Reply With Quote
It looks like an improvement.
The driest category would be 0-5 instead of 0-10 ml?
That's certainly true - while nose deep in working on this it's easy to forget that the goal is for the script to provide a good model for speculative worlds - and so the input data for such applications will probably be a bit 'cleaner' in the sense that it'll probably be closer to having the two extremes for everywhere, rather than a time snapshot that catches temperature/precipitation extremes for many places but not actually all.
Using the F aridity category for all pixels that have the same temperature in both input datasets gives the following result (after I toned down the top precipitation estimate to 300mm):
earthTest17.png
Or, using the W category as alternatively suggested in those same cases:
earthTest18.png
In practice the metric of 'same temperature in both seasons' will probably only hit equatorial regions, polar regions, and maybe a few other anomalous spots, and all the E climates are exempted from aridity checks already in my script - but it does still seem a kinda iffy way to check. I could restrict the assessment only to climates that would fall into A barring potential aridity and have the same temperature in both seasons, if that seems like it would be a better representation of equatorial climates where precipitation in either part of the year would have relatively equivalent impact on vegetation growth (although as Azélor mentioned this concern might apply to other regions where temperature has very little seasonal variation, where they exist besides equatorial zones).
Once I've set my script up to use configurable input color profiles it shouldn't be too hard to give an extra precipitation category a shot for very-dry-dry-season Am detection and so on.
It looks like an improvement.
The driest category would be 0-5 instead of 0-10 ml?
My Deviantart: https://vincent--l.deviantart.com/
The added category would be a new very wet category (break 200mm+ into I think 200mm-400mm and 400mm+ categories - the former would have a mean of 300mm and the latter would use a stand-in 'assumed mean' of 450mm) like Charerg mentioned, since Am regions with a very dry dry season can end up Am by virtue of having sufficiently high precipitation in the rainy season, but a 200mm+ category at the top isn't rainy enough to do it:
Originally Posted by Charerg
Yes, I agree it's fairly safe to just use the standard criteria everywhere, even for equatorial areas. Since the "baseline threshold" is already 500-700 mm in equatorial areas with high temperatures, it doesn't make as drastic a difference whether the threshold has +140 or +280 extra applied to it as it does with the cooler climates. And of course those numbers are somewhat arbitrary to begin with.
Also, I agree that in practice when creating precipitation maps for a fictional world the equatorial areas are very likely to get messed up a bit because the abrupt season reversal. In any case, it's almost inevitable that some degree of manual cleanup work will be needed with the final climate maps.
Last edited by Charerg; 02-14-2018 at 02:14 AM.
Have I understood right that you folks are working on a program to model climate in speculative worlds? That would be very cool, and if so I'd be curious to know how far it's come along.
I'm afraid it's not a program. We've created a script that takes the precipitation and temperature maps for January and July as input and outputs the climate map. There's the original version of the script made for PhotoShop by Azelor, and the more recent GIMP script myself and Azelor created in collaboration (and which is slightly updated from the original, I guess). There's also an unpublished version by AzureWings (which is written in Python, I believe, and should be compatible with GIMP, not sure about PhotoShop).
Running the script saves you the trouble of manually creating the climate map, but you still need to create the aforementioned precipitation and temperature maps.
That is way more ambitious that want we could accomplish. Depending on the level of precision we would need a full team of climatologist/programmer and a bunch of very powerful computers.
My Deviantart: https://vincent--l.deviantart.com/
That at least sounds easier than doing the whole climate map by hand like I did. So many spreadsheets... You have my full respect for this thread.
Btw, I think it might be a good idea to put a link to the latest version of the GIMP script in the OP, so at least people know that there's one available, since the thread has grown so long that it would be a pretty heroic undertaking for a potential tutorial-user to read through all 37 of the pages.
Have you thought about Universe Sandbox ?
Of course it's not a NASA soft and the climate aspect still seems to be underdeveloped.
But they have hired this year an "Astrophysicist, Climate & Simulation Developer" and their incoming Planet Grid could improve the climate simulation.
Mays be i'm too optimistic, but is there nothing you could do with this soft ? (i do not own it btw)
