Snowball, Slushball, Desert, Greenhouse…

Thought Experiments on the Extinction of Civilization

What can we learn about past civilizations by surveying them from orbit?


The following thought experiments are predicated upon the effect of major climate change events upon low technology civilizations, as well as being predicated upon these climate events becoming so severe that they eventually result in the extinction of a civilization or civilizations. Several obvious variations on the theme of civilizational collapse through climate change would include thought experiments about early civilizations driven extinct by snowball and slushball events, by desertification and aridification, and by greenhouse events. Please note that I am not here concerned with civilizations in possession of technology more-or-less equivalent or better than our own, but only with civilizations that are primarily agricultural and pastoral in character. However, in terms of searching for the artifacts of such civilizations I will assume technology levels at or above our current technological accomplishment.

Pieter Bruegel the Elder, “Winter Landscape with Skaters and Bird Trap” (1565)

Snowball and Slushball

I have written several posts on the possibility of very old civilizations that stagnate at a level of development at or below that of contemporary terrestrial civilization (cf. Stagnant Supercivilizations). It is entirely possible that, if the breakthrough to the industrial revolution made in western civilization had not occurred when and where it did, that breakthrough might not have occurred at all, and civilization on Earth might have remained below the level of industrialization, either stagnating or caught in a high level equilibrium trap — an agricultural civilization from start to finish.

If such a civilization had gotten its start during a protracted period of climatological stability for Earth, like the “boring billion,” then it might have been possible for even a rudimentary agricultural civilization to endure for millions or years, or even billions of years. But human civilization got its start during an interglacial warming period (what I have called the “Holocene window”), likely to end at some time, and sooner rather than later. And the Quaternary glaciation of which the Holocene warming is a part, involves significant climate fluctuation over periods of thousands, tens of thousands, and hundreds of thousands of years.

In addition to the possibility of an agricultural civilization that emerges in a protracted period of climatological stability, and so may grow to a great age without great change, there is the possibility of an agricultural civilization that emerges during a period of glaciations (like our own), but which fails to make the breakthrough to industrialization and so cannot look to technological solutions to climatological change. Once the warming period that had incubated the civilization had passed, that civilization would have to retreat toward the equator and crowd into the remaining ice-free zones. This is a mirror image of the existential risk scenario that we may be facing if the planet heats up, and human populations abandon the equator and move to the cooler poles.

If the end of a civilization-incubating warm period were followed by a “snowball” event, in which the planet entire were covered in ice, that would mean the inevitable extinction of any but the most advanced technological civilizations that could leave the planet, or engage in large-scale climate engineering, like using mirrors in space to focus more solar insolation on the planet.

How difficult would it be to find the remains of a civilization that had been extinguished by a catastrophic snowball event? If the snowball event meant the entire planet encased in ice, probably many kilometers thick, the physical difficulty of any investigation would be considerable. It would be necessary to drill down through the ice to sample the layers below where once a civilization might have existed. What little evidence might be derived from ice cores would be fragmentary in the extreme. But who would even think to look?

If you arrived to investigate a planet and found it entirely iced over, in the grip of a snowball event, why would you even suppose that anything was under the ice? According to the assumptions of our thought experiment above, a pre-technological civilization would be driven extinct by a snowball event because it would not possess the technological wherewithal to do something about runaway glaciation. Such a pre-technological civilization would leave only very subtle signs of its past existence.

There might be some faint traces of atmospheric gases that could not be entirely accounted for by natural processes, but this would be a very subtle technosignature, easily overlooked. With a technological civilization (even if long defunct), the technosignatures would be more easily detectable and less ambiguous to interpret. The higher the technology, the more exotic the materials produced by a civilization. And if a technological civilization attained spacefaring before going extinct, its remains could be found in orbit around the planet or on other worlds in the planetary system. This would be unmistakable evidence, but civilizations below the level of industrialization would not leave obvious technosignatures of this kind.

In the event of a less severe snowball event, what is sometimes called a “slushball,” where there is still a band of liquid water around the equator (perhaps with icebergs floating in it), some few, hardy representatives of the intelligent progenitor species of the now defunct civilization might maintain themselves in the condition of nomadic hunters. If you arrived for a planetary survey and found such an intelligent species, you might observe that the conditions on the planet at that time were not such as would allow an intelligent species of this kind to evolve, and so you might want to dig deeper into the history of the planet.

