Chromatic microlensing means that the measured color changes during a gravitational magnification event. In the usual geometric-optics description, gravity is not separating colors like a prism. The color signal comes from the source’s wavelength-dependent structure, or from light mixed into the measurement.
Separate the lens from the measured image
An isolated point source, magnified achromatically, retains its intrinsic flux ratio between filters. A real stellar disk can have different brightness profiles in different filters. A lens near that disk then weights its center and limb differently over time, changing the integrated color.
There is a second route: an unresolved neighbor can supply a different-colored background. The lensed star brightens while that neighboring contribution stays approximately constant. The combined image changes color even when the source magnification is identical in both filters.
A two-filter experiment that exposes blending
Hypothetical flux units, not measurements: suppose the source contributes 10 units in a blue filter and 10 in a red filter. A red neighbor adds 2 blue units and 10 red units.
| State | Blue total | Red total | Blue/red ratio |
|---|---|---|---|
| Before magnification | 10 + 2 = 12 | 10 + 10 = 20 | 0.60 |
| Source magnified by 5 | 50 + 2 = 52 | 50 + 10 = 60 | 0.87 |
The image becomes relatively bluer although the source was magnified by exactly five in both bands. Flux ratios are used here for clarity; astronomical color indices are logarithmic and also depend on the filters’ zero points.
The useful model is Fobserved,band(t) = Aband(t) × Fsource,band + Fblend,band. A changing total color does not, by itself, establish that A differs by band.
When the star’s surface is the signal
During a close passage or caustic crossing, narrow regions of an extended star can receive very different amplification. Wavelength-dependent limb darkening can then imprint a structured color curve. The timing and shape matter, not just the largest color difference.
Astronomers fit the photometry in several filters together, allowing for source size, surface brightness, blended flux, and the lens trajectory. The study of microlensing chromaticity explicitly compares limb darkening with blending. Later work on stellar color variations investigates additional source-related complications.
Why quasars require a different checklist
In a strongly lensed quasar, foreground stars can preferentially magnify compact emission regions. Different wavelengths can originate from different-sized regions of an accretion disk. That makes wavelength-dependent magnification useful for testing source-size models.
But multiple quasar images also have different light-travel delays, and dust extinction can change their colors. Comparing images taken at the same observing time does not necessarily compare the quasar at the same intrinsic moment. A stellar-event blending calculation and a quasar accretion-disk inference therefore should not be treated as the same analysis.
Four checks before interpreting a color curve
- Were filter observations sufficiently close in time for the event’s rate of change?
- Has unresolved blended light been estimated in each band?
- Does the signal coincide with a source-crossing feature predicted by the lens model?
- Could calibration, extinction, or intrinsic variability explain the pattern?
A color change is an additional constraint, not an automatic measurement of a stellar atmosphere. For the geometric background, read finite-source microlensing. The distinction to keep is simple: an achromatic physical deflection can still produce a chromatic measurement.
Revised October 26, 2026; the flux experiment is illustrative.