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

[Azélor]

So you end up with 20*T_ann+x formula.

I end up with the same result but it's calculated differently.

[Charerg]

I end up with the same result but it's calculated differently.

I think the formula itself is probably correct, if you end up with 20*T_ann+280 for the "W category", and so forth.

The point I was making was that I don't recommend doing the following simplification:

Hey, I got some numbers!

Replacing this :

• If less than 30% of annual precipitation occurs in the summer : Annual precipitation (mm) < 20 × average annual temperature (°C)
• If more than 70 % of annual precipitation occurs in the summer: Annual precipitation (mm) < 20 × average annual temperature + 280
• Else : Annual precipitation (mm) < 20 × average annual temperature + 140

o If annual precipitation is < 50 % of the threshold = BW: desert climate
o If annual precipitation is between 50 and 100 % = BS: steppe climate

by this:

if the annual precipitation (in centimetres)

are Greater than R= humid
are Smaller than R but greater than R/2= semi-arid
are Smaller than R/2= arid

R=2 x T if rainfall occurs mainly in the cold season (s=summer dry)
R=2 x T + 14 if rainfall is evenly distributed throughout the year (f)
R=2 x T + 28 if rainfall occurs mainly in the hot season. (w= winter dry)
(T= mean annual temperature)

Because if you replace the "aridity categories" with the climate categories, you're inevitably going to end up with results differing from real köppen maps, because you're essentially using a different definition for the B climates. And in any case I don't think it's a necessary simplification. It does require an additional table though, in addition to the climate tables for A and C/D climate groups.

Although if you're aware of this then I've been essentially ranting about nothing, but just making sure so you don't end up overlooking this, since it's a pretty significant detail.

[Azélor]

Although you're right in that wikipedia does not define exactly what the "mainly in summer/winter" means. Although the B climate class page in wikipedia does provide that information (in percentages).

I don't get your point at all.

f you replace the "aridity categories" with the climate categories,

aridity = s,f,w ?
climate = A,B,Ca,Cb,Cc,Da,Db,Dc,Dd,E?

I think I understand what you mean now?
B and A climates use the aridity threshold of s,f,w depending where they are in the table, They use the same thresholds as the C and D climates,
for example, despite being As, some will still use the f threshold

[Charerg]

I don't get your point at all.



aridity = s,f,w ?
climate = A,B,Ca,Cb,Cc,Da,Db,Dc,Dd,E?

I think I understand what you mean now?
B and A climates use the aridity threshold of s,f,w depending where they are in the table, They use the same thresholds as the C and D climates,
for example, despite being As, some will still use the f threshold

No, let me try and rephrase that.

By "climate category" I refer to s, f and w as they are defined per the climates. So, the following familiar definition:

• At least 10/11 (or ~91%) of rain occurs in summer (w climate, winter dry)
• Between 1/4 to 10/11 (26-90%) of rain in summer (f climate, no dry season)
• Less than 1/4 (25%) of rain falls in summer (s climate, summer dry)

Of course, those are different for the A climates, but let us ignore those for now (also, I'm aware in reality it's a comparison between rainiest and driest months, but in our case the above applies since there is only data from two).


With the "aridity category/group", I refer to the categories that are used to define which aridity threshold to use:

• At least 2/3 (or ~67%) of rain occurs in summer (W aridity group, use threshold T_ann*20+280)
• Between 1/3 to 2/3 (34-66%) of rain in summer (F aridity group, use threshold T_ann*20+140)
• Less than 1/3 (~33%) of rain falls in summer (S aridity group, use threshold T_ann*20)

I hope the definitions are clear now. The point I was making is that there is a pretty big difference between the two definitions. So, if you use the same criteria to define both the climate group (f, w, s) and the aridity category (which defines which threshold to use), the aridity groups would look as follows:

• At least 10/11 (or ~91%) of rain occurs in summer (W aridity group, use threshold T_ann*20+280)
• Between 1/4 to 10/11 (26-90%) of rain in summer (F aridity group, use threshold T_ann*20+140)
• Less than 1/4 (25%) of rain falls in summer (S aridity group, use threshold T_ann*20)

I'm not sure if you've done the latter, but if that is the case, then this will cause a different distribution of B climates compared to official Köppen maps, since the definition of which aridity threshold is applied is drastically different from the one usually used with the Köppen system. Again, just making sure that you're aware that there is a pretty big gap between the definitions. Of course, you might already be on top of things.

[Azélor]

I wrote this before reading your last reply.

Back on the spreadsheet:
I've looked at some of the wrong climates to find out that Excel is right and did not miscalculate but I did mistakes. In the latest version of the script, 2 precipitations cells have the s threshold instead of the f they should have. Therefore, the aridity for them is wrong in some cases. To put things differently, Europe should be less arid.

That is what I've found so far.

Which brings me to the other point: the actual threshold system is bad. Some places are too dry other too wet and if you move from s to f, the aridity increase abruptly, faster than the actual increase in precipitations. Using a 3 thresholds is not enough, we need more in order to have something more gradual.

We could assing a maximum and minimum value. They would be located in top right and bottom left.
Everything in between gets a value depending on the % precipitations falling in winter for example.

Since the highest value is 1080 (or something, its just an example) if 100% ( or maybe a slightly lower % to fit or categories) of the precipitations fall in summer. The lowest is 760 if 0% or so fall in summer.

It would make s climates more likely to be arid. F on the bottom left might be more wet if I get it right, resulting in a smoother transition. Less steppes in Europe ?

And f climates on the top right would be more arid.

At least I hope I get this right. I know it is a good idea but I often get the resoning behind the threshold wrong.

[Charerg]

I wrote this before reading your last reply.

Back on the spreadsheet:
I've looked at some of the wrong climates to find out that Excel is right and did not miscalculate but I did mistakes. In the latest version of the script, 2 precipitations cells have the s threshold instead of the f they should have. Therefore, the aridity for them is wrong in some cases. To put things differently, Europe should be less arid.

Haven't read your whole post yet but just a quick comment: remember the S threshold is actually the lowest (defined as 20*T_ann). So, if moving from S to F, then the threshold would become higher (F being defined as 20*T_ann+140), and the climate therefore more likely to be considered arid. Sorry if I'm repeating myself a bit here.

Maybe replacing the threshold with a more gradual "x factor" (20*T_ann+x) that increases based on the percentage that falls in summer would work, but it could be difficult to determine the maximum value. Should the "x factor" be 280 mm at 100% falls in summer? In that case most areas would be less prone to aridity than is normally the case with köppen maps. And if it should be higher, then what would be the correct value?

All in all, it's perhaps easier to stick to the standard definition. That way, at least it's easy to make comparisons between generated maps and actual Köppen maps of Earth. Things are never going to be 100% accurate when working with data from just two months and a discrete number of categories that use average values instead of exact values.

[Azélor]

Basically you are saying that for instance, f climates don't use just 1 threshold but up to three different ones depending on the precipitation distribution during the year? I assuming it makes things smoother. But is it smooth enough.

I'm not sure it solves the jumping aridity problem.

[Azélor]

Haven't read your whole post yet but just a quick comment: remember the S threshold is actually the lowest (defined as 20*T_ann). So, if moving from S to F, then the threshold would become higher (F being defined as 20*T_ann+140), and the climate therefore more likely to be considered arid. Sorry if I'm repeating myself a bit here.

Good I got it right this time.

[Charerg]

Basically you are saying that for instance, f climates don't use just 1 threshold but up to three different ones depending on the precipitation distribution during the year? I assuming it makes things smoother. But is it smooth enough.

I'm not sure it solves the jumping aridity problem.

Yes, that is it exactly. I don't know if it solves any problems per se, but in any case that is exactly how it works in the real-world Köppen maps we're using as references. It could have an effect in southern Europe for example, though: many of the f climates that are borderline s would undoubtedly use the lowest aridity threshold (S) instead of the higher F threshold.

[Azélor]

Ok now I understand we you said the letters could be confusing. Indeed using the same letters for

• At least 2/3 (or ~67%) of rain occurs in summer (W aridity group, use threshold T_ann*20+280)
• Between 1/3 to 2/3 (34-66%) of rain in summer (F aridity group, use threshold T_ann*20+140)
• Less than 1/3 (~33%) of rain falls in summer (S aridity group, use threshold T_ann*20)

as s,f,and w climates, doesn't make much sense since they are 2 different things.
I guess I'm just repeating you.

I personally don't really care how Koppen name them, if it's too confusing we should change the naming, assuming we keep them.

Arid
Average
Wet

West coast
Something (Faoofa?)
East coast (the dry summer is more prominent on east coasts, or the eastern sides of mountains in the case of Central Asia). it is the opposite for w pattern.
Now I remember I used to say precipitation pattern. For example, Brest in Brittany has a Cfb climate with a s pattern because the summer is drier than winter but not dry enough to qualify as a s climate.