It would probably not be difficult to determine that a snowball or slushball planet formerly possessed a temperate biosphere, and that traces of this temperate biosphere would be detectable by scientific investigation. But what kind of traces would a civilization leave under the ice? A melt probe could go into the subsurface oceans, but the remains of a civilization would be on the former continental landmasses, which would be covered by an enormous weight of ice. The movement of the ice over these surfaces would scrape away any traces of a prior agricultural civilization. I’m not saying that such a civilization would leave no technosignatures whatsoever (there may be traces of lead in bubbles in the ice, as there are on Earth from the Roman lead industry), only that the technosignatures would be few and difficult to detect.

A scientific mission, then, might readily sample the ice and the subsurface oceans and find signs of the former temperate biosphere, but it is quite likely that they would find no traces of past civilization by these methods. And if they specifically were on a survey for civilizations, the prospects of finding anything might be so minimal that they wouldn’t bother to investigate (it might well depend on their schedule and their budget). I could imagine a survey team that makes a quick scan for radionuclides, and, finding none, moves on, as any civilization that hasn’t even developed the technology of nuclear fission may be unworthy of further investigation.

In the event of a glaciation that fell short of a slushball — say, something like the last glacial maximum — even a rudimentary civilization might be able to keep itself going near the equator, or wherever there remained arable land and sufficient rain and sunshine to grow crops. Here the evidence for civilization would be apparent, and the remains of the former planetary-scale civilization would not be so completely obliterated by the ice that a thorough archaeological survey could not reconstruct much of that now vanished civilization.

One of my all time favorite Chesley Bonestell paintings, of human explorers visiting a ruined Martian temple. Note the slenderness of the columns and the wide space between them, possible in the lower gravity of Mars.

Desertification and Aridification

A melancholy motif that was commonplace a hundred years ago, when Percival Lowell was looking at Mars through a telescope and thought that he saw canals crisscrossing the Martian surface, was the idea that there was an ancient civilization on Mars that was dying, and they were building canals to bring water from the polar caps to the equatorial regions. In other words, Martian civilization was under existential threat from desertification. The idea that a civilization might be destroyed by desertification, then, has been around for a while.

This idea of Martian civilization facing existential risk was further elaborated by C. S. Lewis in his Space Trilogy (comprising Out of the Silent Planet, Perelandra, and That Hideous Strength). Lewis wanted to respond to the ideas of Olaf Stapledon and J. B. S. Haldane, which he believed to be “demonic.” Lewis used the character of Dr. Weston to place the potential spacefaring expansion of humanity in the worst possible light. In the Space Trilogy, the Martians choose to die nobly and stoically with their dying planet, rather than to use the technology they possess in order to travel to another planet and to displace those inhabitants in favor of themselves.

Now, in the case of a multi-planetary ecosystem — on which cf. Emergent Complexity in Multi-Planetary Ecosystems — such a scenario might be possible, although I would observe that, even in the case of panspermia distributing life among closely spaced planets in a single habitable zone, these individual planets would still have radically different biospheres because of their separate evolutionary paths. In the case of worlds separated by interstellar distances, the biospheres would be so distinct, even if originating from a single origin of life event, that organisms evolving in any one biosphere would not likely be able to survive in any other distantly related biosphere.

Our own solar system came close to being one in which multiple inhabitable planets were to be found in the same habitable zone. If Venus had been a little farther out, or if Mars had been a little closer in, we could have found ourselves in the dilemma that C. S. Lewis contemplated. Indeed, Mars may have been inhabitable in the past. At some time in the distant past (and by “distant past” I mean billions of years ago) Mars probably had a liquid water ocean on its surface, but Mars got cold and dry, and the ocean dried up. Probably some of the water went underground, some went into the ice caps, and some into the atmosphere and was thence lost into space. Mars essentially freeze dried itself and is now a cold desert.

If Mars’ period of apparent habitability had endured for a longer period of time, life might have arisen, and civilization, and the kind of ruins depicted in the Chesley Bonestell painting above might now exist on Mars. But when Mars had an ocean between three and four billion years ago, life on Earth was only getting started, and was little more than pond scum. Mars would have had to have remained habitable for almost as long as Earth has been habitable in order for it to evolve forms of life sufficiently sophisticated to build a pseudo-Grecian post-and-lintel temple.

Deserts are great preservers. If there had been a civilization on Mars driven extinct by desertification, its artifacts would likely be exceptionally well preserved. Some of the oldest examples of textiles that we possess come from Peruvian mummies preserved by the desert (cf. Oldest Indigo-Dyed Fabric Ever Is Discovered in Peru by Stephanie Pappas), so we see that deserts can preserve artifacts other than durable goods and monumental architecture. Textiles, books, and even bodies are preserved by deserts — the Nag-Hammadi library was in part preserved due to exceptionally dry conditions, as were the Children of Llullaillaco.

