## Thursday, April 9, 2015

### Weather System Mark 6 - Wind & Other Trends

Step 3: More Stations

When working on any new project, I always end up having to graph out certain elements in order to keep them clear in my head.

This is the Stations Map:

The lines are tenths of a degree of latitude or longitude.  This particular part of the world, surrounding Palmyra, is comparatively short on stations, so the distances are greater than they would be for Europe.  However, since this part of the world also experiences a more homogeneous climate than Europe does, we need less stations.

What we want are four stations surrounding Palmyra, each of which are calculated the way that Rutbah was in the last post.  This gives us four different comparisons with Palmyra, so that we can see the movement of air between four different 'pairs' at a glance:

We don't need the temperature right now, so we can concentrate on the air-masses alone.  The table above shows how the four stations around Palmyra affect the weather there.  The "Higher" column indicates the station with the higher elevation.  I've designed this so that the reader can clearly see that Palmyra - H & Hama - L make the pair 'HL.'

Elevation for the station is given as a number after the station name.  Note how Palmyra is lower than Rutbah and Irwaished, but higher than Deir Ezzor and Hama.

Because LH and HL are both "storms," the wind direction is unstable for the entire region.  Both the higher stations are producing downbursts into Palmyra, while the lower stations are experiencing an inversion wind shear with the warm air emerging from Palmyra.

If the reader looks at the map, however, it will be noticed that Deir Ezzor is directly north of Rutbah - and ultimately, those two stations ALSO form a pair.  Rutbah's weather would also be directly affecting Deir Ezzor just as it does Palmyra - and that pair's code would be 'LL' (subsidence).  This gives us a pretty stark indication of where the storm begins and how far it reaches.

What we have, on the whole, is a High sitting overtop of Palmyra, getting blasted with cold air from above, then hitting the cooler air to the northeast and west.  The atmosphere is swirling and wind directions are shifting from hour to hour (unstable).

All this from very little data.

There are two important points we want to take away from this, giving us both good and bad news.

The bad news is that when I recalculate, the numbers will change and the weather pattern will produce something potentially different.  The most common pair that will come up is LL, because this is night (remember?) and the way the system is designed, night time temperatures have a 2/3rds chance of being below the average mean.  Most of the time, LL will come up all across the five stations and the weather will be a gentle subsidence in the direction of the lowest elevation.

However, we don't have to say that any given situation lasts a day!  We can roll a d4 and propose that the latest conditions last 1-4 days.  It is up to us.  That, however, is beyond the point where I've reached; at the moment, I'm just trying to get these patterns to manifest rationally.

The good news is that every station everywhere in the system is linked.  They don't interact in different tables; they interact altogether, at the same time.  Therefore, the air mass out of Rutbah that affects Palmyra also affects Deir Ezzor and Irwaished, as well as stations to the south and east, which are in turn affected by stations around those and so on, potentially right around the globe.  If we wanted, we could create an air mass chart for the whole system.  If I had the ability to program a visual tool, that chart could be automatically generated!

Sadly, however, excel is as far as my expertise takes me.

Step 4: Wind

So far, I have done the work for four stations: Palmyra, Hama, Tripoli and Beyrut.  This is the direction I expect my players are going to take the next time they run, as they are bound for Egypt.  Here's a complete list for all the pairs affecting those four stations (things are going to get a wee complicated now):

As the reader can see, I've added wind information for each pair.  We'll get to that in a moment.

It would be easier at this point if the above table were organized visually in a way that mapped out the pairs in a 2-dimensional fashion, like the stations map above.  However, in the end there are going to be thousands of pairs; and I only need this chart as reference for the next chart.  I felt, therefore, that the reader ought to look at this the way I have.  As a mass of data that has to be pieced together visually inside our heads.

Note how Hama having a high in the above creates swirls and eddies all around it, as does the one pair between Irwaished and Palmyra.  Were I to recalculate the table, that would change; sometimes every pair from top to bottom is LL.

Wind Shear is the effect of the paired air masses we've been discussing since the last post.  Wind direction is determined by the movement of air; most of the time at night, this will be either be subsidence or unstable masses churned up by a night time high.  Note that in the daytime, the most likely pairing is HH, so that most wind movement will be up elevation, not down.

Conditions are a more descriptive term for what is actually happening visually from the character's point of view.  Characters don't understand words like 'subsidence' and 'downburst.'  For them, explaining that there are thunderheads, a wind storm or that it's overcast makes much more sense.

Now, some readers will be realizing that this area is a desert and the 'overcast' doesn't make much sense - particularly if LL is common and means that an overcast is occurring.  Well, settle down; I do take that into account later.  For now, don't read too much into the term 'overcast.'  It is, at this point, a loose framework upon which we shall hang other details.

Wind speed is determined by the amount of drift we generated two posts ago. Remember drift?  Drift was the difference between the randomly generated temperature and the mean.  Drift told us that the station was experiencing a H, L or M air mass.  Well, the amount of drift determines the wind speed.

Take the example above:  Irwaished's drift is 12, whereas Palymra's drift is -1.  The difference between those two drifts = 13.  Because the local wind shear/conditions are an inversion/thunderhead, this number is taken as read and the wind speed is 13 mph.  On the Beaufort Scale, this gives us a moderate breeze.  That looks like this:

The wind speed is modified by the given pairing.  HH and LL both cut the wind calculation by 50%.  HL, the inversion, takes the wind speed as calculated.  LH, the wind storm, increases the number by 20%.  HM, the unstable thermal, drops the calculated number by 50% as well; a stable thermal, MH, to 40% of the total.  LM, the downdraft, reduces the number to 10% of the total and the adiabatic cooling, ML, to 30%.  These numbers are entirely ad hoc, but it 'feels' right.  Where am I going to get accurate numbers for this sort of thing?

Having made the calculations with excel, the last column calculates the Beaufort scale.  I don't know what 13 mph looks or feels like, but I can make a reasonable assessment from the above image and description.  I find that extremely helpful.

1. Thoroughly interesting stuff so far Alexis. Do you have any concerns over the variation between stations inherent in generating random numbers? What I mean is, does the variation between stations relatively close to one another create weather that will be much different than the weather one would otherwise see reported on the TV news? It strikes me that weather systems seem much larger and regions more homogenous than what your results will indicate. Will there be more downbursts in your Palmyra than in the real one because of the variation between it and nearby stations? Any concerns over what will happen in Europe when the stations are more numerous and closer together?

2. Well, there is variation, no question. Not as much as might be thought, however. While I feel it would be useless for a global simulation, I suspect it will be perfectly fine for a role-playing game.

It needs more experimentation, but at the same time the structure should allow me to tweak it in future, after some testing.

I'll have the rest of the system up and in place before the end of today, and post the program so people online can play with it.

3. I agree on the structure and the approach seeming sound enough to support further tweaking. In reading and considering all of this I'm only frustrated in that as I don't run in the real world, I'm going to have to pull a lot of data out of my ass that would otherwise be available online. I suppose that's a small price to pay for a working RPG weather system for which somebody else did all of the heavy lifting. ;-)

4. Might be less trouble than you think, James.

You should be able to select a part of the world and then just copy the details from various sources. Your world is unlikely to be as large as mine, so an area the size of New England or Georgia-Alabama should do for you, depending on where your world is.

You could take a lowland like East Georgia and mix it with upland Colorado. Doesn't have to have the kind of verisimilitude that my world has, right?

I've found you can copy and paste into excel from worldclimate.com, so there's no tedious copying.