**Gavin A. Schmidt** is a climatologist, climate modeler and Direct...
I’ve pondered this ‘possibility’ all my length - that there have be...
One of the most common searched structures that can be seen as an e...
If you are curious here is below is a clip of the first appearance ...
The Negev Desert in Israel has been dated to about 1.8 million year...
It has been estimated that only one bone in a billion gets fossilis...
Carbon has 3 naturally occurring isotopes - $^{12}C$, about 99% of...
Another element that would be detectable is technetium-98 ($^{98}T$...
The problem is that if the activity of a hypotetical industrial civ...
International Journal of
Research Article
Cite this article: Schmidt GA, Frank A (2019).
The Silurian hypothesis: would it be possible
to detect an industrial civilization in the
geological record? International Journal of
Astrobiology 18, 142150. https://doi.org/
Received: 11 October 2017
Revised: 21 February 2018
Accepted: 25 February 2018
First published online: 16 April 2018
Key words:
Astrobiology; Drake equation; industrial
civilization; Silurian hypothesis; Anthropocene;
Author for correspondence:
Gavin A. Schmidt,
© Cambridge University Press 2018
The Silurian hypothesis: would it be possible to
detect an industrial civilization in the
geological record?
Gavin A. Schmidt
and Adam Frank
NASA Goddard Institute for Space Studies, 2880 Broadway, New York, NY 10025, USA and
Department of Physics
and Astronomy, University of Rochester, Rochester, NY 14620, USA
If an industrial civilization had existed on Earth many millions of years prior to our own era,
what traces would it have left and would they be detectable today? We summarize the likely
geological fingerprint of the Anthropocene, and demonstrate that while clear, it will not differ
greatly in many respects from other known events in the geological record. We then propose
tests that could plausibly distinguish an industrial cause from an otherwise naturally occurring
climate event.
The search for life elsewhere in the universe is a central occupation of astrobiology and scien-
tists have often looked to Earth analogues for extremophile bacteria, life under varying climate
states and the genesis of life itself. A subset of this search is the prospect for intelligent life, and
then a further subset is the search for civilizations that have the potential to communicate with
us. A common assumption is that any such civilization must have developed industry of some
sort. In particular, the ability to harness those industrial processes to develop radio technolo-
gies capable of sending or receiving messages. In what follows, however, we will define indus-
trial civilizations here as the ability to harness external energy sources at global scales.
One of the key questions in assessing the likelihood of finding such a civilization is an
understanding of how often, given that life has arisen and that some species are intelligent,
does an industrial civilization develop? Humans are the only example we know of, and our
industrial civilization has lasted (so far) roughly 300 years (since, for example, the beginning
of mass production methods). This is a small fraction of the time we have existed as a species,
and a tiny fraction of the time that complex life has existed on the Earths land surface (400
million years ago, Ma). This short time period raises the obvious question as to whether this
could have happened before. We term this the Silurian hypothesis
While much idle speculation and late night chatter has been devoted to this question, we
are unaware of previous serious treatments of the problem of detectability of prior terrestrial
industrial civilizations in the geologic past. Given the vast increase in work surrounding exo-
planets and questions related to detect ion of life, it is worth addressing the question mo re for-
mally and in its astrobiological setting. We note also the recent work of Wright (2017) which
addressed aspects of the problem and previous attempts to assess the likelihood of solar system
non-terrestrial civilization such as Haqq-Misra & Kopparapu (2012). This paper is an attempt
to remedy the gap in a way that also puts our current impact on the planet into a broader per-
spective. We first note the importance of this question to the well-known Drake equation.
Then we address the likely geologic consequences of human industrial civilization and then
compare that fingerprint to potentially similar events in the geologic record. Finally, we
address some possible research directions that might improve the constraints on this question.
Relevance to the Drake equation
The Drake equation is the well-known framework for estimating of the number of active, com-
municative extraterrestrial civilizations in the Milky Way galaxy (Drake, 1961, 1965). The num-
ber of such civilizations, N, is assumed to be equal to the product of; the average rate of star
formation, R*, in our Galaxy; the fraction of formed stars, f
, that have planets; the average num-
ber of planets per star, n
, that can potentially support life; the fraction of those planets, f
We name the hypothesis after a 1970 episode of the British science fiction TV series Doctor Who where a long-buried race
of intelligent reptiles Silurians are awakened by an experimental nuclear reactor. We are not however suggesting that intelligent
reptiles actually existed in the Silurian age, nor that experimental nuclear physics is liable to wake them from hibernation. Other
authors have dealt with this possibility in various incarnations (for instance, Hogan, 1977), but it is a rarer theme than we ini-
tially assumed.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1473550418000095
Downloaded from https://www.cambridge.org/core. IP address:, on 21 Apr 2019 at 20:40:28, subject to the Cambridge Core terms of use, available at
actually develop life; the fraction of planets bearing life on which
intelligent, civilized life, f
, has developed; the fraction of these civi-
lizations that have developed communications, f
, i.e., technologies
that release detectable signs into space, and the length of time, L,
over which such civilizations release detectable signals.
