Hurried Thoughts: Phanerozoic Koppen-Geiger Climate Maps

So a while back I ran across the paper "A high-resolution climate simulation dataset for the past 540 million years", which is...exactly what it sounds like it is, simulating climate over the Phanerozoic (time of complex life) at 10-million-year intervals with the CESM climate model. The paper provides some basic information on climate trends over that period, but they also provide some of the model outputs and I realized that I could run these through my koppenpasta script with only slight modification.

Some notes on the model limitations:

• First off, the paper details the approach whereby they adjusted the model CO₂ level to achieve a temperature profile matching geological records, which often resulted in a level higher than that predicted from geological CO₂ proxies; the authors attribute this mostly to errors in the vegetation model. 
• They also use that vegetation model for all the model runs, even though widespread land vegetation only appeared after around 400 million years ago. 
• They used modern orbital parameters for all runs, which is reasonable enough given that we have no evidence of them substantially changing aside from the minor fluctuations of the Milankovitch cycles.
• The CESM model also has some known issues at the equator, including an odd tendency to produce a pair of closely spaced tropical rain belts rather than the single belt we see in reality; I don't think this affected the climate maps much but it's probably why you can see patches of cooler water over the equator in a couple of the maps.
• All model outputs up to 300 million years ago have a small patch of land over the north pole that doesn't correspond to any real landmass; this may be an input or output error or a hack to fix some issue with ocean dynamics, but again I don't think it causes any real issues with the rest of the global climate.

In producing the maps, I also had a slight issue with the sea climates; the outputs included no direct data on sea ice cover, but it did show surface albedo, which clearly showed the contrast between the reflective ice and open sea. It only showed albedo for areas exposed to sunlight, though (so excluding polar areas in mid-winter), and there were odd artifacts in the albedo data at the edge of the sunlit region. I ultimately went for a somewhat awkward approach of using temperature to estimate ice cover in the unlit regions (using equation 11 here) and help filter out some of the artifacts, with some dialing in by eye to try and match the lit and unlit areas as best I could; but you'll still see a couple spots on some maps with oddly less or more ice cover than their neighbors.

Anyway, all the resulting maps can be found in this folder (if you're a patron of mine, it's the same I put on the Patreon a while back). It includes both Koppen-Geiger maps, labeled by the year, and maps with my Holdridge approximations (using average temperature and precipitation but no potential evapotranspiration ratio), labeled with "h" and the year. The model was run at a resolution of 192x288, and for ease of viewing I scaled the maps up to 512x1024 without any interpolation (this rescaling also makes them Plate Caree equirectangular projections if you want to reproject them or look at them in Gplates).

 

While we're here, we might as well look over a few highlights.

 

The Global Mean Surface Temperature in the model outputs (black line) compared to reconstructed temperatures based on geological evidence (red stars). Li et al. 2022

530 Ma (Fortunian, early Cambrian)

This is the world during the "Cambrian explosion", the period when many modern animal groups first appear in the oceans; an extreme hothouse with global temperatures averaging around 26 °C, though even here the massive continent of Gondwana has some cooler areas where it stretches over the south pole—not that it matters much given that little more than microbes inhabits the land at this point. 

 

Most of our best fossils from the Cambrian come from the coastal seas of south China (the small equatorial continent on the far right) or North America (the large continent on the middle left), showing a variety of unusual early animals like Anomalocaris and Hallucigenia, as well as ultimately more enduring groups like trilobites and the earliest chordates.

440 Ma (Aeronian, early Silurian)

Temperatures remain high through the Cambrian and most of the Ordovician but then plummet towards a nadir of around 18 °C—high by modern standards but quite a drop from the previous hothouse and enough to trigger glaciation over much of Gondwana (and it may have dipped lower at the coolest points). This may have caused the End-Ordovician Mass Extinction at around 443 Ma, but there's some nuance to that we'll discuss another time.

 

The Ordovician (and to a lesser extent the Silurian) saw significant diversification of mollusks (including the great orthocone nautiloids), arthropods (sea scorpions, trilobites), echinoderms (crinoids, sea stars), and early vertebrates, leaving abundant fossils mostly in North America and Europe (the cluster of landmasses on and south of the equator in the center), with some from Morocco (along the Gondwanan coast at the bottom), and some particularly good fossils close to the extinction event itself in the Soom Shale in South Africa (Gondwanan coast near the left edge).

390 Ma (Eifelian, mid Devonian)

This is the climate when plants first begin dominating the land and developing the woody tissues required for tall growth. Temperatures have gradually recovered to another peak at around 25 °C, and a fortuitous arrangement of continents provides plenty of wet, warm territory for plants to conquer. Later in the Devonian, temperatures will fall again, which may contribute to the Late Devonian Mass Extinction around 372-359 Ma, but that's an even more complicated story we'll have to come back to another time.

 

The famous Rhynie Chert in Scotland (on the east coast of the large equatorial landmass in the center) shows some of these earliest terrestrial ecosystems, including early plants, algae, crustaceans, arachnids, and the 8-meter-tall fungus Prototaxites; and some of the earliest tetrapods (vertebrates with 4 limbs and feet) have been found in Canada and Greenland (northwest coast of the same landmass). But the Devonian is still often known as the "Age of Fishes" thanks to the diverse jawed fish, including the armored placoderms, found mostly in Europe (east coast of the same landmass again) and Australia (the large peninsula near the right edge).

