The Köppen–Geiger climate classification made simpler (I hope so)

[Pixie]

I'm still behind in my project (which is bound to take years), but I follow this thread with renewed interest each time you guys post something this rich into the discussion.

Just wanted to say, Charerg and Azelor, that I am a passionate lurker of your work Keep it up.

[Charerg]

Thanks for the comment, Pixie, it's nice to know that there's an audience reading this stuff .

But to continue my "Jurassic World" project, I've finished with the temperature maps. So, first off, the Zone of Temperature maps:

January:
180 Mya - Jan Influences.png

July:
180 Mya - Jul Influences.png

And then the temperature maps themselves:

January:
Jan Temp - 180 Mya.png

July:
Jul Temp - 180 Mya.png

I have to say that I had a much harder time making these than I expected. Mostly because it's really difficult to tell which climate each location should have, and multiple interpretations tend to be possible from the evidence. This would be much easier for Early Cretaceous era and beyond, since that's when flowering plants turn up. For the Jurassic, it's mostly an extinct ecosystem, even in terms of plants, with only some living fossils around. That said, I generally used the appearance of umbrella ferns (Dipteridaceae) as an indicator for a subtropical-to-tropical climate (since the modern species of the family don't grow outside of subtropical/tropical areas). Likewise, I've regarded the appearance of Ginkgo trees as indicators of a temperate climate (the tree is a living fossil, nowadays it only grows in China).

That said, this is all very speculative, owing to the limited evidence, and the difficulty in the interpretation. However, I think this is "close enough" that it should produce results that may be considered acceptable (if not necessarily accurate).

One thing worth noting about the Jurassic "greenhouse climate" is the low temperature gradient from the equator to the poles. Generally speaking, it seems the polar 70-90 latitudes had climates comparable to maybe 40-50 latitude temperate climates of today. New Zealand, for example, while situated very close to the South Pole (at least probably, several interpretations exist about this as well), had a climate that appears very similar to the climate New Zealand has today!

[Charerg]

So, I finished the precipitation maps. Before I get into the maps themselves, some notes:

1. ITCZ Fluctuation

In my earlier atmospheric features maps, I had envisioned the ITCZ fluctuating wildly from 30 N to 30 S in mainland Pangaea. When making the precipitation maps and comparing them to lithological and floral climate indicators, it soon became apparent that the ITCZ could not have fluctuated so wildly. The Kalahari area, the regions around the Gulf of Mexico and the Parana basin all have Jurassic deposits of aeolian sandstones, indicating the presence of vast dunes. It seems that the ITCZ fluctuation was comparable to that of modern Africa (approx. 20 N to 20 S), which in hindsight is perhaps obvious. In any case, that's an important note to take home. It seems that extreme ITCZ fluctuation as seen in modern Asia only occur if an equatorial ocean is present (the Indian Ocean on present-day Earth), with a major subtropical landmass. So, in the Jurassic the ITCZ may have reached 30 N and 30 S latitudes in the regions arround the Tethys Ocean (in Australia and Asia, in other words), but probably not in mainland Pangaea.

2. The Amazon Basin

On the Paleogeographic Atlas Project page I linked previously, Reese et all. (2000) reconstruct the Amazon Basin as having a tropical summerwet climate (savannah or hot steppe in Köppen). However, this conclusion was apparently made without data from the Amazon Basin itself, instead deposits from regions as far from each other as Equador and Morocco were taken, and an assumption made that the area in between had a comparable climate. I haven't studied the issue in detail, but on a cursory search it seems that more modern studies seem to disprove this idea, and instead suggest a very arid desert environment for Jurassic Brazil (see this web article from 2016, for example).

I suspect the "desert hypothesis" is more accurate, since the following factors were present:

a) The Amazon Basin wass surrounded by mountains (Guiana Highlands to the north, the Cordillera to the west and the African-Brasiliano mountains to the east), blocking oceanic influence.
b) There were vast desert both north and south of it.
c) It was far from all major bodies of water except the Panthalassic Ocean, but I think Panthalassic influence would have been reduced by factor a, and also by the general easterly direction of the dominant trade winds, which would have brought dry desert air into the Amazon region from the surrounding deserts.

