Gravity-only test for a central IMBH (or compact dark mass) in Omega Centauri. As background stars drift behind the lens, their apparent positions shift. Gaia DR4 (≈ Dec 2026) and Roman (≈ 2028 onward) will either see it or rule it out.
When a compact mass passes near our line-of-sight to a background star, the star's light is gravitationally deflected. The star's apparent position shifts on the sky by a fraction of the Einstein radius — at peak alignment, δθ_max ≈ θ_E / (2√2). This is the astrometric signal. It is distinct from the photometric magnification (the brightness bump that OGLE, MOA, and KMTNet hunt for in bulge surveys). Astrometric shifts persist on time scales of years and survive even for "dark" lenses (isolated stellar-mass black holes, neutron stars, IMBHs) that produce no light themselves.
Gaia DR4 (≈ December 2026) releases the full astrometric time series for each Gaia source for the first time — not just the 5-parameter fit, but the per-epoch positions. Pipelines like Wyrzykowski et al. 2025 (A&A) have been waiting for this data specifically to look for transient astrometric microlensing events caused by isolated compact objects. ω Centauri's sightline crosses dense bulge regions, so every IMBH-mass body in OC has a non-negligible lensing cross-section against bulge stars. With a per-source astrometric noise floor of order ~30 μas at G ≲ 19, DR4 should detect — or strongly constrain — any IMBH that produces centroid shifts above this floor on year-to-decade timescales.
NASA's Roman Space Telescope (launch as early as fall 2026; nominal May 2027) carries a Galactic Bulge Time Domain Survey designed to find isolated stellar-mass black holes by astrometric microlensing across millions of bulge stars — see Penny et al. 2019 (ApJS 241:3). ω Cen sits in front of one of Roman's calibration fields. Sensitivity reaches ~10 μas, roughly 3× better than Gaia. Roman is therefore the medium-term test: if Gaia DR4 doesn't detect an IMBH signature in OC but the mass is real, Roman should still catch the smaller-shift, fainter-source events that Gaia misses.
A single IMBH produces rare but large-amplitude events: one mass × big θ_E × moderate event rate. A dark cluster of many smaller masses (stellar-mass BHs, neutron stars) produces frequent but smaller-amplitude events: many masses × small θ_E × high event rate. The two scenarios give numerically distinct distributions of (δθ_max, t_E) across detected events — which Gaia DR4 and Roman together will be able to discriminate. Compare to the alternative-hypothesis tool to see the signature difference.
Compare with Dark Cluster Alternative, IMBH Constraint Stacker, Pulsar Timing, and JWST Accretion. Astrometric microlensing is the only test in this set that depends solely on gravity — no assumptions about gas, radiative efficiency, or pulsar populations enter.
The event rate estimate Γ ≈ 2 θ_E σ_bg μ_rel is a back-of-envelope cross-section calculation; it assumes a uniform background density and ignores blending, finite-source effects, parallax, and binary-lens complications. The "detectable" verdicts use simple step-function thresholds (30 μas for Gaia, 10 μas for Roman) — real pipelines use full likelihood fits and can squeeze 2–3× more out of marginal events. The Einstein radius formula assumes a point lens; an IMBH is effectively a point at these scales, but a dark cluster is not.
Gaia DR4 release schedule — ESA/DPAC, planned ≈ December 2026 (cosmos.esa.int/gaia/release). Roman launch — NASA, fall 2026 – May 2027 (roman.gsfc.nasa.gov). Penny et al. 2019 (ApJS 241:3) — Roman microlensing survey simulations. Wyrzykowski et al. 2025 (A&A) — astrometric-only microlensing pipeline targeting DR4. Dominik & Sahu 2000 (ApJ 534:213) — astrometric microlensing of dark compact objects, including the δθ(u) = u θ_E / (u² + 2) result used here. Paczyński 1986 (ApJ 304:1) — original photometric microlensing paper; the A(u) = (u² + 2) / (u √(u² + 4)) magnification formula used in this tool.