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

[Charerg]

I don't know about the usefulness of having

0-5 and 5-10 m. I think 0-10 does a good job. When I look at the summer map, it's all desert except maybe in Israel and lower California, which are humid/steppe but they receive significantly more precipitations in winter.

 

Something like that maybe?

1. 0-10
2. 10-20
3. 20-40
4. 40-60
5. 60-100
6. 100-150
7. 150-200
8. 200+

And yes, I was talking about the instructions.
I had an idea. A new map to visualize precipitation distribution. The idea is that the world is divided in 3 zones. Rainy, dry and transition. Each zone (except the transition) has some sort of centre from which precipitations increase/decrease. It work a bit like a gradient.
of course, it's possible that this model doesn't work everywhere. In the northern latitudes, the polar front distributes precipitations differently.

Attachment 103992

I think I originally included the 0-5 mm category to help in identifying those ~0 mm rain in winter-cases of w climates, particularly in very high latitudes that have relatively low summer precipitations. But I agree, I don't think that was a necessary inclusion. Your suggested categories look very good. Though maybe the 2nd highest could be a bit rainier, would 160-240 and 240+ work better? That should help in identifying Am, I think (though maybe at the cost of making 100-160 mm too broad, not sure).

Also, the precipitation instruction does sound fairly interesting. Though tbh I'm kind of thinking that maybe the instructions for oceanic currents and atmospheric circulation should be more detailed, since people seem to struggle with those parts as well, and if you mess the currents and pressures up, there's no way the precipitations are going to be accurate.

[Azélor]

Since I don't see any holes in the climate map, the script probably works as intended (given that I just copy-pasted all those hexadecimal codes from your Excel file into the various climate zones).

Ok but since we lowered the number of temperature categories, almost 200 combinations don't exist anymore.

Yes, S climates are definitely too widespread. Blame Bishkek. Maybe W a little bit too but not as bad.
But actually, these area have a summer low precipitations. Some areas are close to the Mediterranean climate even if the difference is smaller. If I remember correctly, in 10-20 years from now, some area in Brittany will have a Csb climate, as well as some areas in Southern England.
The problem comes from when I start to doubt the numbers and go check city climate charts. Just looking at the lowest value or the Jan/Jul values is not a good indicator. You need to check the 3-4 months average, or just check the other mouth to avoid abberant data that is completely off the chart but just for one month.
We should go back to the old model then, regarding S and W.

About Brazil: the climate distribution is different. IRL the coast is drier, the west and south west and wetter and the hole is a bit drier.
But in general, it's a shift between Am and Af on the edges. You can try a different configuration and compare the results but I don't think it's a big concern, less than the other one

[Azélor]

Is there something special with 240 ml?

[Charerg]

Ok but since we lowered the number of temperature categories, almost 200 combinations don't exist anymore.

I did use the search function in Excel to eliminate the redundant stuff from the climate list.

Yes, S climates are definitely too widespread. Blame Bishkek. Maybe W a little bit too but not as bad.
But actually, these area have a summer low precipitations. Some areas are close to the Mediterranean climate even if the difference is smaller. If I remember correctly, in 10-20 years from now, some area in Brittany will have a Csb climate, as well as some areas in Southern England.
The problem comes from when I start to doubt the numbers and go check city climate charts. Just looking at the lowest value or the Jan/Jul values is not a good indicator. You need to check the 3-4 months average, or just check the other mouth to avoid abberant data that is completely off the chart but just for one month.
We should go back to the old model then, regarding S and W.

I don't think the w climates are a problem, their distribution looked pretty good too me. But yeah, I'll shift those two over 30% rain in summer categories back to f. It makes more sense since those regions don't really have a dry season in summer, even though they receive more precipitation in winter.

About Brazil: the climate distribution is different. IRL the coast is drier, the west and south west and wetter and the hole is a bit drier.
But in general, it's a shift between Am and Af on the edges. You can try a different configuration and compare the results but I don't think it's a big concern, less than the other one

Yeah I guess it's ok. I guess I'll check how the results would look with the Am/f categories shifted to Af/f (since I'll just have to shift a single precipitation combo to test that), but it's a pretty minor detail.

Is there something special with 240 ml?

No, but it would push the average precipitation for the 2nd highest category up to 200 mm. That should improve the resolution in regards to Am distribution, since you'd get more combos that produce an Am climate.

For example, if 2nd highest (150-200 mm) has 175mm avg. and the bordering sub-60mm categories (40-60 mm and 20-40 mm) have an average of 50 and 30 mm, the numbers would look as follows:

Pann:
(175+50)*6=1350mm
(175+30)*6=1230mm

With 200 mm avg:

(200+50)*6=1500mm
(200+30)*6=1380mm

Am thresholds:
25*(100-50)=1250 mm
25*(100-30)=1750 mm

So I guess it doesn't make a difference regarding Am distribution after all (unless using a higher average for 240+ than 200+). Never mind then.

[Azélor]

for future purposes:

I cleaned the spreadsheets and simplified them where I could.

