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

[acrosome]

I think I'm ready for some criticism on my climate work, now, if any of the gurus can chime in...

[acrosome]

I'm getting a puzzling error when I attempt to run AzureWings' Python script:

dean@deans-thelio:~/speculative-koppen-master$ python ./skcc.py --tempns="./JulTemp.png" --tempnw="./JanTemp.png" --precns="./JulPrecip.png" --precnw="./JanPrecip.png" --outfile="output.png"
Error: Invalid precipitation map color value: (107, 165, 210)

But 107,165,210 is the color defined as ocean in the readme files. Any idea what is up, here? I can't just leave the ocean blank, either, since it says 0,0,0 is an invalid color too.

[AzureWings]

I could probably make that bit a hair more robust... but offhand the first thing I'd check is whether one or both of your precipitation maps has one or more pixels of ocean in pixels where your temperature maps have land (current script version assumes the land pixels will line up between temp and precipitation maps, and so checks to see if a pixel is ocean only by looking at the temp maps. It's an oversight - thanks for causing me to notice it, even if it turns out not to be the cause of your issue - I'll release a fix when I'm not in dire need of getting to sleep).

[acrosome]

Whoa. Ok, I'll have to figure out how I can find such a problem pixel in Gimp.

I guess I can select all land on my land mask and erase everything else, on all four maps...

EDIT-- After a LOT of pixel-hunting, I got it to work!

[AzureWings]

I've released a bugfix for my script in the case where one or both precipitation maps have ocean in pixels where the temperature maps do not. The version of the script on the Github repository (see the link in the first post of the thread) should be fixed now; the former behavior of the script is that pixels where either temperature map had ocean were treated as ocean. Now the script behaves as intended in that regard; if any of the four input maps have ocean for a pixel, it is treated as ocean for all four (my intent here was to make it easier to use source maps that might have slightly different coastlines, such as might occur for high-detail maps where temperature and precipitation were generated separately and/or not by hand).

And for GIMP - try, for each precipitation map, selecting all land on your land mask and then with that selected try the "intersection" selection type (I think it's the last of the four selection types in the row of select types for the different select tools) with the select-by-color tool (with anti-aliasing, feather, etc. off and a threshold of 0) on the ocean color. If you're left with any pixels after doing so, they were ocean-colored pixels but were inside the land mask.

(Sorry if you're already plenty familiar with the selection types.)

[acrosome]

And for GIMP - try, for each precipitation map, selecting all land on your land mask and then with that selected try the "intersection" selection type (I think it's the last of the four selection types in the row of select types for the different select tools) with the select-by-color tool (with anti-aliasing, feather, etc. off and a threshold of 0) on the ocean color. If you're left with any pixels after doing so, they were ocean-colored pixels but were inside the land mask.

(Sorry if you're already plenty familiar with the selection types.)

Yes, that's what I did. It turns out that I had a lot of off-color pixels. Somehow. I'm guessing that at some point I used a brush instead of a pencil. I also started painting a sample of the color listed in the error on the map, then selecting that color to find the other pixels.

[Naima]

Step 4, winds:

Use the pictures to figure out how the winds are blowing. Figure A
Or use the main map at the bottom at the page.
All the figures are from the North Hemisphere except E.

Attachment 76626


Dominant winds

• Near the equator, the dominant winds are usually blowing to the west (Trade Winds)
• In mid latitudes, it’s blowing to the east (Westerlies)
• And close to the poles they are blowing to the west again (Polar Easterlies)


Wind usually flow from the highest pressure to the lowest. The larger the difference in pressure between two areas, the stronger the winds will be.
Inside a large high or low pressure zone, the winds can be very weak. Figure B

Mid latitudes winds: Starting with your low pressure bands at high latitudes, the Westerlies should blow from west to east where it’s blue, and where there are no colors. Avoid the red.

Low pressure have 2 types:

• Hot season: converge like the ITCZ in Asia, see figure F
• And those of the mid and high latitudes, the North Pacific (round , isolated) and the North Atlantic (continuous band of low pressure) Figure G

The high pressure systems at mid latitudes (also called subtropical highs):

• Draw the poleward winds first. They have a curved shaped because they quickly change direction when encountering the Westerlies. Figure C
• Winds blowing from the equatorward side of the system tend to blow toward the equator, or if any, toward the closest low pressure center.
Winds are converging near the equator; they tend to blow to the west. Figure D

Polar highs: The high pressure systems on the poles brings dry and extremely cold temperatures. Figure E (South Pole)

Overland highs are spinning according to the schema. See Figure A, or look at Eastern Asia in winter.

January winds

Attachment 76627

July winds

Attachment 76628

---

Due to polar cell and coriolis effect shouldn't the wind go from east to west around the polar region ?

[Naima]

Also another question I noticed that winds and sea currents do not always follow the same direction while building the world , is this normal?

[antillies]

Been a few months since someone has posted here so allow me to break the silence.

I'm taking a hard look at automating the entire process with as little input from the user as possible. One question I have about currents though (and Charerg, I'd appreciate your input on this as well): for the warm current that flows north past the Philippines, why does it continue north/northeast? Why doesn't it instead flow toward the Chinese coast? Are the winds already beginning to affect it at that point? Or is it a matter of where the continental shelf is? The latter would confuse me further because (at least according to this map) there is a current, albeit a weaker one, that flows adjacent to the coast.

[Charerg]

Been a few months since someone has posted here so allow me to break the silence.

I'm taking a hard look at automating the entire process with as little input from the user as possible. One question I have about currents though (and Charerg, I'd appreciate your input on this as well): for the warm current that flows north past the Philippines, why does it continue north/northeast? Why doesn't it instead flow toward the Chinese coast? Are the winds already beginning to affect it at that point? Or is it a matter of where the continental shelf is? The latter would confuse me further because (at least according to this map) there is a current, albeit a weaker one, that flows adjacent to the coast.

The oceanic circulation always forms a loop, and since there is water "draining east" around the 45°, it follows that water must flow south->north to replace it. The winds would also be a factor. Also, due to that mass of water having been deflected northwards by the Philippines, it already has some momentum behind it in that direction.

It's not dependant on the continental shelf, as these are surface currents (and indeed primarily tied to atmospheric circulation, ie. wind patterns).

Edit:
I guess the main takeaway when it comes to automating the process is identifying bodies of water that would contain a "closed loop", since it would be good to replicate the circular pattern of real oceanic currents. Most smaller bodies of water and inlets/bays could mostly be ignored as they wouldn't make that big of a difference when it comes to the global climate. I made a sort-of simplified example based on the 1943 map of the North Atlantic circulation a while back (might have posted this already), that might be helpful in identifying the patterns: