View Full Version : [Award Winner] Where does the wind blow?

09-19-2010, 09:43 PM
Want to figure out wind, weather, and ocean current patterns for an arbitrary world?

I'm not a metrologist, nor do I even play one on TV. I do have an insatiable appetite for odd topics, especially geographically-related ones. The level of expertise I try for is enough to make a plausible fictional climate. And remember while I present this, that I'm mostly talking climate, not weather. Weather systems will override these average winds often. This tutorial should help you figure out believable winds across your own world map. Since he was asking about how to ensure a certain area really *would* be desert, I've started with Slipguard's planet Megalaos (http://www.cartographersguild.com/showthread.php?11915-Whaddyall-think-about-The-stories-behind-maps&highlight=slipguard).

The general surface winds across an idealised globe, whose axis is vertical, and whose surface is uniform, would be something like this. (Non-idealized Megalaos used as a background).

09-19-2010, 09:49 PM
These surface winds are paired with opposite circulation aloft, like so:

09-19-2010, 09:51 PM
But our planet is tilted. Even our supposed planet-sized cue ball would only have winds like that on the fall and spring equinoxes. Those cells of circulation are driven by both rotation and sunshine. So in the northern hemisphere's summer, the pattern shifts north, until at northern midsummer it would look like so:

09-19-2010, 09:52 PM
And in the northern winter the 'solar equator' is down at the Tropic of Capricorn. Here that's 23.5 degrees south latitude, since earth's axial tilt is 23.5 degrees. That's why I asked what Megalaos' tilt is (http://www.cartographersguild.com/showthread.php?11915-Whaddyall-think-about-The-stories-behind-maps&p=128574&viewfull=1#post128574). The more pronounced the tilt, the bigger the swing would be from summer to winter and back. Less tilt, somewhat more placid weather and climate both. Hmm - or maybe the term should be *even*, not necessarily calm. Midwinter like so:

09-19-2010, 09:57 PM
I'm assuming Megalaos' tilt is about like Earth's, for the sake of the argument. Ahhh, but neither Megalaos nor Earth have a uniform surface. Ocean and land soak up vastly different amounts of sunshine. Then there's ice.....

Our understanding of weather and climate is based thus far on ONE planet, so it's fair to figure for vastly differing surface conditions, the circulation would be different. I don't know how things would work on an ice world (http://en.wikipedia.org/wiki/Atmospheric_circulation) or a dried-up desert planet (http://www.cartographersguild.com/showthread.php?4653-Arnoria-the-dying-world&p=121133&viewfull=1#post121133).

At any rate, what seems to happen is that in a given hemisphere's winter, there's a general high pressure concentrated over big land masses, at least those around the default high pressure peak of 30 degrees, shifted to follow this 'solar equator' (http://en.wikipedia.org/wiki/Solar_equator). Smaller land masses, not so strong. In summer, things shift (http://geography.uoregon.edu/envchange/clim_animations/flash/mslp.html) and the strongest highs wind up over ocean. In either case, if there's not much land under the belt of high pressure - what I labelled 'horse latitudes' (http://en.wikipedia.org/wiki/Horse_latitudes), then there are high pressure blobs over ocean. The pressure differences 'clump', so instead of a band of high, you get a string of ovoid regions of high. If you have a really big landmass the effect might be more pronounced - Earth's Eurasia, for instance. Megalaos' Boris continent isn't nearly as big as Eurasia.

High pressure, low pressure - who cares? Well, the pressure is relatively high or low based on what the air is doing. Look at the black-background globe above. See the air rising at the equator? Particularly if it's passed over ocean lately, that air is fairly wet, and rising, it cools. Cooling, it can hold less moisture, and it dumps a lot of it. There's a strong reason that equatorial areas tend to heavy rain. Coming back down, that's fairly dry air, hence the deserts on Earth that are around 30 degrees north and south. The other generally high pressure zone, with descending drier air, is at the poles. Antarctica may have kilometers of ice, but it's technically a desert.

