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

[acrosome]

Ah, as I had thought. Though it does seem odd to have a stark border between a precip 5 area and a precip 0 one...

Thanks.

[Coggleton]

So I've started to look at using Azure's script for the holdridge biomes, and I'm wondering if there's an issue with it coloring in oceans correctly- for the below picture, oceans are colored as (107, 165, 210) while in the output file I passed it should have oceans as (155, 205, 230). Below are the command line, resulting map, and actualOutputProfile - what's interesting is that everything else besides the oceans used the correct color.

Function call:
> py ./skcc.py --tempns="suru jul temp.png" --tempnw="suru jan temp.png" --precns="suru jul prec.png" --precnw="suru jan prec.png" --tempprof="actualTempProfile" --precprof="actualPrecProfile" --outprof="actualOutputProfile" --outfile="suruClimateAzure.png"

Output Profile (Pasted as a text file, I have a file on my computer without the .txt so it ran well):
actualOutputProfile.txt

Result:


I also ran it using the holdridge mode, and same issue: everything but the oceans conformed to the output profile.

Function call:
>py ./skcc.py --tempns="suru jul temp.png" --tempnw="suru jan temp.png" --precns="suru jul prec.png" --precnw="suru jan prec.png" --tempprof="actualTempProfile" --precprof="actualPrecProfile" --mode=holdridge --outprof="actualHoldridgeOutputProfile" --outfile="suruBiomeAzure.png"

Output Profile:
actualOutputProfile.txt

Result:


Once Charerg's climate script runs on my (actual scale) map, I'll plan on posting both outputs to see how they compare in case someone finds it useful/interesting.

[AzureWings]

You are correct, thanks for pointing that out! When I backported improved and reorganized I/O file handling code from my temperature project I overlooked the fact that I'd changed a few keywords. I've updated the output profile loading code to properly recognize the 'Ocean' keyword (it had been replaced with 'Ignored').

[Coggleton]

Glad I could help point that out! And as promised, here are the results for those interested(both scripts had the same color profile and map inputs; the only thing changed was the ocean color to help make some zones more visually distinct, as well as the legends being added):

Charerg's Climate Script:


Azure's Climate Script:


Azure's Biome Script:


It seems like Charerg's script tends to result in drier climates being preferred over others, such as in a couple of areas where As and BSh border each other (and a rare sighting of Dwa in Azure's output as a side note). Though a question for Azure: how much does your script differ procedurally from mbartelsm's biome guide? If not at all, then because mbartelsm's relies on the temperature/precipitation maps and only uses the climate maps to adjust tropical forests I would expect this to have minimal impact on biome placement.

[AzureWings]

So I don't know how the color categories in mbartelsm's guide work out in comparison since my script just crunches raw numbers under the hood (that way if you have input data with more/fewer/different category values you can use it as an input still via a custom color profile), but basically it takes the same temperature/precipitation input data, uses the temperature input to approximate a biotemperature by averaging the summer and winter temperatures interpolated across points on a hardcoded sine wave, and then uses the biotemperature approximation and the precipitation input to index into the Holdridge biome grid/triangle. Warm temperate and subtropical are arbitrarily split at a biotemperature of 17 degrees C. The default Holdridge output profile follows mbartelsm's output colors but you could give each individual biome hex in the triangle a distinct color if you wanted with a custom profile. The script doesn't do anything at all with the climate map for biomes (indeed it doesn't have any way to take one as an input; and it doesn't compute the Köppen category internally when doing biome classification either).

In terms of output differences on the biome end, they'd most likely be a result of the biotemperature approximation step.

With respect to Charerg's script and arid regions, we might be using slightly different calculations to determine the aridity threshold/evapotranspiration term (I think we had some discussion in the thread about that a while back). I think I preferred the notion of a continuous model but for whatever reason I left my script doing the piecewise version.

[Charerg]

As Azure points out, the most recent iteration of my Gimp script uses a gradual aridity threshold, which causes a slight difference regarding which areas are considered arid, when compared to the traditional Köppen formula for determining the aridity threshold.

[acrosome]

Where can I find AzureWings' biome script?

[AzureWings]

It's the same script as my climate one, just add "--mode=holdridge" to the command-line invocation (and there are different valid categories for custom output profiles if you want to use those).

If you meant the script in general, you can download it from Github at the link Azélor put in the first post of the thread.

[acrosome]

Awesome, thanks. I guess I have to download the new version. Mine is quite old.

And while I'm at it, I'm sure that you have heard it before, but thank you for this excellent tool.

[acrosome]

EDIT-- FYI, I'm going to post this separately, now, since no one seems to be answering. I guess you guys have developed lives outside of the guild or something...

Howdy. I'm looking for advice on how to handle the rainshadows of large mountain ranges.







The issue tends to be the huge ranges on Ishtar and the shield volcanoes on Atlu. Does this look right? Bear in mind that much of eastern Ishtar is pretty dry to begin with. Any pointers about any other problems you see with these is welcome, too, of course. I'm much worse at precipitation than I am with temperature maps. And regarding that, just in case you are interested:





And here is the climate output: