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

[Azélor]

I'm trying to figure how to simplify the temperature groups down to 36 in Excel. It can be done I assume, but it's far easier to do in PS/Gimp. I can't figure out how.
People would have to change the colours. it should not be too complicated.
One thing I'm wondering then that would make the process easier would be to directly use the RGB colours instead of using this : https://www.cartographersguild.com/a...1&d=1453783031
Should we dump them? They look pretty but don't have any particular usefulness.

I could increase the average precipitation for the highest category since 200ml is just the minimum. it would not be too complicated.
Assuming I just continue the progression, the last category would be between 200 and 400, with an average of 300 ml. Do you think that would be good?

here is the modified version for the A climates

sheet7.png

i filled the gaps and Aw received more since it is more common. It's likely to be a arid though.
For Am, there are 2 combo that are the most likely. the light one is the rainiest. The dark one is just above 2000 ml per year. Do you think I should include 1 or both?

[Charerg]

I'm trying to figure how to simplify the temperature groups down to 36 in Excel. It can be done I assume, but it's far easier to do in PS/Gimp. I can't figure out how.
People would have to change the colours. it should not be too complicated.
One thing I'm wondering then that would make the process easier would be to directly use the RGB colours instead of using this : https://www.cartographersguild.com/a...1&d=1453783031
Should we dump them? They look pretty but don't have any particular usefulness.

I could increase the average precipitation for the highest category since 200ml is just the minimum. it would not be too complicated.
Assuming I just continue the progression, the last category would be between 200 and 400, with an average of 300 ml. Do you think that would be good?

here is the modified version for the A climates

sheet7.png

i filled the gaps and Aw received more since it is more common. It's likely to be a arid though.
For Am, there are 2 combo that are the most likely. the light one is the rainiest. The dark one is just above 2000 ml per year. Do you think I should include 1 or both?

Yeah I think that works better. Probably including both combos as Am will work best, I think, though that's something that will just have to be tested out (maybe the light one should still be Af, not sure). Though one thing that I realised from those more detailed precipitation maps you posted is that actually Am doesn't always have to be a transitional climate. Sometimes it can be just a rainier version of Aw. For example, the Am areas in India (the Western Ghats) have as long a dry season as the rest of the subcontinent, they just receive over 2500 mm rain during the monsoon season, so they're still classified as Am even if the driest month has close to 0 mm. But with our precipitation categories we're going to totally miss these cases (mainly occurring in SE Asia), unless we introduce an "above 500 mm" category for extreme monsoon rains or something.

The classification of the A climates is a bit weird in the sense that both Aw and As use the criteria that they are classified based on which season the "below 60 mm" month belongs to. But what if it's below 60 mm in both January and July? If it's 50+50 for example, the annual precipitation would be 600 mm. If we consider that an A climate has at least 18 °C annual mean (because the coolest month must be at least 18 °C), the min. F threshold would be 500 mm (=20*Tann+140). So that would technically not be dry enough to be considered BS, but neither does it receive more than 60 mm in either season. There just doesn't seem to be a climate category for this climate. Personally I did the same thing as you and just classified it as Aw because it's more common. But I'm not sure if this is an ideal solution.


Edit:
One solution that comes to mind is to modify the criteria for B in the case of A climates. I'm not even sure if the usual F, W, S thresholds make much sense for tropical climates to begin with, because the temperature just doesn't vary that much between the seasons in tropical areas. In some cases winter can even be the hotter season in the tropics because it's the dry season (so no clouds to reflect sunlight back to space).

Maybe use the following:

If coldest month > 18 °C:

• At least 2/3 of rain occurs in winter: Pth=20*Tann+140 (this is normally the F threshold)
• otherwise: Pth=20*Tann+280 (the normal W threshold)

That way the precipitation thresholds for Tann 18 °C would be 500 mm (if at least 2x rain in winter compared to summer) and 640 mm (all other cases). This should ensure that all areas that don't receive more than 60 mm in at least one season are always classified as arid.

Edit2:
Although I don't think this comes into effect using your categories, since all combos that don't have 50-100mm or above in at least one season seem to fall below the threshold in any case. Here's the "A sheet" with those categories that will always be considered arid (since the minimum Tann is 18 °C with A climates) marked out:

A_sheet_modified.png

[Azélor]

I was wondering what your 14 climates are since I get 10 only.
For example, what is located at Hot/cool?

Yea about the script. Excel uses marco.
Once the map is imported (still need to figure out how to export hexadecimal code)
The users will convert the file using a macro.
It is similar to using a script in PS or Gimp.
It will look into the hexadecimal code and replace the values with new ones, 1 colour representing each climate. Using the Ctrl+F search and replace
Creating a map that looks like this but in hexadecimal code : https://en.wikipedia.org/wiki/K%C3%B...h_authors).svg
Or maybe with a different colour scheme.
After that, run another script that will use your new values to fill the cells with the appropriate colors. Creating some sort of pixels if the cells are square.
Then delete the hexadecimal code to have the finished product.

The only problem with this maybe, is that the scale of this new created map could be different form the original map.

[Azélor]

Yeah I think that works better. Probably including both combos as Am will work best, I think, though that's something that will just have to be tested out (maybe the light one should still be Af, not sure). Though one thing that I realised from those more detailed precipitation maps you posted is that actually Am doesn't always have to be a transitional climate. Sometimes it can be just a rainier version of Aw. For example, the Am areas in India (the Western Ghats) have as long a dry season as the rest of the subcontinent, they just receive over 2500 mm rain during the monsoon season, so they're still classified as Am even if the driest month has close to 0 mm. But with our precipitation categories we're going to totally miss these cases (mainly occurring in SE Asia), unless we introduce an "above 500 mm" category for extreme monsoon rains or something.

By looking and the older version, the dark one was Am and the light one was Af.
I think I will put the 2 of them and we will see how it look.

The classification of the A climates is a bit weird in the sense that both Aw and As use the criteria that they are classified based on which season the "below 60 mm" month belongs to. But what if it's below 60 mm in both January and July? If it's 50+50 for example, the annual precipitation would be 600 mm. If we consider that an A climate has at least 18 °C annual mean (because the coolest month must be at least 18 °C), the min. F threshold would be 500 mm (=20*Tann+140). So that would technically not be dry enough to be considered BS, but neither does it receive more than 60 mm in either season. There just doesn't seem to be a climate category for this climate. Personally I did the same thing as you and just classified it as Aw because it's more common. But I'm not sure if this is an ideal solution.

I don't know, i'd just leave it that way. Last time you used it to generate a map, was it a problem?

In some cases winter can even be the hotter season in the tropics because it's the dry season (so no clouds to reflect sunlight back to space).

That is very unlikely to happen in the model because we are using categories and averages. Another thing that happens in real life is that the maximum temperature in not always July. For example, it is in May in southern India and I recall an oceanic climate were it was September or October. But I don't think this is relevant. There ain't much we can do about it anyway.

If coldest month > 18 °C:

• At least 2/3 of rain occurs in winter: Pth=20*Tann+140 (this is normally the F threshold)
• otherwise: Pth=20*Tann+280 (the normal W threshold)

I don't know if it is a good idea. I'm compiling some sort of master aridity table with all the 3600 combinations. Actually, almost 2/3 are redundant but they are included anyway.
It look like this so far:
sheet8.png

The red is for desert, green is humid, the rest is a steppe. If you change the temperature, precipitation, or aridity threshold in the original tables, the results here will update.
We will be able to see how everything is placed on the table.

[Charerg]

I was wondering what your 14 climates are since I get 10 only.
For example, what is located at Hot/cool?

There are 10 climates of course, but I merged the temp combinations down to 14. For example, Hot/Cool is Ca, but it's split from the other Ca climates because the annual mean temp is below 18 °C (so it has Bk instead of Bh). Likewise I merged the warm Da climates into a separate category (Da - Temperate) and the cooler Da climates into two separate categories (Da - Cold and Da - Very Cold, the latter being already merged in the original tutorial). I painted the areas I merged with the same colour in the prior post. In terms of actual climates it's the exact same as your climate table, no changes there. I've just merged temp comboes with similar values that share the same climate class in order to simplify things.

I don't know, i'd just leave it that way. Last time you used it to generate a map, was it a problem?

I'm not sure, I'd have to check how commonly it actually occurs.

Edit:
Ok, checked it out, the areas painted in purple are the A climates that have the 40-70 mm category in both seasons (I used the 50 mm avg. for this category, so Pann 600 mm):

50+50 Avg Areas.png

Since I merged all the cooler A climates into a single temperature group and used 23 °C as an avg. annual temp for all of them, the F threshold is exactly 600 mm, as well. So these areas are sitting directly at "borderlne BS". They occur primarily in E. Brazil, equatorial Africa, and as a small strip in Sri Lanka. In Africa, they are caused by having limited data I think (only two months). The regions in Brazil might work better as BS though. But overall I guess it isn't a huge problem to classify these areas as Aw.

[Azélor]

I edit the s f w table to include less combo in the f range. I've found several Mediterranean climates areas that were categorized f. Like Bishkek that barely has a dry season but overall low precipitation. The change in the w is to take in consideration the colder climates like in Eastern Asia.

sheet9.png

[Charerg]

I edit the s f w table to include less combo in the f range. I've found several Mediterranean climates areas that were categorized f. Like Bishkek that barely has a dry season but overall low precipitation. The change in the w is to take in consideration the colder climates like in Eastern Asia.

sheet9.png

So the darkened categories are shuffled over to f, right?