310 Ma (Moscovian, late Carboniferous)

This one may surprise some people; the Carboniferous is widely thought of as a time of sweltering swamps and jungles, but in fact it was one of the coldest periods of the Phanerozoic, with average temperatures falling to around 13 °C near its end. 

It just happens that despite the low average temperature, the arrangement of continents still allowed for lush tropical rainforests over much of North America (middle and west coast of equatorial Pangea in the center), Europe (east coast of the same landmass), and China (the equatorial landmass on the right and the long peninsula above it), populated by scale trees, horsetails, amphibians, and immense insects and other arthropods. These would ultimately produce many of the rich coal deposits mined today; the southern continents were largely glaciated at the time and so are poorer in coal today, though some substantial deposits in South Africa and Australia formed later.

250 Ma (Olenekian, early Triassic)

The Pangean climate: temperatures rise through the Permian to a peak of around 25 °C, likely even higher during the climax of the End-Permian Mass Extinction. Vast deserts stretch across much of the supercontinent's interior, but rainforests still persist along the equator. The arc-shaped arrangement of landmasses around the equatorial Tethys Ocean creates the conditions for the Pangean megamonsoon; the ITCZ moves far north and south with the seasons, such that much of the subtropical east coast swings between torrential summer rains and intense winter droughts, and some of the equatorial coast in between is left notably dry.

 

The Permian saw the rise of the synapsids, including some of the earliest large terrestrial vertebrate grazers, such as Scutosaurus and the dicynodonts, and predators, such as Dimetrodon and the gorgonopsids, commonly preserved in the deserts of North America (equatorial west and northwestern parts of Pangea), savannas and steppe of Africa (much of the central and eastern parts at and below the equator), and forests of Siberia (northern part of the large northeastern lobe). After the extinction, a number of unusual reptiles emerged in the Triassic, many best preserved in central Asia (southern part of the same region), before the dinosaurs emerged to dominance in the later part of the period.

170 Ma (Bajocian, mid Jurassic)

Temperatures gradually decline through the Triassic and early Jurassic, but only as low as about 19 °C. As Pangea breaks up, more moisture can reach into the interior of the continents and the deserts recede.

 

Glass sponge reefs extend across much of the young Atlantic, and fluctuating sea levels often cover much of the lower areas of Europe (upper center) in shallow seas populated by ammonites, icthyosaurs, and plesiosaurs, as pterosaurs fly overhead. Later Jurassic fossils attest to sauropods, stegosaurs, and allosaurs in North America (upper left) and the earliest birds and their close cousins in Europe and China (east coast of the upper right continent).

80 Ma (Campanian, late Cretaceous)

 

Breakup of the continents and increased volcanism helps drive temperatures back up to a peak of around 24 °C, and their particular arrangement and topography helps limit the extent of deserts and favor broad tropical and subtropical forests including the earliest flowering plants. The polar regions still dip below freezing in winter, but warm, wet summers keep them lush. Shallow seas extend across much of the continents, slicing apart North America, Africa, and Eurasia.

 

By some estimates, Earth at this point may have had more total biomass than at any other point in its history, and fossils from the Cretaceous are found on just about all landmasses; tyrannosaurs, dromaeosaurs, ceratopsians, and hadrosaurs roamed North America and China (I presume by now you can more-or-less recognize the modern continents), immense titanosaurs reached their peak in South America, abelisaurs and unusual crocodilians diversified in Madagascar, Spinosaurus swam the coastal waters of North Africa while mososaurs and elasmosaurs swam further offshore, and azhdarchid pterosaurs flew the skies or terrorized the islands of Europe.

40 Ma (Bartonian, Eocene, mid Paleogene)

Temperatures stay high through the Cretaceous and early Paleogene, perhaps spiking to over 30 °C during the Paleocene-Eocene thermal maximum at 56 Ma, but begin to decline after 50 Ma. This point in the late Eocene, with an average temperature of around 20 °C, is the latest in this dataset before the initial glaciation of Antarctica at around 34 Ma, after which temperatures would fall by around 4 °C. We can see the modern global climate starting to take shape: despite broad warm subtropics, polar tundra is beginning to appear; and this world is still wetter than the modern day, but the collision of India into Asia, rise of the Rockies, and northward drift of Australia are starting to form the modern deserts.

 

By this point, the world has largely recovered from the End-Cretaceous Mass Extinction at 66 Ma; dinosaurs have been replaced by ungulates like Uintatherium and the brontotheres found across North America, Europe, and Asia, while Hyaenodon and the pig-like entelodonts have emerged as predators, though they must compete with flightless terror birds. Cetaceans have taken to the seas and quickly grown to large whales like Basilosaurus. South America, still isolated from North America, is developing its own unique assemblage including astrapotheres and borhyaenids. But the cooling and drying of the planet is stressing these ecosystems; coastal Antarctica began the Eocene as a subtropical forest but is transitioning to tundra, and the spread of savanna and steppe in the later Oligocene will favor the rise of grasses.

Recent (Meghalayan, Holocene, late Quaternary)

This is CESM's attempt at the pre-industrial climate (in this particular configuration), as a point of reference. The greater detail here is largely just due to more detailed input topography. Where ExoPlaSim tends to be a little too arid, this seems a bit too wet, particularly in Australia; but this model is much more accurate to the East Asian monsoon and the current-aided warming of coastal Europe.

I'll finish off with a quick animation of the whole sequence. See you next time.

Buy me a cup of tea (on Patreon)