N = R
· f
· n
· f
· f
· f
· L.
If over the course of a planets existence, multiple industrial
civilizations can arise over the span of time that life exists at all,
the value of f
may in fact be >1.
This is a particularly cogent issue in light of recent developments
in astrobiology in which the first three terms, which all involve
purely astronomical observations, have now been fully determined.
It is now apparent that most stars harbour families of planets
(Seager, 2013). Indeed, many of those planets will be in the stars
habitable zones (Dressing & Charbonneau, 2013; Howard, 2013).
These results allow the next three terms to be bracketed in a way
that uses the exoplanet data to establish a constraint on exo-
civilization pessimism. In Frank & Sullivan (2016)suchapessimism
line was defined as the maximum biotechnological probability
(per habitable zone planets) f
for humans to be the only time a
technological civilization has evolved in cosmic history. Frank &
Sullivan (2016)foundf
in the range 10
Determination of the pessimism line emphasizes the
importance of three Drake equation terms f
, f
and f
history often serves as a template for discussions of possible
values for these probabilities. For example, there has been con-
siderable discussion of how many times life began on Earth dur-
ing the early A rchean given the ease of abiogenisis (Patel et al.,
2015) including the possibility of a shadow biosphere com-
posed o f d escendants of a different origin event from the one
which led to our Last Univers al Common Ancestor (LUCA)
(Cleland & Copley, 2006). In addition, there is a long-standing
debate concerning the number of times intelligence has evolved
in te rms of dolphins and other species (Marino, 2015). Thus,
only the term f
has been co mmonly accepted to have a value
on Earth of strictly 1.
Relevance to other solar system planets
Consideration of previous civilizations on other solar system
worlds has been taken on by Wright ( 2017) and Haqq-Misra &
Kopparapu (2012). We note here that abundant evidence exists
of surface water in ancient Martian climates (3.8 Ga) (e.g.
Achille & Hynek, 2010; Arvidson et al., 2014), and speculation
that early Venus (2 Ga to 0.7 Ga) was habitable (due to a dimmer
sun and lower CO
atmosphere) has been supported by recent
modelling studies (Way et al., 2016). Conceivably, deep drilling
operations could be carried out on either planet in future to assess
their geological history. This would constrain consideration of
what the fingerprint might be of life, and even organized civiliza-
tion (Haqq-Misra & Kopparapu, 2012). Assessments of prior
Earth events and consideration of Anthropocene markers such
as those we carry out below will likely provide a key context for
those explorations.
Limitations of the geological record
That this papers title question is worth posing is a function of the
incompleteness of the geological record. For the Quaternary (the
last 2.5 million years), there is widespread extant physical evi-
dence of, for instance, climate changes, soil horizons and archaeo-
logical evidence of non-Homo Sapiens cultures (Denisovians,
Neanderthals, etc. ) with occasional evidence of bipedal hominids
dating back to at least 3.7 Ma (e.g. the Laetoli footprints) (Leakey
&Hay,1979). The oldest extant large-scale surface is in the Negev
Desert and is 1.8 Ma old (Matmon et al., 2009). However,
pre-Quaternary land-evidence is far sparser, existing mainly in
exposed sections, drilling and mining operations. In the ocean
sediments, due to the recycling of ocean crust, there only exists
sediment evidence for periods that post-date the Jurassic
(170 Ma) (ODP Leg 801 Team, 2000).
The fraction of l ife that gets fossilized is always extremely
small and varies widely as a funct ion of time, habitat and degree
of soft tissue versus hard shells or bones (Behrensmeyer et al.,
2000). Fossil ization rates are very low in tropical, forested envir-
onments but are higher in arid environments and fluvial systems.
