Astrobiology and Technosignature Research with the Vera C. Rubin Observatory
Dr. Andjelka B. Kovačević on Multiscale Coherence with LSST for Astrobiology and Technosignatures for IAUS404
What if the signs of alien life and alien technologies don’t come to us as a single bright spectral line but as a geometric fingerprint woven across spectral scales and scales in time and space?
Rather than searching for one loud signal, what if the right approach is to ask whether the data “hangs together” in ways that random astrophysical noise simply cannot?
That question is at the heart of what Dr. Andjelka B. Kovačević presented during IAUS404: Advancing the Search for Technosignatures.
About the Presenter
Andjelka B. Kovačević is Associate Professor in the Department of Astronomy at the Faculty of Mathematics, University of Belgrade, and President of the National Committee for Astronomy in Serbia. Her research spans signal mining from red noise, active galactic nuclei (AGN), and the application of deep and machine learning to large-scale surveys: particularly the Vera C. Rubin Observatory’s Legacy Survey of Space and Time (LSST). She is also the first Serbian representative on the Board of Directors of the journal Astronomy & Astrophysics.
The poster was co-authored with Professor Nigel J. Mason of the University of Kent, a leading figure in astrochemistry and molecular physics, and currently President of the Europlanet Society. The third collaborator is Aleksandra Ćiprijanović, a Wilson Fellow Associate Scientist at Fermilab who leads the Cosmic AI group and is an active member of the Rubin Observatory/LSST Science Collaborations. Together, the trio brings together expertise in time-domain astronomy, astrobiology, and AI-driven data analysis.
More About Kovačević’s Work in Advancing the Search for Technosignatures
The Vera C. Rubin Observatory’s Legacy Survey of Space and Time will monitor billions of astronomical sources across six photometric bands, repeatedly, for a decade.
Every object it observes doesn’t just accumulate data—it traces a trajectory through a high-dimensional color-time space. Kovačević and her collaborators ask a direct and ambitious question: can organized physical processes, whether biological or technological in nature, reveal themselves as coherent departures from the natural manifold populated by ordinary astrophysical objects?
The conceptual engine behind their approach is multiscale coherence.
The logic is elegant: noise factorizes. When fluctuations are independent across wavelength bands and observational epochs, the data behave as though each measurement is unrelated to the others. But physical processes (photosynthesis modulating a surface, a Kuiper Belt Object’s coma scattering sunlight, a structured technological emitter, etc.) impose correlated structure across color, time, and population context simultaneously. Detection, in this framework, focuses on identifying structured dependence that cannot be explained by independent fluctuations.
To test this framework, the team built forward simulations probing three orthogonal axes of coherence:
Color coherence: Simulated Kuiper Belt Object (KBO) surfaces—including vegetation red-edge injections as surface-coherence proxies in exoplanetary spectra—produce aligned multiband displacements rather than gray dimming, detectable at roughly 5-sigma with LSST’s photometric precision.
Phase stability (temporal coherence): A surface with stable physical structure can generate repeatable light-curve patterns—a temporal fingerprint that persists across observational epochs.
Manifold distance: How far does a signal lie from the natural LSST color-time-population manifold? Different physics leaves different geometric footprints, and the distance from the baseline defines the anomaly.
A particularly striking result involves simulated exoplanetary reflected-light spectra, with 700 cases, orbiting M-type stars at 10–15 parsecs, with stellar spectral energy distributions included explicitly. When vegetation’s red-edge reflectance is injected as a smooth surface modulation, detectability shows threshold behavior: below roughly 20–30% surface coverage, no detection occurs. Above that threshold, detection probability rises sharply.
The time-domain analysis is equally striking. Simulating 1,000 LSST light curves across three regimes (pure noise, stochastic natural variability, and phase-locked coherent signals) reveals that no single metric separates the populations. Only when color behavior, phase stability, persistence, and amplitude stability are considered jointly do coherent signals emerge as a distinct cluster in the joint observable space.
Noise scatters incoherently; organized physical processes cluster.
Key Takeaways
The LSST data serves as a high-dimensional manifold and astrobiologically relevant signals may be detectable as geometric anomalies within it.
Noise factorizes; physical processes don’t. The key to detecting biosignatures or technosignatures in LSST data is measuring correlated structure across bands, epochs, and populations simultaneously.
Vegetation-like surface coverage on an exoplanet produces a threshold detection effect: LSST becomes sensitive above roughly 20–30% surface coverage, where the signal crosses a coherence boundary.
KBO composition (regolith-dominated, organic-rich, or ice-dominated) maps onto distinct, structured regions of LSST color space.
The team proposes establishing an LSST Astrobiology Science Team to coordinate community efforts and ensure the survey’s full potential for astrobiology is realized in the coming decade.
Thanks for joining us here with Planetary Perspectives, an editorial endeavor and science communication project from Blue Marble Space.