OK, so straying even further into fantasy :-) ... in that I'm guessing even more, but it's all still plausible... we could figure Megalaos' seasonal highs and lows like so - winter first:

09-19-2010, 10:05 PM
Did I mention I stretched your map to show 1 1/3 diameters? Makes it easier for me to visualize this stuff if I'm not having to mentally wrap vectors left edge to right edge. I'm also assuming the north and south edges of your world map don't run all the way to the poles.

Since I don't want to presume to repost someone else's map, here's Paidixira and Megalaos (http://www.cartographersguild.com/showthread.php?8713-Paidixira-My-first-worldmap!-HMOG!&p=96555&viewfull=1#post96555). Well, except for the fact I am using a simplified Megalaos map as the base for the first part of this tut :-).... I have been referring to the continent map further down that thread for climate preferences - the one where The Cusp is mostly desert.

Now, given that these highs and lows space themselves out apart from one another, there's multiple arrangements that would be about as plausible. Knowing that you want the midwest of Paidixira to be dry, we'll say that's where the main continental high lurks. I'll go further in a minute, but like you can guess from the way the wind bands slant, northern hemisphere highs have an outflow that spirals clockwise. So a high over western Paidixira leaves that dry-ish air curving down across the SE of the continent too - I see you have a desert across The Cusp. We try to accomodate :-). Air crossing land doesn't add much moisture, so there's not a lot of rain there in the SE. Those crisp Times Roman highs and lows are in reality very blobby, and shift around. Maybe some of the winter half of the year there's two weak highs, one on either end of the continent.

In the (northern) summer, you might get an arrangement like so:

09-19-2010, 10:17 PM
The N hemisphere lows are counter-clockwise spiraling inflow. I can plausibly imagine a couple of weak lows bracketing Paidixira like shown, or one larger one more centered on it. In either case, the combination of clockwise outflow on a high and counterclockwise inflow on a low still conspire to drag wet air across those mountains to dry it out before it gets to the desired deserts. I'm thinking the pair get the SE dryer, so your depicted The Cusp desert is sensible.

I'm sitting here twisting imaginary airmasses with my hand above the screen. Son #3 is asking if I'm drawing doorknobs :-). It's time to do some more realistic wind than those idealized belts. Southern hemisphere highs and lows keep the same in vs out, only the spiraling is opposite-handed. Crossing the equator you can apparently get a shifting of 'handedness' of the curve to the winds, thanks to the coriolis force. Another rule of thumb I haven't mentioned is that equatorial belt labeled as the doldrums on the ideal globe is more of a trough than a string of pits, if you want to visualize pressure as height ("low", "high" - works for me). And the effective line of that 'solar equator' is called the intertropical convergence zone (http://en.wikipedia.org/wiki/Intertropical_Convergence_Zone), and bends farther away from the real equator over land than it does over water. The 60-degree(-ish) low area can be a trough too, especially in a hemisphere's winter. Honest, I got all that stuff from climate and metrology books & sites... any wrongness is me, any rightness is luck, or somebody else's research.

09-19-2010, 10:23 PM
Make some arrows. Hmm - The bigger ones will do for a crude set of vectors. I used smaller ones for a lot of detail on the Aurora Wind Maps (see far below), and it drove me nuts. You guys who can drop in a curve with arrowhead with a couple of clicks, please do so and make prettier pictures than this.

09-19-2010, 10:26 PM
Done my clunky way, dupe a bunch of those - enough for each high and low. Or get one low about right and dupe THAT for the other lows.

By the way - I dunno if there's a convention for a software-agnostic tutorial. I'll note each time I have created a bunch of something I go back and collect all of those on a single layer.

