[WIP] Building a world from tectonics onward

[MrBragg]

Apologies for the thread spam, this is probably (hopefully!) the last one for a while. Anyway, here's an attempt at pressures and winds for Jan and July. There are quite a few places where I'm not sure if I'm doing things correctly, so please don't be shy with comments / criticism!

Jan


July

[Naima]

Old post ... delete plz.

[MrBragg]

Congratz on finishing the topograhy, MrBragg! I think you must be the first one to have finished a planetary-scale elevation map built with this kind of technique and detail (makes me think I should get off my lazy arse and get some progress done on my own project).

Thanks, Charerg! I guess being stubborn does actually have some advantages Would also love to see your latest map since it was one of the big inspirations for me.

1, I've done some tests lately with exoplasim runs starting with 1 meter of initial global glacial cover; it takes a bit longer for the climate to settle out, at least 50 years whereas they usually are mostly settled within 20 years otherwise, but the resulting glacial cover is a lot more accurate, as well as the global average temperature. it does make some mid-latitude areas even colder than they should be, though.

So I tried this at T42 and starting with glaciers did converge to a marginally cooler overall temperature at 300 ppm CO2 (22.4 C vs 22.9 C). Starting with <=200 ppm CO2 resulted in a snowball, though, so the hysteresis caused by starting with glaciers may constrain the accessible final temperatures somewhat.

On a somewhat related note, @worldbuilding pasta, do you have any experience running resolutions >T85? For giggles I tried everything larger and each and every one immediately crashed before the model even got going, even when using an identical input from T85 (only resolution and sra inputs changed) that worked fine.

[worldbuilding pasta]

So I tried this at T42 and starting with glaciers did converge to a marginally cooler overall temperature at 300 ppm CO2 (22.4 C vs 22.9 C). Starting with <=200 ppm CO2 resulted in a snowball, though, so the hysteresis caused by starting with glaciers may constrain the accessible final temperatures somewhat.

You might try running it first at 300 ppm for 10-20 years and then reduce co2

On a somewhat related note, @worldbuilding pasta, do you have any experience running resolutions >T85? For giggles I tried everything larger and each and every one immediately crashed before the model even got going, even when using an identical input from T85 (only resolution and sra inputs changed) that worked fine.

I haven't even done a proper T85 run yet, I wanna test it more but I've sort of paused my exoplasim work for the winter because of the energy costs here. It might be a ram limitation or something, it might need more layers, or it might be the same bug people have been having with t63 that I still have yet to investigate.

[MrBragg]

You might try running it first at 300 ppm for 10-20 years and then reduce co2

Yeah, if I were super ambitious I could try that, though my results at ~120 ppm CO2 I think are already decent enough to move forward. I've already baked my ice caps into the terrain anyway, so as long as the EPS output gets close I'm happy.

It might be a ram limitation or something, it might need more layers, or it might be the same bug people have been having with t63 that I still have yet to investigate.

Ok cool, those were my first thoughts as well. Going to 15 layers didn't affect it and the error I got was different than when I tried T63, which very clearly told me that that resolution wasn't supported. For memory, EPS at T42 doesn't seem to use all that much RAM (I have 64 GB) and I can run a T85 and a T42 simultaneously, so assuming memory usage scaled ~linearly then I'm imagine at least T106 should be able to run. Next time I run a T85 I'll pay closer attention to the memory usage during the cycles and make sure it never spikes higher than I'd expect.

I haven't even done a proper T85 run yet, I wanna test it more but I've sort of paused my exoplasim work for the winter because of the energy costs here.

If it's at all helpful, I could run some calculations for you when my computer is open, so just let me know.


EDIT: OK so I did some more digging and for T106 and T127 it looks like EPS is missing files it needs for those resolutions. For T106 for example, it starts out with the message "/home/users/.local/lib/python3.6/site-packages/exoplasim/plasim/run/most_plasim_t106_l10_p16.x...." and then crashes; for T85 and T42 this step passes almost instantly and then the model begins. It turns out that the t106_l10_p16 (resolution_levels_processors I'm guessing is the nomenclature) file isn't present in the requested directory and neither are any files associated with T127. So, unsure if this is an issue on my end or with EPS, but figured I'd post it here in case it's useful.

[Naima]

Yeah, if I were super ambitious I could try that, though my results at ~120 ppm CO2 I think are already decent enough to move forward. I've already baked my ice caps into the terrain anyway, so as long as the EPS output gets close I'm happy.





Ok cool, those were my first thoughts as well. Going to 15 layers didn't affect it and the error I got was different than when I tried T63, which very clearly told me that that resolution wasn't supported. For memory, EPS at T42 doesn't seem to use all that much RAM (I have 64 GB) and I can run a T85 and a T42 simultaneously, so assuming memory usage scaled ~linearly then I'm imagine at least T106 should be able to run. Next time I run a T85 I'll pay closer attention to the memory usage during the cycles and make sure it never spikes higher than I'd expect.




If it's at all helpful, I could run some calculations for you when my computer is open, so just let me know.


EDIT: OK so I did some more digging and for T106 and T127 it looks like EPS is missing files it needs for those resolutions. For T106 for example, it starts out with the message "/home/users/.local/lib/python3.6/site-packages/exoplasim/plasim/run/most_plasim_t106_l10_p16.x...." and then crashes; for T85 and T42 this step passes almost instantly and then the model begins. It turns out that the t106_l10_p16 (resolution_levels_processors I'm guessing is the nomenclature) file isn't present in the requested directory and neither are any files associated with T127. So, unsure if this is an issue on my end or with EPS, but figured I'd post it here in case it's useful.

What pc specs do you have?

[worldbuilding pasta]

I think it compiles the "most" files on demand for each configuration the first time you attempt to run it, so it's not that they're missing but failing to compile (or something fails before it even gets that far). I can shoot the dev a message that these resolutions still aren't working, but I think they've just been too busy to work on exoplasim for a while because there are fixes I know they've wanted to do for a while that they haven't implemented.

[worldbuilding pasta]

If it's at all helpful, I could run some calculations for you when my computer is open, so just let me know.

If you really wanted to, I'd be interested to see what happened with Earth topography (max elevation 6300 m on this file) and standard settings (300 ppm etc) at t85 with 1 m initial glacial cover run for 50-100 years, to see if it fixes some of the issues at t42

[MrBragg]

I can shoot the dev a message that these resolutions still aren't working, but I think they've just been too busy to work on exoplasim for a while because there are fixes I know they've wanted to do for a while that they haven't implemented.

Thanks!

If you really wanted to, I'd be interested to see what happened with Earth topography (max elevation 6300 m on this file) and standard settings (300 ppm etc) at t85 with 1 m initial glacial cover run for 50-100 years, to see if it fixes some of the issues at t42

Yeah sure, can start that in a few days. Quick question though: for the Image2sra script, is the max elevation the script asks for the value corresponding to fully white (255) or the highest elevation in the image, regardless of its color value? I've been assuming the former, but if it's the latter I'll make that adjustment.