WIP (sort of tutorial to be) : Climates, applying Geoff's Cookbook at detail (some)

[Pixie]

Hi folks.

Been working every now and then at the detailed heightmap of this world but I also spent some time in related projects (some hobby free time... which isn't as much these days as I would like it to be). One of them is this one.

With a fairly detailed map of the one continent, I could work a nice climate map as well. This I did, and then redid, and then again. And every time I added a little more depth and a better understanding of climate. I also used Geoff's Climate Cookbook everytime. In the meanwhile other members were also working on climates and I thought there could be more in terms of guidance for that. Hence, I thought I could, perhaps, redo my climate map in nice smallish steps and make it public.

Should the outcome and the process be of any "learning use", I will turn this into a pretty pdf for tutorial purposes..

So this is how I am doing it:

[Pixie]

As I said before, my canvas is a topographical map I am working on, which is, at this stage, detailed only for one continent. That is the continent I am working on, it's called Palamb (and there is one very large country in it also called Palamb).

The first step is to have a regional/continental map which is clear of all clutter but shows the physical geography. This map needs to have latitude lines, preferably (or at least) the equator, the 30º and the 60º lines. I got my whole world map onto g.projector and then tried a few projections. My choice went to Hammer projection. It shows a relatively large portion of the map, in places emphasis on the center of it and visually it keeps an idea of the planet as a sphere.
So I exported my base-map-to-be from g.projector with this end result:
hammerproj_palamb.png

You will see in the next post it is important to show not only where land ends but also where the continental plateau ends. Those boundaries between shallow waters and the deep ocean are very important for currents, and oceanic currents very important for climate, so go through the trouble of setting them, please. On my map, that is the light blue around the continents.

As you can see, the latitude lines are visible, there is ample space for scribbling on top of it all and it gives the viewer a clear view from equator to poles. Task #1 is done.

[Pixie]

So now you have a base map, you will add a bunch of layers on top of it... a lot. So get yourself onto a "layers" drawing program (in other words, Paint won't do).

Time to find the "logical" ocean currents. Now, this is pretty easy and I'll trace it in a few steps.

1. Close to the equator there are two currents, flowing westwards. These are the strongest and the first to be drawn. In between them two (most of the time, exactly on the equator) is a counter current flowing eastward - Draw these currents in black. As the currents are more or less parallel to a latitude line, there isn't a significant exchange of heat between the water and the surroundings.

2. As these two strong currents meet a continental basin (not necessarily the shore, see last post) they get diverged away from the equator (northern one will move north, southern one will move south). They will stick to the eastern border of that continent approximately until somewhere between 30º and 45º of latitude. During this time, they release heat brought from the equator and are, hence, warm currents - you should draw them in red.

3. At the 45º West -> East winds (the westerlies) are blowing strongly enough to create an eastward current. However, this is no sudden right angle turn, it actually starts much sooner. The warm current you drew is no longer as warm, is now moving in a more or less stable latitude and will not exchange heat - draw in black.

4. As soon as this current meets continental shores, it spreads north and south (and wikipedia says it sticks pretty glued to the coast both because of the properties of the colder water and because the westerlies don't blow offshore at these latitudes on the western coasts). The currents moving back to equator will take reasonably cold water, which removes heat from the air in the region they pass through - draw them blue. The part that flows north takes relatively warm water to the poles - draw in red.

5. All it's missing is the circulation at the poles. It isn't clear in my map because the continent is mainly tropical/equatorial and it wasn't needed. All you need to do, if you have polar oceans/seas, is to close the loops, keeping the color code for currents moving poleward (red), "equator-ward" (blue) and laterally (black). Remember, unless you close every loop, you are saying water piles up somewhere, and if you are saying that, you are wrong .

palamb_currents.jpg
On the map there are a few currents which have been placed to close the existing loops. Kuy and Santellan currents are the feeding currents for the Northern Equatorial. Since the latter needs to exist the loops needed to exist and to be closed. The Kuy current is also forced to exist as the path eastward is blocked by the warm Palamb current that is moving north and partially blocking the West-East surface movement.

The Hondan Southward current is generated following the same logic as before. Warm water flows south/southwest from the equator in every continental shore facing east in the southern hemisphere. This generates the said current. To close that loop there's the Ohamana cold current.

EDIT - noted a couple of errors in the instructions for the temperate westward currents, I think it's better explained now and error-free.

[Pixie]

From now on, all steps will come in pairs. That is to say, you need to make maps for Winter and for Summer. But best call it July and January - they represent the extremes in terms of how far south or north one can find the hottest areas on a planet and not the typical temperate seasons.

So, new layer. And you can toggle off the currents one. This new one could be called "January weather outline" and will contain the major features in the atmosphere.

The main features are the areas where air rises, causing rain, and where cold air from the upper atmosphere sinks, causing blue skies and moisture free air. In simple terms, here's what matters for the map:

- Close to the equator is the Intertropical convergence zone (ITCZ). Humid air from both tropics meets in the equator and rises, releasing it's water content. Draw this as a very blurry line as it isn't a well defined place. Inland, the ITCZ moves up and down according to the seasons (it is at is northernmost point in july and at its most southern in january). In the oceans it hardly moves because the surface temperature of the oceans varies much less. The ITCZ has a year-round bias towards the hemisphere with large continents, if you have those - this happens on Earth, with Eurasia. On the other hand, if you have large oceans, the ITCZ might split into two, with a clear sky area just over the equator - this happens on Earth in the Pacific.

- Close to 60º we have the Polar Fronts (PF). This is even less defined than the ITCZ and moves southward/northward during the year in the same manner. Here, the cold air from the poles meets temperate air and again, clashing air masses will rise and make rain. You should now have three East-West lines.

- Around the latitude of 30º, on every ocean, a high pressure center will form. This is where the dry air sinks. Those high pressure zones are more or less stable around the year but they also move up and down as the seasons progress. They tend to stick to the eastern side of any given ocean, not far from where the cold currents moving to the equator appeared on the previous map.

- Contrary to the high pressures at the oceans, the warmest areas of a large land area will be indeed very hot and that causes the local atmospheric pressure to drop. This generates low pressure centers, which only form if the land mass is pretty big, at latitudes close to the tropic lines but maybe as far as 45º (equal latitude to the tilt of the planet, I'm assuming 23º as on Earth).

- In this particular map we can't spot these ones. But, usually, terrific high pressure centers are found at the center of the continents when at their coldest - which means, during Winter. On Earth this happens only in the core of Eurasia, north of the Tibetan plateau.

palamb_januaryweather.jpg ... January

New layer for July...
I struggled a bit to envision the movement of the ITCZ in this continent because of the huge mountain range on the eastern side. There were two options: either it shifts between north of that range and south of that range during the year or it is an area where it is stable throughout the year. Given the proximity of the ocean, I went for the second.

palamb_julyweather.jpg ... July

Note how the PF's have moved up a little, like the ITCZ in same parts. Also, the high pressure centers moved to the north, only slightly, except for the Western Danip High Pressure Center, which was able to stretch eastwards with the change of position.

Have a look at how atmospheric pressure and wind formation happens on Earth with this awesome link, for comparison:
Global Climate Animations Circulation and Winds

... and once you're happy with your work...
Task #3 is done.

(EDIT: corrected January map and added extra instructions - to add low pressure centers in continents, when warmest and high pressure centers, when coldest)
(EDIT 2: added link to Global Climate Animations)
(EDIT 3: added further instructions for placing the ITCZ)
(EDIT 4: made clearer instructions for continental High Pressure Centers during Winter time)

[WoodytheClimateGuy]

One Question though: In what manner would the ITCZ would move? Would it move upwards based on elevation or something?

[Pixie]

Now it's time to work out the dominant winds. This stems directly from the previous work, the distribution of atmospheric pressure, and it's simply adding on top of it.

So get a new layer to scribble and a rather thin and simple brush. This time all you do is drawing arrows.

Keeping the previous work in sight (but you can reduce its opacity, to clear the view if you need, like I did in the example below), your job is to "preview" which way air moves near the surface. The key point is that air whirls instead of moving in straight lines, like the whirlpool in the tub. And it circles in opposite directions in the northern and in the southern hemisphere, as well as in low pressure and high pressure areas.

Here's a kind-of-step-by-step approach:

1. northern hemisphere: air surrounds the high pressure circles clock-wise. It will drift into the PF moving eastward and into the ITCZ westward.
2. southern hemisphere: high pressure circles get surrounded counterclockwise, resulting in drifting into the PF and ITCZ in the same direction as before.
3. even if pretty far from any high pressure center, air gets sucked into ITCZ or the PF just like you worked out before.
4. low pressure centers in each hemisphere suck air, it whirls into them in the opposite direction to the high pressure centers in that hemisphere.
5. winds north of the polar fronts: they always draw a kind of curve from the pole to the west, into the PF.

If you get confused whether it's clockwise/counter-clockwise (as I did so many times), use this for reference:
250px-Winds_pattern_highlow_pressure.gif

Now, don't shy from drawing arrows but don't draw arrows very far away from any of these influences. There are effectively areas in a planet where the winds just never blow strongly...

Again, this can be of assistance if you want to compare your prediction with a planet that we actually know well
Atmospheric Circulation and Winds

Here's my result:
palamb_julywinds.jpg ... July

And since this example has absolutely no land out of the PF, I didn't bother drawing wind arrows close to the poles on this one
palamb_januarywinds.jpg ... January

[Schwarzkreuz]

Nice work. I dont always try it this road for realistic weather conditions. Its somehwat a huge topic of its own, so nice to see someone tries that way.

[Pixie]

Thanks scharzkreuz. Yep, it's a complex topic (next post - rains - is a huge one). But some people like to indulge in it, and some, like me, might get obsessive over detail . Hardly necessary for the typical map one can find in the guild.

I shall add a disclaimer, maybe, advising that it will take labor-days to get the climates done.

[Pixie]

So, here comes the rain.

This is the trickiest and longest task and it takes some time. Geoff's Cookbook has 8 different factors to include and we will cover them all in a layer upon layer technique to add them together. As before, this needs to be done twice, for the extreme months of January and July, and the previous layers of sea currents, atmospheric circulation and dominant winds will all be used.

Also, you need to have a good idea of elevation. Not necessarily an over-detailed map like the one I'm using in this example, but something with a little more detail than just the linear super-ranges. If you haven't got anything else than this, you can make a draft layer marking other ranges and elevated areas (plateaus, like southern Africa, East Africa, Iran and Turkey, Central Brazil, have climates specific to their elevation, so you can't ignore that sort of terrain completely)


Now, first thing to do is to create a few layers to work on, you can call them
- baseline rain
- ITCZ, extreme
- ITCZ, moderate
- onshore winds
- mountain rains
- warm coastal currents
- polar front

I am going to break down what to do for each layer, then we turn all this into a single layer mapping rain in colors ranging from dry to very wet, which we will relate later to Koppen's climates.

1. baseline rain
All the layers will be colored with one color only. I use a not-too-bright blue, RGB:100.100.150. Brighter colors will make the end result hard to read.
For the baseline rain you just select one very big area (the whole map if you want) and bucket-fill it with this color. Then toggle off the layer and toggle on atmospheric features. Make a selection (using the lasso tool) around the high pressure centers (stretch it further east and west than the actual centers drawn before). Don't loose the selection, toggle off that layer and toggle on the dominant winds. Adding to the selection choose coasts where the wind direction is offshore and the areas behind mountains that winds cross (normally called the rain shadows of those mountains). Still, keep the selection. Toggle off the winds. Go back to the baseline rain layer and just press delete. Now your initial baseline rain has gaps big and small. Toggle it off again so you can clearly see the land. Make a selection encompassing all land that is far away from the coast or protected from coastal influence by mountains. Back to the baseline rain layer and delete that as well. The end result, for me, was this:
baselinerain.jpg
Toggle it off, and leave it there...

2. ITCZ, extreme
Same process, but this time way faster. Toggle on your atmospheric features layer. Make a selection that covers the ITCZ you marked before. Bucket fill this with the same color as before, in the appropriate layer. It's done. Here's what it "could" look like:
ITCZextreme.jpg
In hindsight, I may have made it a little too narrow. Because it's a simple single color area, it's easy to adjust and you can always go back at any point.

3. ITCZ, moderate
The ITCZ isn't a very defined area and its effects may be both broader and stronger than any other effect, so we have a second layer for it - no other factor weighs as much.
We're already in the third layer. Make a much wider selection than before (I'd say, at least twice as wide) and bucket fill in this new layer.
ITCZmoderate.jpg

4. Onshore winds
Obviously, you need to view your dominant winds layer for this, but also the elevation map. With those layers visible search for every piece of coast where the dominant wind blows inland from the sea. Select the area where these winds would carry clouds/moisture into. If it's a wide plain, the whole thing, if it's a mountainous coast, only up to the mountains, if they blow exactly onshore, further inland, if they are parallel to the coast, only a thin strip of land. Islands, it's plain to see, get onshore winds whichever direction they blow, so cover them in your selection as well.
My example works out this way:
Onshorewinds.jpg

5. Mountain rains
For this one, you need to see the dominant winds and the elevation map as well. Wherever winds meet mountains, select the areas just before the mountains. The extent of those areas will depend on the kind of mountains, but the amount of rain would also depend on the kind of slopes, so don't worry about this too much (we need to keep it practical to a certain level). If a dominant wind is coming from a particularly dry area (a hot inland region), ignore it in this step as it wouldn't create rain even against an everest style concrete wall).
Plateaus also work as barriers for this, although you need to extend the selection somewhat into them and not just "before".
Mountainrain.jpg
If the layers are looking a little strange to you, don't worry. In the end they'll get added in a neat way into a rain patterns map.

6. Warm coastal currents
These influence the weather in a large area. Unlike cold currents which are narrow and stick in the coast, warm currents are more spread out and the moisture they carry gets transported by the dominant winds. So toggle on the sea currents and the dominant winds layers. Your selection will encompass the red arrows you drawn for currents but stretch it into the areas where warm currents meet cold currents and also in the direction the winds are blowing in the area. A warm current with onshore winds and a deep coastal plateau will generate a very large area of influence.
In the end, without loosing the selection, toggle off the current and winds layers and have a look at the elevation map. You may have to cut down the selection a bit if you went inland over mountains - those areas should not be influenced by the oceanic affairs.
This layer ended up like this for me:
Warmcoastalcurrents.jpg

7. Polar Front (lastly, but not least)
The Polar front has a broad effect as the low pressure centers that form in it are prone to move about and form in different locations. So all you need to do is to select a wide area around your previously drawn polar front (as seen in the atmospheric features layer). Now, these low pressure centers make an eastward movement between birth and death so they normally carry more moisture when they enter the west side of a continent than on the eastern coast. Because of this, make your selection wider in the West coast, stretch it inland but narrower over the distance and, eventually, non-existent on the east coast (this depends on continent width and on mountainous terrain in the way).
Polarfront.jpg
Because this continent is mainly tropical/sub-tropical, the PF has virtually no influence on its weather... less to worry about.

And so the different factors have been considered. Apart from the ITCZ, which we accounted twice, the assumption is that they all have the same weight, which is naturally wrong, but it's a compromise we can work with. I'll continue in the next post to break give you a break.

[Pixie]

The ultimate goal will be this (here shown covering only a small area)
wetnessmap.jpg

In this image I labeled the 5 levels of rain we need later, in order to compute the climates: DRY, LOW, MODERATE, WET and VERY WET (this technique will eventually yield a 6th level - even higher - we can call it INSANELY WET, but treat it as just very wet )

Here's how we transform the previous stack of layers into this:
- set all those layers to "screen" mode (instead of normal) (can please any photoshop/gimp savvy person explain what this means, if necessary?)
- duplicate them (as not to lose your work in the next step and easily go back and quickly correct anything if need be)
- make only the duplicates visible and turn off all other layers in the file (all, including background)
- when you set them to screen mode, they turned very whitish, but now there are areas in the original color and brighter areas where those layers superimpose, right? (just checking)
- merge visible layers

You can name the resulting layer with an appropriate name like "Rain Patterns, July". Once the background is visible again, it might look like this (again, just showing a little area):
addedrainpatterns.jpg
The difference between this and the end result above is that I have rounded off the corners. Because the map is supposed to be showing a "likelihood" of rain, transition between levels should always cover the intermediate levels. Depending on the level of detail you want, you may skip this part altogether and adjust later on when we get to the actual climates.

One last thing that can be done, if you want to, is to delete all the parts of the "wetness" map that are over the sea.