I do the lows first, since the arrowheads are at known locations (next to the symbol):

09-19-2010, 10:33 PM
Now - inflow to the lows has to come from one high or another. Just start filling in the map with flow, remembering highs source clockwise (in the northern hemisphere. Whoops - that thought reminds me: I have already loused up the southern Lows - they spiral the other way. Siiiigh. I illustrate this to make the point it's easy to get carried away and do all alike... :

At least fixing handedness is a simple horizontal flip.

09-19-2010, 10:35 PM
That said, how about prevailing winds like so in northern summer:

09-19-2010, 10:40 PM
Stare at that a while and think - does it follow all the rules of thumb? Mmmmm - no, one I left out is that when the ITCZ migrates away from the equator, the inflow on the equator-ward side of it tends to switch direction. Think of it as the same as the way the southern lows have air spiraling in one way, while in the northern hemisphere it spirals the other. If there's a pronounced low *spot*, the spiraling bit holds, whereas if it's a trough, the 'ideal shifted ITCZ behavior' rules. So that means my two seasonal ideals need slight revisions, like so:

09-19-2010, 10:44 PM
Yeah, my Aurora wind maps don't show that back-curving when the wind crosses the equator. Call them Version 0...

So that makes the 'completed' northern summer pattern look like:

09-21-2010, 04:25 PM
Well, phooey. Now that makes SE Paidixira wetter in Summer. Dryish air sourced from that high to the SE, but it passes over ocean before it crosses The Cusp. The area's still pretty dry in winter, based on the location of the high. If this was my planet, I'd evaluate whether I needed The Cusp to be desert. If I was intent upon it being seriously dry, I'd think if I could plant a set of coastal mountains across the very south coast of The Cusp. If it was a pretty strong set of mountains, it would dry out the air on the lee side.

If you DO want The Cusp to be desert, hmmmmm.... I could probably consoldate the two summer highs south of Paidixira into one, that hangs near The Cusp. Of maybe strengthen the leftmost and drag it near The Cusp, and shove the right one out in the ocean a bit to the right of where it sits now. <shrug> What do you think, Slipguard?

Before I go on to the northern winter, does this make sense? I have further explanation in mind related to this atmospheric circulation (http://en.wikipedia.org/wiki/Atmospheric_circulation) stuff, such as the rain shadow effect (http://en.wikipedia.org/wiki/Rain_shadow), how these surface winds drive ocean currents, and typical tracks of tropical cyclones (http://en.wikipedia.org/wiki/Tropical_cyclone).

09-23-2010, 03:40 AM
This all clears the waters, yes. I think the direction you're going in is clearly defined, and I think I want to follow it. The Cusp simply has to be impenetrable for an invading force. It could definitely work as a dense tropical jungle story-wise. Summer precipitation gets my green light, sir!

P.S.If the Cusp goes that direction could those kinds or densities of jungle show up elsewhere in Paidixira, like say, in Valpyra or Komagi, or would the Cusp's biome be unique compared to the rest of the continent?

P.P.S. What do you know about how the placement of trade winds and pressure centers effect sailing routes?

09-01-2011, 10:33 PM
I'm working on a planet that's rather different to Earth, and I'm unsure how that will affect the atmospheric circulation. It's rather smaller (5716 km equatorial radius), but has higher surface gravity (1.1g) due to having a big core. Its day length is 11 hrs 32 minutes, and the axial tilt's about 10 degrees. I'm wondering if that increased rotation speed will result in there being more than three circulation cells.

09-02-2011, 01:07 AM
I've read that with faster rotation, the cells can "tear" into more - five N and five S next, I assume. Look at Jupiter and Saturn for a "many bands" effect. Of course their parameters are WAY different from Earth or your planet. My gut feeling is doubling the speed won't spawn more circulation bands, but that's based on no expertise whatsoever.

Hmmmm... but in looking up books on the subject of planetary circulation, I ran across a guy I work with as a reviewer.... let me ask him if he knows useful texts on the subject, or if he has one I can borrow.

Small world....