So, I've decided to accept the desert hypothesis as the more accurate. Although given that the likelihood of equatorial deserts forming within supercontinental landmasses was one of the main questions I wanted to answer with this particular "climate study", I might do a bit more research on the Jurassic of Brazil in the future.

3. Tien Shan and the Tarim Basin

The northern Tarim basin has very rich coal deposits from early-to-mid Jurassic, suggesting vast marshlands and a generally humid climate. Furthermore, paleocurrent indicators suggest that no significant rainshadow existed (although currents seem to flow away from the Tien Shan mountains, indicating that it was a geologic feature already). From the late Jurassic onwards the Tarim becomes increasingly arid, and the Tien Shan develops a clear "wet flank" and "dry flank". Some of this humid->arid transition in the Tarim may be explained by the southward movement of Eurasia (bringing the Tarim region into the influence of the Tethys High). However, it is also likely that my elevation reconstruction is wrong, and that the Tien Shan was quite low during the period, only reaching greater elevations with the collision of the Lhasa Terrane with Eurasia.

As I'm too lazy to redo the elevation map (which depicts the Tien Shan close to its present-day elevation), I've given the range the significant rainshadow it probably would have if it was as high as in my reconstruction, but it's worth noting that it was likely much lower in reality (and lacked a major rain shadow). Likewise, the Tarim basin probably had a more humid climate in reality than what it will have in the eventual climate map.

4. Antarctica

It's worth noting that the precipitations in Antarctica are largely guesswork. Because of it's glacier-covered nature, Antarctica represents a large data gap in the knowledge of Gondwanan climates. Although there are Jurassic deposits from the Antarctic peninsula (which suggest a subtropical climate), next to no data exists for continental Antarctica.

5. Eastern North America and Europe

The Early Jurassic in eastern NAmerica and southern Europe appears surprisingly dry. Evaporite deposits exist in Nova Scotia, Portugal and Southern England, suggesting that a relatively arid climate was the norm. Considering their location on the eastern edge of Pangaea, this is unexpected (I would have presumed a relatively humid climate, like modern-day Mesoamerica). I can only presume that the area was heavily affected by arid winds from the interior of Pangaea, perhaps something like the island of Socotra today. It seems that in general, southern Europe had a fairly arid climate, though the rainfall was spread more-or-less evenly throughout the year (the absence of clear growth rings in fossilized tree stumps suggests the lack of a clear dry season).


And with the disclaimers out of the way, here are the the precipitation maps themselves:

January:
Prec Jan - 180 Mya.png

July:
Prec Jul - 180 Mya.png

So, next up are the climates themselves. I'll post the unmodified map, since the idea with my modifications to the tutorial is to eliminate the need to do major corrections to the climate map (presuming that the temp and precip maps have been sufficiently carefully done). It will be interesting to compare that to other reconstructions of early Jurassic climates, and see how the results look like.

[Charerg]

Ok, here are the climate maps:

First off, the unmodified version:
Köppen Climates Alt.png

Regarding the goals of my modifications to the tutorial, the results seem good: there is virtually always a band of BS between BW and more humid climates (although it's a bit narrow in some cases like southern Africa, but that's because I didn't make the different precipitation belts sufficiently broad). Likewise, A climates have a nice Af->Am->Aw/As transition. That said, the equatorial region in Africa is a bit messed up (this happens often with the tutorial), and in general there is clear need to do some cleanup, if only to make the map more readable. So, all in all, it might be said that my modifications have improved the situation (particularly the distribution of the A climates and BS), but in practice it's largely impossible to create sufficiently high quality temp and precip maps to completely eliminate the need for manual modifications on the climate map.

In the map itself, I seem to have made a few regions a bit too dry, and the BS belt should be much broader in tropical Pangaea, but other than that it looks fairly plausible for the most part. It's worth noting that there is an As->Cs transition in the western areas of the continents, which looks a bit weird because the As climate looks like Am in this colour scheme (the Köppen colour scheme wasn't exactly designed with Jurassic climates in mind).

Here's the "cleaned up" version:
KöppenC Alt Cleaned.png

In the cleaned version, I've modified the equatorial climates so they make sense, and also adjusted the extent of BS so the result looks more plausible. I also got rid of some weird high-latitude steppes and deserts. A general overview of the climates:

Overview:

As expected, D and E categories are almost entirely absent. This is consistent with known evidence: it appears that very few locations on Earth experienced sub-zero temperatures during the Jurassic. The only substantial D climate to make an appearance is Db, which occurs in the Altai-Sayan region due to the high elevations of the area. Some ET and EF also appear around the very highest mountains, but in general both D and E are extremely marginal.