As an example, for all the dinosaurs that ever lived, there are
only a few thousand near-co mplete specimens, or equivalently
only a handful of individual animals across thousands o f taxa
per 100,000 years. Given the rate of new disc overy of taxa of
this age, it is clear that species as short-lived as Homo sapiens
(so far) might not be represented in the existing fossil record
at all.
The likelihood of objects surviving and being discovered is
similarly unlikely. Zalasiewicz (2009) speculates about preserva-
tion of objects or their forms, but the current area of urbanization
is <1% of the Earths surface (Schneider et al., 2009), and exposed
sections and drilling sites for pre-Quaternary surfaces are orders
of magnitude less as fractions of the original surface. Note that
even for early human technology, complex objects are very rarely
found. For instance, the Antikythera Mechanism (ca. 205 BCE) is
a unique object until the Renaissance. Despite impressive recent
gains in the ability to detect the wider impacts of civilization on
landscapes and ecosystems (Kidwell, 2015), we conclude that for
potential civilizations older than about 4 Ma, the chances of find-
ing direct evidence of their existence via objects or fossilized
examples of their population is small. We note, however, that
one might ask the indirect question related to antecedents in
the fossil record indicating species that might lead downstream
to the evolution of later civilization-building species. Such argu-
ments, for or against, the Silurian hypothesis would rest on evi-
dence concerning highly social behaviour or high intelligence
based on brain size. The claim would then be that there are
other species in the fossil record which could, or could not,
have evolved into civilization-builders. In this paper, however,
we focus on physicochemical tracers for previous industrial civili-
zations. In this way, there is an opportunity to widen the search to
tracers that are more widespread, even though they may be subject
to more varied interpretations.
Scope of this paper
We will restrict the scope of this paper to geochemical constraints
on the existence of pre-Quaternary industrial civilizations, that
may have existed since the rise of complex life on land. This
rules out societies that might have been highly organized and
potentially sophisticated but that did not develop industry
and probably any purely ocean-based lifeforms. The focus is
thus on the period between the emergence of complex life on
land in the Devonian (400 Ma) in the Paleozoic era and the
mid-Pliocene (4 Ma).
International Journal of Astrobiology 143
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1473550418000095
Downloaded from https://www.cambridge.org/core. IP address:, on 21 Apr 2019 at 20:40:28, subject to the Cambridge Core terms of use, available at
The geological footprint of the Anthropocene
While an official declaration of the Anthropocene as a unique
geological era is still pending (Crutzen, 2002; Zalasiewicz et al.,
2017), it is already clear that our human efforts will impact the
geologic record being laid down today (Waters et al., 2014).
Some of the discussion of the specific boundary that will define
this new period is not relevant for our purposes because the mar-
kers proposed (ice core gas concentrations, short-half-lived radio-
activity, the Columbian exchange) (e.g. Lewis & Maslin, 2015;
Hamilton, 2016) are not going to be geologically stable or distin-
guishable on multi-million year timescales. However, there are
multiple changes that have already occurred that will persist.
We discuss a number of these below.
There is an interesting paradox in considering the
Anthropogenic footprint on a geological timescale. The longer
human civilization lasts, the larger the signal one would expect
in the record. However, the longer a civilization lasts, the more
sustainable its practices would need to have become in order to
survive. The more sustainable a society (e.g. in energy generation,
manufacturing or agriculture) the smaller the footprint on the rest
of the planet. But the smaller the footprint, the less of a signal will
be embedded in the geological record. Thus, the footprint of civ-
ilization might be self-limiting on a relatively short timescale. To
avoid speculating about the ultimate fate of humanity, we will
consider impacts that are already clear, or that are foreseeable
under plausible trajectories for the next century (e.g. Nazarenko
et al., 2015; Köhler, 2016).
We note that effective sedimentation rates in ocean sediment
for cores with multi-million-year-old sediment are of the order
of a few cm/1000 years at best, and while the degree of bioturb-
ation may smear a short-period signal, the Anthropocene will
likely only appear as a section a few cm thick, and appear almost
instantaneously in the record.
Stable isotope anomalies of carbon, oxygen, hydrogen and
Since the mid-18th century, humans have released over 0.5 trillion
tonnes of fossil carbon via the burning of coal, oil and natural gas
(Le Qué et al., 2016), at a rate orders of magnitude faster than nat-
ural long-term sources or sinks. In addition, there has been wide-
spread deforestation and addition of carbon dioxide into the air
via biomass burning. All of this carbon is biological in origin and
is thus depleted in
C compared with the much larger pool of inor-
ganic carbon (Revelle & Suess, 1957). Thus, the ratio of
the atmosphere, ocean and soils is decreasing (an impact known as
the Suess Effect Quay et al., 1992) with a current change of around
1 δ
C since the pre-industrial (Böhm et al., 2002;Eideetal.,
2017) in the surface ocean and atmosphere (Fig. 1(a)).