01-24-2014, 10:37 PM
Wow, this is a great tutorial, really informative. I was wondering though, how can you tell what latitudes the high pressure (30 degree-ish) areas and low pressure (60 degree-ish) troughs will move to during the summer/winter? I understand that the ITCZ will move to the tropics of cancer and capricorn, and will bend away from the equator while over landmasses, but how do you tell how far these areas will move?

01-25-2014, 12:25 AM
How far? Insanely wild guesswork, tempered by a look at what happens on Earth. Honestly, any of this wind & climate work is so far into pure fantasy that I should be drawing dragons on all the updrafts, and fairies on the downdrafts -- it's plausibility built on a duct-tape and baling wire foundation that I'm shooting for. Nobody, absolutely nobody (excerpt internet-type trolls) will question your work, but will instead be blinded by your science and will marvel in awe at the pretty winds. Don't tell yourself your suppositions are tissue-paper constructs, but rather repeat to yourself that this is how it IS on this world, because you SAY SO. If you are male, do this in a manly growl, thumping your tankard of ale on the table authoritatively. If female... I dunno - some aspiring female climatologist tell us how to be assertive in one's head, ok?

My goal in adding complexity to a world this way is to get all the delightful odd turns and unexpected revelations that I wouldn't have thought of in a vacuum. "Oh, so over here would be an eddy, huh? That's cool." "and over there the rain would be monsoon-crazy for two months then would be missing for ten... " and from that starts to grow in my head a picture of the culture, the people, the social geography.

Back to your question - how far will the ITCZ move poleward over landmasses? As far as suits you. Figuring that these curved winds can't plausibly make right angle turns... maybe letting how the parallel circulation lines 'want' to fill in gaps be the driver. If the ITCZ is the 'heat equator' and vertical circulation is dumping wetness thereabouts, then let your imagination paint jungles and rain forests and swamps as far poleward as that line wanders -- if imagination is intrigued by a certain peninsula maybe being heavily treed and populated by more Holliphants than lawyers, then tell imagination "imagination, good job. that's right where physics says the ITCZ line goes." Then look over at nearby people and lift an eyebrow, with a firm "What?!". What business is it of theirs if you voice both parts in a dialogue between Reason and Imagination?

Then based on that 'seasonal heat equator' I'd shift highs and lows to be somewhat evenly distributed. If that pushes two highs rather close together, I'll squint and make a snap decision that maybe those would coalesce into one larger high. Or I'll take a less snappy tack and say some years one gets two weak highs, vs. other years a stronger single, with consequently differing weather driven by the climate generalities. SO all of a sudden my mariners have a reason to be unsure WHICH year they'll return to port, since favorable winds happen only one year out of three across this certain stretch. And ashore, the wheat crop fails for that one year in three, when rains are too heavy for optimum growth. More autogenerated details, like the fractals beneath the terrain generation I started with, or genning character traits for roleplaying with dice.

01-25-2014, 06:26 AM
Don't tell yourself your suppositions are tissue-paper constructs, but rather repeat to yourself that this is how it IS on this world, because you SAY SO. If you are male, do this in a manly growl, thumping your tankard of ale on the table authoritatively. If female... I dunno - some aspiring female climatologist tell us how to be assertive in one's head, ok?

If you are female, do this in an only-dogs-can-hear-this high pitched shriek, thumping your broom against the head of anyone who questions your god-like world-building authority. ...Authoritatively.


01-25-2014, 06:05 PM
Haha, fair enough. I suppose a good deal of what worldbuilding is is simply playing God, "And here there shall be land! And here there shall be ocean! And here shall be monsoon winds!" I was wondering though, if you know of any tutorials or good places to learn about how to form ocean currents? I've read the climate cookbook, but its section on ocean currents isn't very clear, and I was wondering if there was anything better to use?

02-07-2014, 12:45 PM
I tried to follow this, and I'm so lost...