The spacecraft that we have sent to the Martian desert will likely outlast anything that human beings have built on Earth. Our monumental architecture and sprawling cities will all be ground to dust by the slow and relentless movement of tectonic plates. On Mars where it is cold and dry and probably sterile, the only forces to wear away our landers and our rovers are the wind and the sand, which will certainly cover them over in the fullness of time, there to await eternity under a dune of red sand.

A desertified planet that once had a rudimentary civilization, driven extinct by aridification, would likely preserve much more of the record of that civilization than a civilization that succumbed to glaciation. Moreover, evidence of a civilization on a desertified planet would probably be relatively obvious. Many of the monuments of Egyptian civilization were toppled and covered over by sand, but the pyramids, the Sphinx, and the largest temples were always visible as ruins so that the memory of Egyptian civilization was never lost to historical memory.

Such a planet would essentially preserve the entirely of a civilization as though it were a vast tomb, and any visiting archaeologists would face an embarras de richesses in terms of scientific investigation. Moreover, since the dry conditions of the planet would continue to preserve all its specimens, there would be no need to rush forward with salvage archaeology. The temptation would be there to loot the best artifacts, and it would be necessary to set guards around the planet in order to deter looters, who would come for just such treasures, planning to sell them on the interstellar antiquities black market. Scientific investigation of the planet could well take centuries, especially if the civilization had endured for a long period of time before succumbing to its dying homeworld.

In so far as a dramatic climate change event could result in the kind of desertification here contemplated, it also would be entirely possible that, with a significant climate change event, parts of a planet would be buried under ice, while other parts of a planet experienced extreme aridification. On such a planet, a civilization could be subject both to being covered over with ice near the poles, while having large tracts of its arable land subject to aridification. A one-two punch like this would almost certainly end the career of an agricultural civilization (even if the planet remained habitable, and it did not self-sterilize, as did Mars) and send its remaining population back to hunter-gatherer nomadism. In other words, even a planetary-scale agricultural civilization (like terrestrial civilization up until the industrial revolution) lies below the dissolution threshold of planetary-scale climate change.

A dissolution threshold, then, is not merely a function of the extent and achievement of the civilization, but also a function of the magnitude of any existential risk that is implicated in the collapse of a civilization. I now realize that, implicit in this and the above thought experiment on civilizations that succumb to snowball and slushball climate events, is the idea that less technologically advanced civilizations have a lower dissolution threshold, while civilizations that have learned to investigate their environment scientifically, and which can harness technologies to change their environment, have a higher dissolution threshold.

In the far future, Earth will be catastrophically desertified due to the expansion of the sun. The emergent complexities of our homeworld will, one-by-one, break down, life will be reduced to its simplest forms until only extremophiles exist, and then even the extremophiles will go extinct, and our lush, living world will be a sterile desert, if it is not actually consumed by an engorged red giant sun. If Earth survives the sun’s red giant stage, the Earth will become a frigidly cold desert once the sun shrinks into a white dwarf. At this stage in the history of our solar system, it is interesting to speculate on whether even the artifacts of a technologically sophisticated civilization would leave any technosignatures after so much time, and after so eventful a history.

A contemporaraneous illustration of the 1607 Bristol Channel floods


One of the most common existential risk thought experiments today is that of civilization faced with a catastrophic greenhouse event due to anthropogenic warming of the global climate. In the above thought experiments on snowball events and desertificaiton, I focused on low technology civilizations primarily dependent upon agricultural and pastoralism. I will continue this theme, so the reader should understand that I am here engaged in a counter-factual exercise (which could well be realized somewhere else in the cosmos, so that it may not be a counterfactual in the strict sense). Our civilization today possesses considerable technological resources that it can deploy in order to mitigate the impact of catastrophic events. Such civilizations are not at issue in the present thought experiment.

Perhaps more important than technological resources, is the relationship that a technological civilization has to its own future. We know that if we study some phenomenon scientifically we can understand it better over time, and our future knowledge can provide us with technologies for existential risk mitigation, and that these technologies will improve with time. We usually simply assume this without explicitly realizing how different this is from the attitude of past civilizations, which understood themselves to be at the mercy of events they could neither understand nor control. Pushing the envelope of some technological development as a race against time, far from being a disheartening acceptance of fate, is an invigorating experience of human agency. This is insufficiently appreciated today, but we may well re-discover the power of purposeful activity when we once again face an existential threat.