Looks ok to me, since the standard criteria for s would translate to max. 25% rain in summer (whereas w would be min. ~91%), though I understand that some liberties may have to be taken since the precipitation categories are so broad that it's hard to follow the classification exactly.

Edit:
Oh, now I understand it, they are shuffled over from f to s/w. Might make s and w a bit more widespread, but that's probably the case in my version as well, since I chose to err on the side of making them too widespread too, rather than having s/w cover less area than they should.

[Azélor]

Oh, now I understand it, they are shuffled over from f to s/w. Might make s and w a bit more widespread, but that's probably the case in my version as well, since I chose to err on the side of making them too widespread too, rather than having s/w cover less area than they should.

That's it. I'm not sure what the impact will be yet.

Also, I've finished the aridity table. it is similar to what I had done previously.
There is one strange thing occurring in that one S area , despite receiving less precipitations, is more humid than another F area on top of it. This is caused by the difference in threshold.
I could change it manually but since both follow the model, neither are actually wrong.

Attachment 103898

There are the different rain combinations for the moment : 21

Attachment 103899
Red lines are used to separate s, f and w
Am climate are split. half of them are in the f.
1 of the remaining is in the s.
The last is in the w.
It is a little annoying that the Am climate force me to have these 2 distinct categories. Otherwise they would be with S1 and W1.

F1 and W1 both have a steppe counted as humid because having a constant yearly mean temperature over 35 seems very unlikely to happen.

F1 is fractured. The area Aw/s is there because that climate does not fit in the F1 area, it's too dry. The other A/w are scattered quite a bit. I will not make a difference between As and Aw since on Earth, As is only found in some area like on mountains that generate local rain shadow effect. Anyway, it is extremely unlikely to have a As cover a large area.

[Azélor]

Moving forward, I have 34 possible temperatures combos once I exclude the tundra and ice caps.
With 22 precipitation possibilities this means 748 possibilities. How many do you have using you method?
If you have ideas to make it simpler, I'll be glad to hear.

The previous version had presumably 16 combinations for the precipitation but I had to do an extra manipulation in the script to sort the A climate.

edit: this mega sheet display all the 3600 possibilities

sheet12.png

edit : actually, I can get rid of the redundant data by leaving the redundant temperatures cells empty.

sheet14.png

I can get rid of them all by using a simple CTRL+F
search **00
replace with nothing


in green the temperature
in blue the precipitation, the value is the same inside each block
in yellow the results of the combination of the two

In order to do the precipitations, I gave a code to each combo.
it looks like that after the simplification:

sheet13.png

Again the beauty of Excel is that it updated the whole table after I realize I had entered the wrong value in one of the green squares.

---

I quickly achieved this:

sheet15.png

After getting rid of the redundant data and unifying all the tundra and ice caps, I have 748+2 combinations. exactly as expected.

After that, i will probably juxtapose the color for each temp groups (A,Ca,Bc ...) over each section.
Then, around the table, I will copy paste the sections and regroup them by their corresponding climate.

[Charerg]

Moving forward, I have 34 possible temperatures combos once I exclude the tundra and ice caps.
With 22 precipitation possibilities this means 748 possibilities. How many do you have using you method?
If you have ideas to make it simpler, I'll be glad to hear.

I didn't really simplify the precipitations much beyound the obvious ones (like merging the f, s and w categories that are too rainy to be ever considered arid). Like you, I also had to include two separate categories for Am (Am/f and Am/s in my case). Furthermore, the Aw/f and As/f had to be split from the Aw/w and As/s groups, because while they would have been Aw/As if tropical, they would otherwise be considered f. Overall, I ended up with 34 precipitation categories (down from 8x8=64). With only 12 temperature groups though (+2 for ET and EF), I end up with just 408+2 possibilities, which wasn't too hard to manage since I used only 5 overall temp/prec tables (Hot, Warm, Temperate, Cold and Very Cold) to determine the extent of BW and BS. I listed in a prior post which temperature groups belong to each overall category.

Tbh, I don't think the precipitations can really be simplified a lot without risking some severe inaccuracies, especially with just 6 categories. I think you're much better off merging the temperature groups further if you want to simplify things, since many combinations have mean annual temps only a few degrees apart (and a 1 °C difference in Tann only changes the precipitation threshold by 20 mm, which is pretty minor).

Personally I chose to keep As as a category, because I don't think it adds too much extra complexity, and it could possibly be more widespread in some cases (like during the Jurassic, when the temperatures were hot enough to push today's Cs regions into the A category). Also, some Köppen maps show quite a bit of As in E. Brazil (this one, for example). That Brazilian As region is actually bit of a mystery, I haven't quite been able to figure out what causes it. Though that map probably exaggerates the extent somewhat, since it has some As->Cw transitions, which seem rather unlikely to be real.