A city seems like an enduring mark on Earth. Buildings, roads, bridges, and machines occupy a landscape on a scale that is difficult to overlook.
But 56 million years changes the question. Would a future researcher find recognizable streets, or would the clearest evidence be a chemical change preserved in sediment?
There is currently no evidence establishing that a pre-human industrial civilization existed. This episode examines what such evidence would have to look like, and why deep time makes the search more complicated than looking for ruins.
What the Silurian Hypothesis Actually Asks
Gavin A. Schmidt and Adam Frank introduced the question in their 2018 paper on the Silurian Hypothesis. They considered whether an industrial civilization's geological traces could be distinguished from natural events millions of years later.
The name is a science-fiction reference, not a claim about intelligent reptiles or a civilization in the Silurian geological period.
The paper is a thought experiment about detection. Its starting point is the known environmental footprint of human industry, rather than an excavated ancient factory.
Why Ruins Might Be the Wrong Target
The episode follows the processes that can make a built landscape unrecognizable. Exposed structures weather. Metals corrode. Sediment buries surfaces. Erosion removes material and carries it elsewhere.
At larger scales, tectonic processes deform and recycle parts of the geological record. Oceanic crust can eventually be carried into subduction zones. This does not mean that every part of Earth's crust is destroyed on one fixed schedule.
Burial can also preserve evidence. The result depends on where something lies, what covers it, and what happens afterward.
A city would occupy only a small part of the planet. Finding its surviving remains would require both preservation and a search in the right place. A widely distributed environmental signal could offer a different kind of target.
What Chemistry Could Remember
The video identifies several possible categories: changes in carbon and nitrogen isotopes, combustion particles, concentrated metals, synthetic materials, and artificial radioactive signatures.
These are candidates with different strengths and limitations. A material must first enter a deposit, survive alteration, and remain detectable in the sample eventually collected.
An isotope anomaly describes a change in the relative abundance of forms of an element. It does not identify the cause by itself. Likewise, a metal-rich layer could reflect several processes besides industry.
The useful question is whether a combination of observations would be difficult to explain naturally. Dating, geographical distribution, and the relationship between samples would matter as much as a striking laboratory result.
Why Mention 56 Million Years?
The Paleocene–Eocene Thermal Maximum, or PETM, occurred approximately 56 million years ago. It involved major warming and a large release of carbon into the atmosphere–ocean system, as described in research on its environmental consequences.
It offers a real geological comparison for a rapid carbon-cycle disturbance. It is not a discovered industrial episode.
The distinction matters because similarity is not identification. If two events share a carbon-isotope signal, that does not establish that they share a cause. A proposed industrial explanation would need evidence that distinguishes it from natural carbon release and climate change.
Would Plastics and Fallout Survive?
Synthetic materials sound like decisive clues because their manufacture can be closely associated with human activity. Their long-term preservation is a separate question.
An intact plastic object, a fragment, and a chemically altered residue are different kinds of evidence. Their survival cannot be assumed across every burial environment or over tens of millions of years.
Radioactive fallout provides another useful distinction. Research on fallout as a geological marker describes how nuclear-testing products became incorporated into sediments and other natural archives.
But radioactive markers have different half-lives and environmental behaviour. A signal useful for recognizing twentieth-century deposits does not automatically remain useful after 56 million years. The isotope and preservation conditions must be specified.
Why the Moon Is a Useful Comparison
Preservation also differs between worlds. NASA explains that lunar impact craters remain well preserved because the Moon has relatively little erosion and other surface activity compared with Earth.
That contrast helps explain why Earth's surface is an imperfect archive. It does not mean lunar objects are indestructible: impacts and other processes still alter the surface.
The comparison is about conditions of preservation, not evidence that another civilization left artifacts there.
What Would Count as a Strong Case?
A persuasive claim would need a documented deposit, secure dating, reproducible measurements, and evidence assessed against natural explanations and modern contamination.
Missing ruins alone cannot establish a vanished civilization. Neither can an unexplained chemical anomaly.
The lasting value of the question is methodological. It asks what our own industrial activity leaves behind, how those traces change, and what a future observer could reasonably infer from them.
What Earth preserves may be very different from what its inhabitants intended to make permanent.
Further Reading
- Schmidt and Frank (2018), International Journal of Astrobiology: The Silurian hypothesis: would it be possible to detect an industrial civilization in the geological record?. Open manuscript.
- Cundy, Waters, Hajdas, and Saito (2022): Radioactive Fallout as a Marker for the Anthropocene.
- Ivany et al. (2018), Science Advances: Little lasting impact of the Paleocene–Eocene Thermal Maximum on shallow marine molluscan faunas.
- NASA: Moon Craters.