As a function of the increase of fossil carbon into the system,
augmented by black carbon changes, other non-CO
trace green-
house gases (e.g. N
and chloro-fluoro-c arbons (CFCs)),
global industrialization has been accompanied by a warming of
about 1°C so far since the mid-19th century (Bindoff et al.,
2013; GISTEMP Team, 2016). Due to the temperature-related
fractionation in the formation of carbonates (Kim & ONeil,
O per °C) and strong correlation in the
extra-tropics between temperature and δ
O (between 0.4 and
0.7 per °C) (and as sensitive for deuterium isotopes rela-
tive to hydrogen (δD)), we expect this temperature rise to be
detectable in surface ocean carbo nates (notably foraminifera),
organic biomarkers, cave records (stalactites), lake ostracods and
high-latitude ice cores, though only the first two of these will be
retrievable in the timescales considered here.
The combustion of fossil fuel, the invention of the Haber
Bosch process, the large-scale application of nitrogenous fertili-
zers and the enhanced nitrogen fixation associated with cultivated
plants, have caused a profound impact on nitrogen cycling
(Canfield et al., 2010), such that δ
N anomalies are already
detectable in sediments remote from civilization (Holtgrieve
et al., 2011).
Sedimentological records
There are multiple causes of a greatly increased sediment flow in
rivers and therefore in deposition in coastal environments. The
advent of agriculture and associated deforestation have lead to
large increases in soil erosion (Goudie, 2000; National Research
Council, 2010). Furthermore, canalization of rivers (such as the
Mississippi) have led to much greater oceanic deposition of sedi-
ment than would otherwise have occurred. This tendency is miti-
gated somewhat by concurrent increases in river dams which
reduce sediment flow downstream. Additionally, increasing tem-
peratures and atmospheric water vapour content have led to
greater intensity of precipitation (Kunkel et al., 2013) which, on
its own, would also lead to greater erosion, at least regionally.
Coastal erosion is also on the increase as a function of the rising
sea level, and in polar regions is being enhanced by reductions in
sea ice and thawing permafrost (Overeem et al., 2011).
In addition to changes in the flux of sediment from land to
ocean, the composition of the sediment will also change. Due
to the increased dissolution of CO
in the ocean as a function
of anthropogenic CO
emissions, the upper ocean is acidifying
(a 26% increase in H
or 0.1 pH decrease since the 19th century)
(Orr et al., 2005). This will lead to an increase in CaCO
ution within the sediment that will last until the ocean can neu-
tralize the increase. There will also be important changes in
mineralogy (Zalasiewicz et al., 2013; Hazen et al., 2017).
Increases in continental weathering are also likely to change ratios
of strontium and osmium (e.g.
Sr and
Os ratios)
(Jenkyns, 2010).
As discussed above, nitrogen load in rivers is increasing as a
function of agricultural practices. This in turn is leading to
more microbial activity in the coastal ocean which can deplete
dissolved oxygen in the water column (Diaz & Rosenberg,
2008), and recent syntheses suggests a global decline already of
about 2% (Ito et al., 2017; Schmidtko et al., 2017). This in turn
is leading to an expansion of the oxygen minimum zones, greater
ocean anoxia and the creation of so-called dead-zones (Breitburg
et al., 2018). Sediment within these areas will thus have greater
organic content and less bioturbation (Tyrrell, 2011). The ultim-
ate extent of these dead zones is unknown.
Furthermore, anthropogenic fluxes of lead, chromium, antim-
ony, rhenium, platinum group metals, rare earths and gold, are
now much larger than their natural sources (Sen & Peucker-
Ehrenbrink, 2012;Gałuszka et al., 2013), implying that there
will be a spike in fluxes in these metals in river outflow and
hence higher concentrations in coastal sediments.
Faunal radiation and extinctions
The last few centuries have seen significant changes in the abun-
dance and spread of small animals, particularly rats, mice and
144 Schmidt and Frank
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S1473550418000095
Downloaded from https://www.cambridge.org/core. IP address:, on 21 Apr 2019 at 20:40:28, subject to the Cambridge Core terms of use, available at