Gravitational microlensing is a form of gravitational lensing in which an intervening object bends light from a more distant source, producing multiple images that are usually too closely spaced to distinguish individually. It is detected chiefly through changes in the source’s apparent brightness or position. The lens may be a star, a planet, a stellar remnant, or another compact concentration of mass, and it need not emit detectable light. Microlensing is used to discover exoplanets, measure the masses of otherwise faint objects, and investigate the distribution of compact matter.
Physical basis
According to general relativity, mass curves spacetime, changing the paths followed by light. For a light ray passing an isolated point mass at an impact parameter , the weak-field deflection angle is approximately
where is the gravitational constant and is the speed of light.
If a foreground lens lies sufficiently close to the line of sight to a background source, light can reach the observer along different paths. An isolated point lens produces two images of an off-axis point source. With perfect alignment, an ideal point source forms an Einstein ring.
For a typical stellar lens within the Milky Way, the image separation is of order milliarcseconds, too small for ordinary imaging observations to resolve. The combined images nevertheless contain more light than the unlensed source would contribute along the same observed direction. Gravitational lensing preserves surface brightness in geometric optics; the increased total brightness results from an increased apparent image area, not an intrinsic increase in the source’s luminosity.
The term microlensing describes an observational regime rather than a sharply defined mass range. Stellar microlensing also occurs within the larger-scale lensing patterns produced by a foreground galaxy, where it can alter the brightness of individual images of a background quasar.
Characteristic scales and the standard light curve
Einstein radius
The natural angular scale of a point-mass lens is the angular Einstein radius,
where , , and are the observer–lens, observer–source, and lens–source angular-diameter distances. For lenses and sources within the Galaxy, ordinary geometric distances are normally adequate, with .
The physical Einstein radius in the lens plane is
For a stellar-mass lens at Galactic distances, this radius is commonly comparable to a few astronomical units. This scale helps explain why microlensing is sensitive to planets with projected separations of a few astronomical units from their host stars.
Point-source, point-lens events
For an isolated point lens and a point source, the total magnification is
where is the angular lens–source separation in units of .
Assuming constant relative motion along a straight line,
where is the minimum separation, is the time of closest approach, and is the Einstein crossing time:
Here is the relative angular proper motion of lens and source.
This model produces a smooth, symmetric brightening and fading, often called a Paczyński light curve. At , the magnification is approximately . The ideal point-source magnification diverges at exact alignment, but a real source’s finite angular size prevents an infinite observed brightness.
The crossing time is not a direct mass measurement. It also depends on the distances and relative motion. Stellar-lens events often last weeks to months, while lower-mass lenses can produce much shorter events, but duration alone does not uniquely identify the lens.
Departures from the simple model
Binary lenses and planetary signals
A lens containing two masses produces a more complicated magnification pattern. Such systems include binary stars and a star accompanied by a planet.
An important feature is a caustic: a curve in the source plane at which the ideal point-source magnification becomes formally infinite. A finite source crossing or passing near a caustic produces sharp or otherwise distinctive features in its light curve.
A planet generally creates a short-lived deviation superimposed on the longer event caused by its host star. The deviation may be an increase or a decrease relative to a single-lens model. Depending on the planet’s mass and the trajectory, it may last hours to days.
A well-constrained planetary model commonly yields:
- The planet-to-host mass ratio, .
- The projected planet–host separation in units of the Einstein radius.
- The source trajectory relative to the lens system.
These measurements do not necessarily determine the planet’s mass in physical units or its full orbit. Different configurations can sometimes produce nearly indistinguishable light curves, including the familiar close–wide degeneracy.
Finite-source effects
When the angular size of the source is significant relative to features in the magnification pattern, different parts of its surface experience different magnifications. These finite-source effects smooth sharp features and can reveal the ratio
where is the source’s angular radius. An independent estimate of , often derived from the source’s color and brightness, then gives .
Differential magnification across the source can also probe its surface-brightness profile, including limb darkening.
Parallax and orbital motion
The standard straight-line trajectory can be altered by the orbital motion of Earth, by observations from widely separated locations, or by motion within a binary lens or source. Microlensing parallax can therefore distort a light curve or cause different observers to measure different light curves.
Lens orbital motion, source companions, and parallax can produce partially similar signals. Distinguishing them requires adequate temporal coverage and careful model comparison.
Astrometric microlensing and mass measurements
Microlensing changes not only brightness but also the apparent position of a source. Even when the two images cannot be resolved, their combined center of light shifts as the lens moves past.