The concern with climate change today is that the carbon released into the atmosphere since the industrial revolution may result in a runaway greenhouse effect, making the planet hotter, melting the polar ice caps, flooding former coastal areas, and changing weather patterns. This causal mechanism would not be a concern to a low technology civilization, but such a civilization could still find itself exposed to a greenhouse existential risk, for example, if a supervolcano were to erupt, injecting carbon into the atmosphere equivalent to or greater than an industrial revolution relying upon fossil fuels. Let us suppose this or some similar scenario transforms a planet with an agricultural civilization into a hot and humid greenhouse. What happens next?

I have often emphasized that the civilization of early medieval Europe was almost purely agricultural in character, and, as such, it was an inland civilization in which cities and ports played a minimal role. In the event of catastrophic rises in sea level due to rising temperatures, the whole of society would be severely disrupted, but if the populations, forced to migrate, were able to find arable land with the right climate, civilization could continue.

Floods have occurred throughout human history, and have been documented in writing and images (and, of course, there is the story of the Biblical flood, and illustrations of this no doubt drew upon contemporaneous experiences of flooding). However, flood waters usually recede, and the residents of briefly flooded areas can return and attempt to rebuild. In a catastrophic greenhouse climate scenario, the flood waters would not recede, but would be followed by further storms pushing the flood waters further inland. It might be hundreds of years, or thousands of years, before new coastlines stabilized, allowing for cities and ports to be built on the newly established coastline. Hence the social disruption mentioned in the previous paragraph.

If the disruption became great enough, or populations forced to migrate from regions no longer capable of supporting agriculture could find no arable land, any semblance of political order would be lost in a free-for-all of bare survival. An agricultural civilization needs little to continue, but it does at least require arable land and a climate in which food crops can be grown. All of this would be disrupted by a sufficiently severe greenhouse episode, which could easily cause the end of such a civilization.

What technosignatures would remain from a rudimentary civilization in a hot and wet greenhouse planet? Former coastal cities would be deep under water, and would be preserved to a certain extent by being submerged. Here a submersible could return important information to any archaeological survey team. It would not be difficult to determine former coastlines of continents, and a submarine could be sent to document these former coastlines, revealing ghostly underwater cities, probably better preserved than if they had been exposed to the heat and humidity of a greenhouse climate.

Anything of the former civilization that remained above ground would be subject to climatological conditions that would be unfriendly at best to the preservation of artifacts, especially the artifacts of a low technology civilization, which are mostly constructed from organic and naturally occurring resources, are subject to rapid degradation. With high temperatures and warm rains lashing down, most of what remained of the civilization would be rotted away. Buildings constructed of stone would endure for a time, but if the warm and wet climate produced something like a planetary scale rainforest, the vegetation would rapidly take over any ruins.

The remains of a civilization under these circumstances would then be like the remains of the Mayan and Khmer civilizations — civilizations of the tropical rainforest biome — overgrown by luxuriant tropical forests, to the extent that it would be difficult to see them. However, one of the interesting developments archaeology since the Space Age has been the ability to detect ruins under the canopy of tropical rainforests (cf. Maya Ruins) as well as in deserts (cf. Space Archaeologist Sarah Parcak Uses Satellites to Uncover Ancient Egyptian Ruins), now called satellite archaeology. There are even satellite surveys that have revealed ancient roads long hidden under deserts (cf. CORONA Satellite Photography and Ancient Road Networks: A Northern Mesopotamian Case Study).

Needless to say, any archaeological survey team in possession of the technology to travel to other worlds would have long before been practicing satellite archaeology on its homeworld, and the first thing they would do in approaching a planet for the first time would be a complete set of photographs that could be rapidly analyzed by AI optimized to look for the telltale signs of past civilization. While these methods would work quite well in the case of desertification, as I noted in my previous post, deserts are greater preservers. Greenhouse climates, however, are not great preservers, so the fact that satellite archaeology can uncover Mayan ruins in Mesoamerica points to the possibility of the utility of satellite archaeology in studying planets on which a civilization succumbed to a greenhouse event. And once the orbital images had revealed traces of a past civilization, the submersibles could be released into the oceans to look for those ghostly cities along former coastlines.

A Final Thought

Civilizations that made a particular effort to leave monumental architecture would be much easier to detect ex post facto than civilizations that did not engage in the construction of monumental architecture. However, it should be noted that monumental architecture need not be merely symbolic in order to be sufficiently massive as to be obvious after long periods of time. A civilization that engaged in the construction of enormous dams, or irrigation projects, and so on, might leave ruins as durable as the ruins of the Egyptian pyramids, which latter may outlast any monuments of technological civilization and still testify to the human experience long after human beings are extinct. The example of the pyramids demonstrates that even early agricultural civilizations, if they have a monumental preoccupation, can leave traces of themselves that will rival or outlast those of technological civilization.