Theia revisited - Climate check (WIP)

[Azélor]

Oceanic current: the dynamic looks good but the temperatures should be colder.
You mid latitude current are partially running under the ice sheets. I would consider then cold instead of mild.
So currents flowing toward the equator are also colder.
I think generally, this will make most of the currents colder.


About pressure:

Your maps lack high pressure systems over the poles.
Since you planet is covered with thick layers of ice in the polar regions, it has a constant high pressure system there. It does weaken in summer but would still, be strong.
It acts like a convection movement, sucking the air from the high atmosphere downward on the pole and then push it toward the equator while being deflected toward the east because of the Coriolis force.

For the rest sadly, there is a lot of guesswork.
You low pressure systems over the ocean in mid-high latitude would shift closer to the equator. Probably around beyond the edges of the ice sheets.
The water being warmer than the ice surface, it can create a big difference in pressure and strong winds like on the southern seas on Earth.
I also think it would shrink in size horizontally, covering a smaller area latitude-wise.

The subtropical high would likely also be pushed toward the equator and also shrink in size.

[nwisth]

Thanks for helping me out here, Azélor! I really appreciate it.

Regarding the exoplanet YBP 1194b, you are right - it can't be a gas giant, can it? It's probably a huge ball of molten rock and metal. For my novel's purposes, it's a weekly astronomical event that serves a few purposes. One is making a week seven days - on a world with no moon but a ring, the equinoxes and solstices are the great, visible indicators of the year's passage, and noticing that the sun dims twelve times between each one makes it easy to split a quarter-year into three four-week months - even without a lunar cycle to back it up. The other reason is religious - to give the people something omnious in the sky, to associate with darkness , cold and evil. With six gods to name the other days of the week, the last one is reserved as Shadowday (or Shadday, as time passes and words get cut down). Since all the other planets of the system are brazenly made up - including Theia, of course - I'll happily hand-wave the details of its size, speed and distance to make it able to play its roles.

What I said about a few billion years and tectonic activity was simply a comment on how the Worldbuilder program doesn't move tectonic plates around in several cycles. On Earth, mountain ranges have risen up and been eroded down several times, our rocks churned and mixed across deep time. My Theia heigh map doesn't have that luxury, due to the limitations of the software - but I still imagine it to be just a little younger than the Earth, and to have gone through pretty similar tectonic processes. The star YBP 1194 is thought to be .5 billion years younger than the Sun, but as wikipedia says, the error bar is high due to the distance, and could be off by as much as 1.6 billion years. I will definitely give Gplates a look! I bet that could texture up my continental plateaus nicely.

As an interesting side-note, as I re-read the YBP 1194 article now, I noticed that the star is placed in a cluster, with over 500 other stars within a distance of 10 lightyears. We only have 17 stars that close, and about 134 stars within 20 lightyears. I had figured that the Ring would blot out most stars, but with 500 of them really close, they ought to be bright enough to actually give Theia a pretty spectacular night sky - especially at midnight, when the worldshadow obfuscates a large part of the Ring.

Thank you so much for the modified currrent map! You are completely right, of course - the water would be a lot colder. Working with the ice sheet layer turned off made me a bit blind to its effects. I will use it to re-create my temperature maps, and also try to find a way to make the temperature bands of the permanently ice-covered regions to shift (my Select/Color Range process made it hard to modify the hard line between blue/white and red/yellow heat map gradients from Worldbuilder).

I will also re-make the pressure maps - and as a result, the precipitation maps as well. Guesswork will still have to be made - but with your invaluable input, Azélor, the guesswork will be a lot more educated this second time around.

-Niels

[nwisth]

So, today I thought I'd just give the script a quick go before I went back and reworked my source maps. Turns out, it wasn't as easy as I believed!

On my next try, I'll make sure to make my brushes as hard as possible. Even at 100% hardness, the feathered edges of brushes caused a lot of ugly pixel stripes. If I want the map to look good as well as be precise, I'll need to do a lot better job painting the precipitation!

I also got a lot of empty areas - marked by black on my recolored map (it was hard to tell the difference between the areas, so I used Select/Color Range to pick out the different climates and recolored them manually to the Köppen standard from wikipedia):



It seems that the black areas are where I left the precipitation map transparent. Does anyone know how I can stop that from happening? I tried using black, as well as a yellow color in case my height map color shining through would confuse things, but the same thing happened.

Also, are there any updated photoshop scripts/actions that output the standard Köppen colors? I tried recoloring the "koppen climate" layer before running the actions, but that didn't work.

-Niels

[Charerg]

Which script did you use (there are several available in the thread, and they're not identical)?