I tried to follow this, and I got to here:


And I realized that it looked AWFUL, *and* I was obviously doing something wrong. >.<

Do you know where I messed up? I didn't save at each step so I can't go back and find it... :(

02-08-2014, 01:43 AM
Cookierobber - if you've followed the generalized wind maps thoughts like above, and you have a set of summer and winter conditions you like, then look at the ocean basins, divided by the equator, and see if there's any consistent 'push' going on. My understanding is that a huge driver - literally - for surface currents is the drag of surface winds. If both winter and summer circulation has a bunch of wind shoving water one way, maybe that's a decently strong current... with the concurrent action being any water that goes somewhere has to come back somehow. So even if some of the year winds are against an obvious return current, both slosh and elsewhere-flow will push along there anyway. I can't find it just now but I've blathered in that vein on here somewhere too - water currents as well as air thus all being influenced by these high & low cells.

Jalyha -- so what tilt does this world of yours have? If much, then you'll get not just the kind of things you show, but also a somewhat different set on the other side of the sun. As far as looking wrong - well, the placement of highs and lows can't be a rigid set of rules - air is quite fluid and can't exactly stay piled up somewhere. So imagine highs and lows once 'initially' placed by the rules of thumb as 'settling out' in a sort of birds-settling-on-a-wire manner. Shuffle, shuffle, shuffle, till things are even-ishly distributed. To some degree a high is just where a low *isn't* and a low is where highs *aren't*. Are you depicting northern-hemisphere winter there, or summer? You seem to have taken a stab at the 'continental high in winter' thought from post five above...

02-08-2014, 01:57 AM
Yeah that's what I was trying to do :) "continental high/winter"

It's... pretty close to the same tilt as earth would have I suppose :)

And I think the "birds on a wire" thing helped. :D I will try shuffling again in the morning ^_^! Thanks so much

03-17-2014, 04:15 PM
I've had trouble with some of the winds, but I found a solution that works well enough I'll share. Since I'm having to redo my continent due to client (GM) requests, I'll use some of those images.

I make a hex overlay of the map - large hexes that are hundreds of miles across. (I'm using GIMP, and RobA's hex grid script found here (http://www.cartographersguild.com/automation-scripts-actions/5338-gimp-script-hex-grid.html).)

I note (specially mark) the centermost hex of any land mass that covers at least one hex. Given a choice of two almost-center hexes I choose the most easterly. If the choices are equidistant east which I choose depends on season. If winter the one furthest from the equator, if summer the closer. And I also note the critical latitudinal lines: equator, 30, and 60.

I make two of these maps, one for summer and one for winter. For each I mark the continental center(s) as high for winter and low for summer.
Summer: 62397 Winter: 62398

(For the rest of this I'll do summer only.)

So now that I've got all this, I start drawing prevailing wind in each hex. It's really simple in principle though you get a couple of fun frustrations. Basically you find the closest high or low and you draw an arrow so it point to the hex side one to the right (north hemisphere) or left (south hemisphere) of that direction. You turn the arrow because of coriolis, which has already been mentioned in this thread.

You've got occasional conflicts. For simplicity I always assume the strongest zone - high or low - is the zone on the continent. This eliminates most problems. However, as you can see by the map I get places where the wind seems to go opposite directions. I've got those highlighted with a red zone on this map:

Even so I can begin to see how the rain falls. First pass I start with an arrow that's over water and pointing toward land then draw a line till it reaches mountain or is past the continent. Second pass I look for hexes that have 'dry' arrows and back up to see if there are any water-lines they may reasonably have picked up. Eventually I wind up with a 'wet' map:

Now as it happens this gives me some dry areas I don't want dry, so I'm going to be tweaking things (mountain ranges, axial tilt of the continent, latitude - that sort of thing) to get some rain where I want it. But it was an easy way to figure out where the heck the winds were blowing.

Note that if I need more precision for rain I can use smaller hexes.

03-18-2014, 11:59 PM
What a clever scheme! This could perhaps be automated - and as is, seems able to generate some plausible results with maybe less wild guesswork than intuitive rules of thumb. Well done: we want to see more !