Low latitude climates (0-30):

Tropical Pangaea is dominated by an overall arid climate, with savannah (Aw), steppe (BS) and desert (BW) being the most common categories. This is consistent with the data: it appears that equatorial rainforests were almost nonexistent during the Jurassic. In fact Reese et all. (2000) reconstructed them as entirely absent on the Paleogeographic Atlas Project page I have linked previously. That said, climate models predict some rainforest turning up in the Indonesian regions and eastern Africa. As the available data is limited, there is no way to verify this, but my interpretation does agree with the climate models in that it seems plausible that some Af and Am would turn up in these locations.

The absence of equatorial rainforest seems to be due to a number of factors: the first is the overall arrangement of the continents, which does not favour the formation of rainforests near the equator. The megathermal climate of the Jurassic was probably also a factor: the tropical latitudes would have experienced extreme temperatures, and evaporation would have been consequently very high.

Mid latitude climates (30-60):

A characteristic feature of the Jurassic were the "Paratropical" climates that occurred in mid-latitude areas. In general, evidence indicates that the greatest biodiversity in both flora and fauna during the Jurassic occurred in the mid-latitude regions. Essentially today's "Temperate Zone" was tropical rainforest during the Jurassic for the most part. In my climate map, I've generally placed the A/C boundary around 45, but tropical conditions may have extended even further polewards (and subtropical conditions probably all the way to around 70 latitudes).

Again, this seems due to a number of factors: first off, falling under the influence of the westerlies (though some areas may have experienced a monsoonal climate, like Asia today), mid latitude areas would have received rainfall on a fairly regular basis. In addition, the megathermal climate would have ment warmer air carrying more moisture, resulting in a more humid climate in these areas than experienced today. At the same time, they weren't quite as hot as tropical regions, reducing the evaporation rate, which probably also favoured the formation of vast forests.

High latitude climates (60-90):

During the Jurassic, high latitude areas had a climate comparable to today's Temperate Zone. In my climate maps, I've retained the Cc climates that the tutorial generates, but the vast majority of Cc should actually be Cb. Evidence indicates that broad-leaved, apparently deciduous conifers dominated high latitude areas, suggesting that these regions likely did experience a mild winter, but even so truly cold conditions probably did not occur, even in areas extremely close to the poles.

So, that's it for this "climate study". Although the overall pattern of the climates should be a fairly good approximation of how the Early Jurassic Köppen climates would have looked like, keep in mind that this isn't necessarily accurate when it comes to the details (for example, the Tarim basin was probably more humid, like I mentioned in my previous post). It has certainly been an interesting and hopefully educative experiment!

Edit:
I updated the "cleaned version" of the climate map. The colour for As has been changed, so it's now much easier to identify. I also made a few minor changes.

Edit2:
I was inspired to refine the climate map a bit further. The main difference is that Asia is much more humid, which reflects the distribution of Jurassic coal deposits much better. My prior version would be more representative of Late Jurassic onwards, when Asia started to become increasingly arid. I also removed the high-latitude deserts from Antarctica. Although no evidence exists from continental Antarctica, it seems unlikely that high latitude deserts existed there, given that none appeared in Siberia. In addition, I replaced all Cc with Cb. Some Cc probably would have appeared in mountainous areas, but as I was a bit lazy, I chose to replace it all with Cb, since Cc climates tend to be fairly marginal. So, here's the final map, this time in the "Azelor colour scheme":

Köppen Climates.png

[Charerg]

Here's another "bonus step" to the tutorial for those interested, this time intended to refine the classification of Cb/Cc and Db/Dc climates.


Definition of the boundary between temperate (a/b) and subpolar (c/d) climates in Köppen:

As those familiar with the Köppen system know, the boundary between the a/b and c/d sub-classes is defined based on the length of the growing season (this is proxied by the number of months with mean temperature above 10 C). Both a and b sub-classes require at least 4 months with mean temperature above 10 C (approximating a growing season of at least 4 months), whereas the c and d climates have 1-3 months.

