Computer simulations of galactic expansion argue that long-ago visits might leave no trace.
The idea that Earth could have been visited or even settled by an extraterrestrial civilization long before humans evolved has a narrow scientific footing: not in claims of artifacts, but in how models of interstellar expansion behave when civilizations have limited lifetimes and move in fits and starts.
In a paper posted to the preprint server arXiv, a team led by Jonathan Carroll-Nellenback and including Adam Frank and Jason Wright uses computer simulations to revisit the Fermi Paradox, the long-running tension between the expectation that technological life could be common and the absence of clear evidence that anyone has reached Earth.
Their central claim is not that ancient visitors existed, but that if they did, the odds of modern researchers finding persuasive evidence may be low. Over deep time, Earth is a planet that edits its own surface.
A “settlement model,” not a sweeping conquest
The researchers frame their work as a “settlement” problem. Instead of assuming that a spacefaring civilization expands in one continuous wave that fills the Milky Way quickly and permanently, their simulations allow expansion to proceed unevenly. Settlements can arise, persist for a time, and then disappear. New settlements may form later elsewhere. Stellar motions also matter, because stars are not fixed points on a map, a detail the paper treats as important for how a settlement front diffuses through the galaxy.
That approach produces outcomes in which a galaxy can be broadly “inhabited” in the sense that settlements exist somewhere, while many individual star systems remain empty at any given moment. It is a way to loosen a common assumption embedded in many versions of the paradox: if anyone ever learned to travel between stars, then every suitable system should eventually be occupied, including ours.
“You can end up with this loose network of settlements,” explains co-author Jason Wright, “where the whole galaxy is settled, but any given star at any given moment might not be.”
In the paper’s framing, Earth could fall into any of several categories: never visited, visited but not settled, settled for a time and then abandoned, or settled in ways that left little durable signature. The modeling does not identify which of those applies. It argues that several are consistent with the same basic facts: a galaxy old enough for many opportunities, and a present-day Earth that does not show unambiguous evidence of a nonhuman technological past.
The “finite horizon” problem on a changing planet
A second piece of the argument is geologic. The team points to what it calls a “finite horizon” for evidence, meaning a limit on how far back physical traces on Earth’s surface can reliably persist.
The basic mechanism is familiar to geoscientists: erosion, burial, volcanism, impacts, and, most importantly, plate tectonics. Oceanic crust is continually created and destroyed. The U.S. Geological Survey notes that the oldest parts of today’s oceanic crust are on the order of a couple hundred million years old, evidence of ongoing recycling at subduction zones and mid-ocean ridges (USGS “This Dynamic Earth”). Even when traces are preserved on continents, they are still subject to metamorphism, uplift, and erosion across spans that dwarf human history.
In that context, “no trace” is not a surprising outcome over tens or hundreds of millions of years. A hypothetical alien presence deep in Earth’s past could leave behind structures or materials that are later destroyed, dispersed, or buried beyond practical recovery. The paper’s point is probabilistic, not absolute: the longer the time interval, the less confident anyone should be that surface evidence would remain available to find.
That matters because popular versions of the Fermi Paradox often treat the present as decisive. If no one is here now, the reasoning goes, then no one ever arrived. The study argues that this link is weaker than it appears once the model includes finite civilization lifetimes and Earth’s finite evidence horizon.
The team formalizes that by combining a steady-state “settled fraction” of star systems with the time window in which evidence could survive. In that combined picture, a galaxy can host settlements over long spans while still allowing long quiet intervals for specific systems, including Earth.
A third outcome for the Fermi question
The authors present their work as a “third answer” to the paradox. The first common answer is that technological civilizations never arose elsewhere. The second is that they are common and should be visible or present. The third is that civilizations can arise and spread, but in fragmented waves that do not guarantee that any specific world is occupied now, or that evidence of past occupation would remain.
Their simulations highlight several outcomes that are consistent with silence today. In some parameter choices, civilizations do not last long enough to colonize widely. In others, they spread, but unevenly, leaving gaps. In still others, a settlement front can fill in regions of the galaxy and later thin out, with settled systems becoming uninhabited as civilizations end and are not immediately replaced.
In this framing, the absence of contact does not require an ongoing visitor. A world can be empty for geological stretches and still sit in a galaxy that has hosted interstellar settlement at other times and places.
The paper also draws on the “Aurora Effect” idea embedded in its title, a way of describing a galaxy that can be broadly settled without being uniformly occupied at every moment. That matters because many popular intuitions about colonization assume permanence. The model does not. It assumes that settlements, like species and institutions on Earth, can end.
What this does and does not claim
The study does not report observations of artifacts, signals, or biological signatures. Its contribution is a modeling argument about plausibility under conservative assumptions: finite travel speeds, finite ranges, and finite lifetimes.
It also does not claim that Earth was in fact colonized. It argues that the lack of clear evidence today is not, by itself, decisive against ancient visitation. That is a narrower claim, and it rests on two ideas that are easy to state but hard to escape: a galaxy can be settled without constant, universal occupation, and Earth has a limited ability to preserve surface traces over deep time.
The most speculative element in the broader claim is the suggestion that any traces might already exist but be misinterpreted or overlooked. The paper itself leans more heavily on the opposite: that long intervals make the survival and recognition of traces less likely.
Still, the scenario the authors outline is internally consistent with known constraints. Civilizations could be transient. Settlement could be patchy. Geological recycling could erase the evidence. The outcome is a version of the Fermi Paradox that does not require anyone to be nearby, talking to us, or leaving fresh, unmistakable ruins.
Because if alien civilizations may have colonized Earth before humans ever existed, then the real question isn’t just where are they now? — it’s why did they leave? And perhaps even more haunting: what did they leave behind?