For a dark point lens and an unblended point source, the angular centroid shift relative to the unlensed source is
The shift reaches its maximum, , at . Light from the lens or nearby stars modifies the measured centroid.
This astrometric signal can provide the angular Einstein radius. Combined with microlensing parallax, it can yield an individual lens mass. Defining the relative parallax as
for Galactic geometry, and the scalar microlensing parallax as
the lens mass is
Angles must be expressed in consistent units. Additional information may come from detecting the lens’s own light or resolving the lens and source after they separate.
These methods are particularly valuable for studying isolated white dwarfs, neutron stars, and black holes, whose masses can otherwise be difficult to measure when they lack observable companions.
Observational methods
Microlensing surveys monitor large numbers of stars repeatedly, especially toward crowded fields in the Galactic bulge and the Magellanic Clouds. The probability that any particular source is strongly magnified at a given time is small, so wide-field monitoring is essential.
Major survey programs have included the Optical Gravitational Lensing Experiment, Microlensing Observations in Astrophysics, and the Korea Microlensing Telescope Network. Automated systems identify candidate events, while frequent observations and coordinated follow-up help capture brief planetary deviations.
In crowded fields, a common technique is difference image analysis, in which suitably matched images are subtracted to isolate changes in brightness. Unresolved neighboring stars and lens light must still be accounted for. A simple photometric model is
where is the unlensed source flux and is blended light.
In the geometric-optics, point-source approximation, gravitational magnification is independent of wavelength. Observed color changes can nevertheless arise from blending, finite-source effects, or wavelength-dependent source structure. Achromaticity is therefore a useful diagnostic, not an absolute observational requirement.
Scientific applications
Exoplanet populations
Microlensing detects planets through gravity rather than through their own emitted light or periodic changes in their host star. It can probe distant planetary systems and is especially sensitive to planets near the lens’s Einstein radius, often at relatively cool orbital separations.
Short events without an obvious host-star signature can reveal planetary-mass lens candidates. However, an undetected host may be very widely separated, so distinguishing a genuinely unbound planet from a bound planet requires additional evidence or statistical modeling.
Microlensing surveys can estimate planet occurrence rates, but the detected sample must be corrected for selection effects. Detection efficiency depends on mass ratio, projected separation, source brightness, observing cadence, weather, and event geometry.
Compact objects and dark matter
Microlensing searches were developed partly to test whether compact objects could account for Galactic dark matter. Surveys toward the Magellanic Clouds and other fields have placed strong limits on the contribution of compact lenses over substantial mass ranges.
These constraints are mass-dependent and rely on assumptions about the spatial and velocity distributions of lenses and sources. They do not exclude every possible compact-object population or every proposed primordial black hole mass range. Lensing by ordinary stars within the monitored systems must also be distinguished from lensing by objects in the Galactic halo.
The microlensing optical depth—the probability that a source lies within an Einstein radius of a lens along the line of sight—is also used to study the distribution of ordinary matter in the Galaxy.
Distant sources
In strongly lensed quasars, stars in the foreground lens galaxy can magnify different emitting regions by different amounts. Smaller regions generally show stronger microlensing variability. Comparing changes across wavelength bands helps constrain the sizes and structure of accretion disks and other quasar-emitting regions.
Near the critical curves of galaxy-cluster lenses, microlensing can also produce large brightness changes in individual, highly magnified distant stars.
Historical development
In 1936, Albert Einstein published a calculation of the lensing action of a star, while judging the observational prospects to be poor. In 1986, Bohdan Paczyński proposed monitoring large stellar populations to search for compact dark objects through microlensing.
In 1993, the MACHO, EROS, and OGLE collaborations reported candidate microlensing events, establishing large-scale stellar monitoring as a practical detection method. The first widely accepted exoplanet discovery by microlensing, OGLE-2003-BLG-235/MOA-2003-BLG-53, was announced in 2004.
Limitations and interpretation
Most Galactic microlensing events are one-time alignments rather than periodically repeating signals. A missed planetary deviation usually cannot be recovered in a later cycle, although subsequent imaging can still constrain the lens.
The principal limitation is parameter degeneracy. A simple light curve supplies a crossing time and trajectory parameters, but not a unique lens mass, distance, or transverse velocity. Binary lenses, binary sources, parallax, and orbital motion introduce further ambiguities.
Without additional measurements, inferred physical properties often depend on statistical models of Galactic populations. Reliable population studies therefore require both explicit detection-efficiency calculations and propagation of measurement uncertainty, rather than treating every event as a uniquely characterized planetary system or compact object.