Since in the tutorial, we only generate the average temperatures of two months (January and July), we can't determine based on the existing maps how long the growing season is in each area. So, you need to create a new map, the number of months with mean temperature above 10 C. Here's one for Earth, generated from the WorldClim data I used earlier:

No. of Months above 10 C MTemp.png

Other uses of the map:

Once you have a similar map of your world, you can straight up define the boundary (just pick the 1-3 months categories, and you have your c/d climates). Btw, you should already have the "no months above 10 C" category covered by default (since that is the definition of E climate group, so it's synonymous with tundra and ice cap). Also, all A climates qualify as "12 months above 10 C", although the "12 months" category also extends to Ca, and even Cb in some cases.

In addition to allowing accurate definition of Cb/Cc and Db/Dc climates, a map showing the length of the thermal growing season can be useful information in other ways. For example, it can give you an idea about how productive the different areas might be in terms of farming (though precipitation isn't taken into account here, but your climate map should cover that).

Also, you can use this map to split the Ca category into "true subtropical" climates with a minimal cold season, and more temperate climates with a longer cold season. As Köppen is often criticized (particularly by Americans who don't think New York should be classified as subtropical ) for having an overly broad Ca category, this may be desirable. Here's an example based on the previous map of Earth:

Climate types by length of the growing season.png

In the Trewartha classification, subtropical climates are defined as having at least 8 months above 10 C. I used a bit more strict criteria here, and drew the boundary at 9 months minimum. Since the definition of "subtropical" is very subjective, to say the least, you can choose whatever criteria you consider most accurate.

Edit:
I should mention that my "subtropical and tropical" category does include b as well, because a requires a hot summer (over 22 C average in hottest month). So, a tropical highland climate where the mean monthly temperature remains always above 10 and below 22 would be classified as Cb, even though it's quite different from, say, England which has a clear cold season. That's one of the advantages of doing a map like this in addition to the main tutorial: you can differentiate between climates that are seemingly similar in Köppen, but in actuality totally different.

[Pixie]

The point that Geoff's and Azelor's techinques only care about the extreme months is indeed a limitation. And you might be into something to break out of that, Charerg, but you would need to come up with a way to predict that kind of map you produced for Earth, for any given arrangement of land/ocean (which we like to call conworld)... Any thoughts?

[Azélor]

I did it with the mean temperatures on a spreadsheet (Excel) assuming the changes of temperatures were constant all year long.
Then counter the number of months the temperature is under a specific threshold.
Following that logic, 1 combination is supposed to be Cc but it's also possible to have a Cb with the same combination if the variation of temperature is different.

The map with the mean temperature above 10 C is indeed interesting but can it be done by using data we already have, such as temperatures?
Using more temperature categories would work but require a lot more thinking when mapping it.

[Charerg]

The point that Geoff's and Azelor's techinques only care about the extreme months is indeed a limitation. And you might be into something to break out of that, Charerg, but you would need to come up with a way to predict that kind of map you produced for Earth, for any given arrangement of land/ocean (which we like to call conworld)... Any thoughts?

As the "months above 10" depends on largely the same factors that influence temperature, it should be possible to figure out approximate instructions for creating a map like this based on latitude, altitude and influences (continental, cold current, hot current). Those pieces of information we already have from the tutorial, only the effects of each need to be determined (like Azelor did in the temperature section).

Like Azelor mentioned, mean temperature could be used to approximate this, if we had more categories, but I think there would still be a lot of anomalies, and ultimately it's probably better to generate the "length of growing season map" separately as a "bonus map", so that the main tutorial remains as it is (since it's probably already complicated enough for most users).

[PaGaN]

Hi Azelor.
Just wanted to stop by and say Hi and compliment you on the excellent work done in this thread.
I will be calling on your hard work here when I (hopefully soon) get to the climate design for the world I'm working on. She's a WIP called Aerlaan over in the Regional/World Mapping forum.
Your feedback and opinions would be very much appreciated.
Thanks.
PaGaN

[acrosome]

I came back here after a couple of years hiatus and, wow, this is awesome. (I've had climate discussions with Azelor and pixie in the past.) I was inspired enough by this method to read a bit about making GIMP plugins since I don't have Photoshop but, well, I never considered myself a luddite but maybe I am because that stuff is clearly beyond me.

OTOH, since I'll really only have to do this once maybe I can do it by hand...