The more up-to-date script by AzureWings takes the .png files as input and outputs the climate in the standard Köppen colours (it doesn't require either PhotoShop or GIMP, being a command line prompt).

[nwisth]

Wow, a command line prompt sounds awesome! I used the set of photoshop actions and PSD file in Step 7, linked from Azélor's first post. I saw the link to a script by AzureWings, but it said GIMP so I left it alone. Is this the one you mean?

https://github.com/PHarvey7/speculative-koppen

-Niels

[Azélor]

I did some math and I think that the closest star is at 1,25 AU from the star. Assuming I got the math right and that all stars are spread out evenly.
That is still pretty far. There are no impacts on the climate at least that is sure.
Interesting enough, they may be in a cluster but when you are inside, you have no idea of this until you have more advanced technology.

The link to the script is on the first page of the tutorial.

[Charerg]

Wow, a command line prompt sounds awesome! I used the set of photoshop actions and PSD file in Step 7, linked from Azélor's first post. I saw the link to a script by AzureWings, but it said GIMP so I left it alone. Is this the one you mean?

https://github.com/PHarvey7/speculative-koppen

-Niels

Yes, that's the one.

Though I'd advice getting Notepad++ so you can look at the code yourself and check what RGB values the script uses for the various temperature and precipitation categories (you can also easily edit the RGB values or even add or remove categories if you want).

[nwisth]

Ok, wow - that was an uphill struggle! I've remade temperature and precipitation maps, then managed to install Python and even got the script to run (after a lot of Google Fu) - but then saw I have to change the colors. I guess new precipitation maps are in order.

At least I made these two new temp maps based on your feedback, and since I'd used masks and full color fields they were easy to modify so they would work.

Summer:


Winter:



New precipitation maps will have to wait - I'll make them to follow these color values:
(Heaviest) 135, 0, 180) :300
130, 60, 200) :170
105, 70, 200) :105
60, 60, 180) :55
70, 95, 150) :30
55, 85, 100) :15
35, 50, 50) :7.5
(Driest) 20, 20, 20) :2.5
Ocean 107, 165, 210) :O

-Niels

[AzureWings]

Sorry to hear it was a rough setup - my directions do assume a bit of Python and command-line familiarity (I'm guessing installing the Pillow dependency was one of the stumbling blocks?). Anything in particular that took a lot to figure out? I could add some tips or an FAQ section to the readme for the script.

As for the map colors the script is designed to let you write up a little text file (like in the format you duplicated in your post for precipitation values, which looks like my default precipitation color profile) and use the optional arguments (--tempprof=<filename> and/or --precprof=<filename>) to specify those custom files that define different color categories and what value (for temperature/precipitation) each color corresponds to, instead of using the color categories for input that the script defaults to. That way you wouldn't need to change the colors of the different precipitation/temperature bands of your existing maps (if I'm understanding what you mean correctly). The one issue here is that the 'transparent category' for precipitation can't be done in this way (yet; I think I can and should fix that).

You can also make a color profile for the output too (the repository includes a couple samples, including the one corresponding to what the script uses by default), if you wanted a specific color scheme for the Köppen categories (then use the file by running the script with the optional --outprof=<filename> argument).

[nwisth]

Hi! Again, thanks for the help, guys. :)

This morning I've spent four hours hunting pixels and struggling with color tables. I still can't get the script to work, always ending up with:
Error: Invalid temperature map color value: (x, x, x) or Error: Invalid precipitation map color value: (x, x, x)

I've gone through each of the four maps time and time again, hunting for those pesky anti-aliased lines and trying to eliminate them with the pencil tool - up to the point where I can't find a single pixel not of the right color.

I've tried converting the PNGs to Indexed Color, specifying only the correct color values, to try and force the diverging pixels out - but the script won't accept indexed color files, so I have to convert them back to RGB. The incorrect coloreded pixels did not return, but I still get the errors.

I've tried adding the erroneous color values to precprof and tempprof files, setting the same value for temperature and precipiation as the color it is a variant of, and that makes the error for that specific color go away - but so far it is always replaced by a new one. Now my precprof list has 40 entries and the tempprof 21. I keep thinking "just one more color" - but now I really, really need to start working on my paid tasks before my clients get too impatient, so I'm waving the white flag for today.

How do you guys make this work? It seems so effortless in your posts. I've added the files in case you can make heads or tails of it.

JanPrecip


JulPrecip


JanTemp


JulTemp


precprof
precprof-colors.txt

tempprof
tempprof-colors.txt


I guess it would have been faster to do it manually...

At least I've learned a few things:
-Work using colored layers and masks - that way I can use LEVELS to reduce the anti-aliasing to zero.
-Always paint with the PENCIL tool
-Keep colors correct at all times.

-Niels