To categorize what climate goes where.
When the hex code for climates is generated, it is reported to another sheet in a single column were it need to be trimmed down to 572 possibilities.
Then create a new sheet for each climate. paste the column.
Filter the column with the climate criteria, like in this example for Am which is simple

sheet22.png


and Af which is more complicated with 18 possibilities. Aw has 72. It is simple, just a lot of copy pasting.


sheet201.png

Damn I forgot about aridity! It was a bad idea then. If the model changes, the time it will take to update this is probably more than the time it takes to manually move the climates from one column to the other.

---

Trying the 8 categories in a new document :
There is now plenty of room for Am climates.
I don't always use exact values, should I?.
For example, it should only be possible to have Am climate if the dry month is over 4% of the yearly precipitations but I included places were it was 3,7%.

sheet23.png

[Charerg]

I think you have the Am climates off again. If we go by the official definition [P_ann > 25*(100-P_min)], there is no requirement about not having a dry season (in fact, the Am climates in India and other regions in SE Asia can have close to 0 mm rain in January). It just requires a high amount of annual precipitation (absolute minimum 1000 mm), and at least one month with less than 60 mm. Actually I'd guess that some Am climates receive more precipitation annually than the more dry Af climates.

As in my previous post, in this case the Am thresholds would be:

Driest month 50 mm: 25*(100-50): 1250 mm
Driest month 30 mm: 25*(100-30): 1750 mm
Driest month 15 mm: 25*(100-15): 2125 mm

And in the case that the driest month was 0mm, this would be the threshold:
Driest month 0 mm: 25*(100-0): 2500 mm

So, any A climate that receives more than 2500 mm annually (and has a month below 60 mm) is classified as Am. This is the case with the Am areas of India, like I mentioned. So, the following combos would be classified as Am:

Drowned (300 avg)+Moist (50 avg): 2100 mm (above 1250 mm threshold)
Drowned (300 avg)+Dry(30 avg): 1980 mm (above 1750 mm threshold)
Soaked (175 avg)+Moist (50 avg): 1350 mm (above 1250 mm threshold)

I'd suggest sticking to the official definition in order to avoid numerous weird Af->Aw and Am->BS transitions.

Also, s climates have the following requirement (if we go by the Kottek et al. definitions of the climates, Köppen himself used 30 mm according to wikipedia):

• Driest summer month must be below 40 mm

So those s climates that have a moist (40-60 mm) summer or rainier should actually be f.

Edit: the Climate Map

I finished shuffling those anomalous s combos over to f. I also tested moving the Am/f prec combo (200+ and 100-50) over to Af, since I think that's closer to Af than Am (though it's easy enough to change it back if you want).

genClimates_test2.png

Overall I think it's pretty good now. Brittany still stays as Cs, since the source maps (in this post) have the region receiving 100-200 mm in winter and 25-50 mm in summer, but that's probably as accurate as we're going to get using 6 precipitation categories.

[Azélor]

Are you referring to this paper?

http://koeppen-geiger.vu-wien.ac.at/pdf/Paper_2006.pdf

Hum, I'm mostly using the Wikiepedia pages. I'm not sure if they are invalid or just using other references.
Well, it comes from a book : Physical Geography: A Landscape Appreciation
but I'm not sure it is related to Koppen.

If we assume that the Wikipedia pages are not good, does it mean I have other data that needs to be changed?

[Azélor]

and at least one month with less than 60 mm.

In the table, the moist category is 40-60ml. It need at least a moth is this category. In our model, it conveniently fall either in July or January. Anything with a dry season above moist cannot be classified Am.

I redid the table;
It should be fine now
sheet24.png

[Charerg]

Are you referring to this paper?

http://koeppen-geiger.vu-wien.ac.at/pdf/Paper_2006.pdf

Hum, I'm mostly using the Wikiepedia pages. I'm not sure if they are invalid or just using other references.
Well, it comes from a book : Physical Geography: A Landscape Appreciation
but I'm not sure it is related to Koppen.

If we assume that the Wikipedia pages are not good, does it mean I have other data that needs to be changed?

Well, most of the wiki pages have references to that paper or other similar ones, so they should be ok for the most part (I think, though I haven't checked). But yes, I've been using that paper as the reference for the climate definitions. Although ofc we use 0 °C Isotherm (instead of -3 °C) as the C/D boundary since we don't have a -3 to 0 °C temp category.

And yeah, the Am distribution looks good now.

Edit:
Btw, is the Af/Am distribution better in the latest test of the script, or should I change it back to what it was?

[Azélor]

Well, most of the wiki pages have references to that paper or other similar ones, so they should be ok for the most part (I think, though I haven't checked). But yes, I've been using that paper as the reference for the climate definitions. Although ofc we use 0 °C Isotherm (instead of -3 °C) as the C/D boundary since we don't have a -3 to 0 °C temp category.

And yeah, the Am distribution looks good now.

As far as I am aware, the 0 vs -3 is not something climatologist agree on. I picked 0 because (Pixie probably) it makes more sense since it's the freezing point.

Not a big thing so I flipped them over to f.

sheet25.png

Oh Ignore the cell in the top. It is still purple.

sheet26.png