around the world, are presently working on various microlensing projects. These projects break down into two major areas. Some of the most notable of the first type, survey projects, are:

MACHOAustralian/American Collaboration- 1 m dedicated telescope on Mt. Stromlo, 6 years of data. One of original three collaborations.
EROSFrench collaboration - 1.5 m dedicated telscope at ... since...One of three original collaborations
OGLEPolish/American collaboration -- 1.5 m dedicated telescope at ... since...
MOANew Zealand/Japanese

The survey projects monitor large number of stars determine the optical depth and rates towards various targets, typically the Galactic bulge, the LMC and SMC.

When the survey projects observe a stellar brightening consistent with microlensing they issue an alert, notifying the comunity that a possible microlensing event is in progress. The second major type of collaboration picks up on these alerts and does detailed photometric, astrometric or spectroscopic observations. Among these follow-up collaborations are:


In addition to verifying the microlensing status of many of the events detected by the survey, a major goal of these follow-up collaborations is the detection of planets around the lens object. Even very small (earth-mass) planets can produce a detectable perturbation on the measured light curve. Another important goal of the follow efforts is the discovery of non-standard microlensing events such as those due to binary lenses or non-inertial motion of the earth. Careful study of these rare events allows one to partially break the distance-velocity-mass degeneracy and may lead to a better understanding of the distribution and nature of the lenses.

Status report

Microlensing experiments have been operating since 1992. Unfortunately, results have been slow in coming. Primarily this is due to the low event rate and the extremely large quantities of data (>6 terabytes to date for the MACHO Collaboration alone) that need to be analysed. Thus the most recent results (based on two years of data) can be summed up as follows: Analysed in the context of halo models this suggests that about half the halo is composed of objects with a mass of about half that of the Sun. The only feasible candidate population for these lenses would be white dwarves produced by an early population of stars. Unhappily this hypothesis sits very poorly with the bulk of our knowledge of early and present galactic conditions. For example the production of so many white dwarves is almost certain to wildly overproduce the heavy elements. To further complicate matters is is now clear that at least in the case of the SMC self-lensing is substantial. Of the two events detected to date towards the SMC, both are believed to be due to SMC lenses. Thus the location of the LMC lenses in the halo is thrown into doubt. Various suggestions have been made placing the lenses in the LMC disk, in a putative LMC halo, in the disk of our galaxy or even in an intervening galaxy. Suffice it to say that the location and nature of the lenses is still unknown and quite contentious.

Probing another halo

M31 microlensing:



surface brightness fluctuations

Difference Image Photometry

DIP image

Intro to mapping

map images


expectation figures

where are we?

Map of Andromeda showing fields

Image Subtraction from MEGA Dataset: Example

KPNO 4m MOSAIC image of the southern MEGA field showing M31 to the left and M31 as the bright spot right of center. The field of view is 36 arcmin square. The vertical streaks are saturated charge overflow from bright stars. The white-outlined square marks a 10 square arcmin region subtracted from a previous image of the same field, shown below. These data were taken in October 1998.

KPNO 4m MOSAIC subimage of M31 just south of the nucleus, subtracted from similar images taken in 1997. Note that most of the signal has disappeared, leaving only a few positive (white) and negative (black) stellar images due to variable stars. In two cases, we believe this residual stellar signals are due to microlensing events. A few stars are so bright as to cause saturated images e.g. in the lower right and upper left corner, and several more places throughout the image. Most of the pixels are filled with light levels near zero (and noise at the expected Poisson level), since most stars have "subtracted away". The same subimage before subtraction is shown below.

The same KPNO 4m MOSAIC subimage of M31 from 1998 shown above, but before subtraction.