gemini://zaibatsu.circumlunar.space/~solderpunk/gemlog/cloud-spotting-blues.gmi

tfurrows wrote a phlog post yesterday centred mostly around responding to severe weather advisories and warnings. Right at the end he mentions participating in "weather spotter / Skywarn training", something provided by the US National Weather Service. This seems to basically be a program to "crowdsource" early detection of heavy weather (what did we call "crowdsourcing" before that became a term?). I would certainly be interested to read more about this experience!

I have never lived anywhere where heavy weather was a routine enough event for there to be an established warning system in place (maybe that statement will sound very strange to everybody in the near future). But in recent years I think I have tried twice to familiarise myself with the standard nomenclature / taxonomy for clouds, partially as a natural response to spending more time outdoors but also as part of a conscious effort to try to reverse where possible the disconnectedness from the natural world that my civilised upbringing has instilled in me. Cloud spotting is the most aesthetically pleasing part of becoming in general more "weather literate", and it doesn't require buying any instruments, so why not start there?

To be honest I have really struggled to get the hang of it. I *want* to blame the system for being a silly one, but I know *that* is silly. It can hardly be said that its not fit for purpose. Professional meteorologists have used it for centuries. It is the system by which observations are conveyed to aviators and mariners who rely on that information for safety of life and limb. And it's silly for me to expect natural phenomena to fit into a perfectly regular and systematic analytic framework, but as a geek it's hard to let go of that mentality.

While there are no shortage of exceptions and special cases, the core of the system seems to be that clouds are categorised into one of five basic shapes or "forms", and one of three levels of altitude. Each form has a name (let's pretend they are A, B, C, D and E) and each altitude has a name (lets pretend they are 1, 2 and 3), and those names can be combined so you have, e.g. type 1A clouds and type 3C clouds and type 2D clouds. So far this is pretty sensible. For the record there are *not* fifteen recognised cloud types like you might naively expect. I assume this is simply because atmospheric physics makes some forms impossible at some altitudes, but I've never found any kind of introductory cloud classification text which makes this explicit.

Where it gets a bit weird is that the five forms and three altitudes don't have single letter of single digit names, but rather Latin names. Which is fine, that's how everything was done in the days this system was introduced. But some of the names assigned to forms are incredibly similar to some of the names assigned to altitudes, and one of the form names is basically a mashup of the two others, all of which is a recipe for disaster, as we shall soon see. The five forms are:

But not every recognised cloud type has this simple two-part structure. Sometimes applying the systematic approach results in a name that just sounds silly. What should we call a low altitude sheet of cloud, "Stratostratus"? Are high altitude wisps "Cirrocirrus"? No, that would be dumb (or rather it's a lot like what linguists call "reduplication", which sounds silly and childish to English speakers but is actually a productive grammatical tool in some languages), so in those cases we just disregard systematicity and call them by their altitude alone (or is it their form alone?), so we have "Stratus" clouds (St) and "Cirrus" clouds (Ci). I guess what has happened is that some of the names for the five forms were chosen based on the altitude at which they most typically form. That is obviously a terrible idea when the plan is to have a system which combines altitude and form, but how else can you explain this mess? Then again, maybe that isn't what happened. I don't speak Latin, but maybe the "strat" part of "stratiform" really does mean something like "sheet", after all archaeologists talk about "strata", which are thin, flat layers of dirt. Is it just a total coincidence that "low altitude" and "sheet" *and* "high altitude" and "wisp" are *both* similar sounding in Latin? That seems extremely unlikely...

Anyway, here's the boss fight of this whole thing. There's a kind of cloud called Stratocumlus (Sc). What *is* that? Is it a low altitude heap of cloud, strato + cumuliform = Stratocumlus? Or is it low altitude patches, rolls and ripples and somebody (quite rightly) decided that "strato + stratocumuliform = Stratostratocumulus" was just getting out of hand? It's the second one. The "system" proposes a name for one combination of altitude and form which is long, cumbersome, and silly sounding, so the "solution" is to abbreviate it, which results in a name exactly the same as the one the "system" would propose for the combination of the same altitude with a *different* form. How much more broken can you get? At first I thought "Oh, maybe this is a rookie mistake, maybe somebody who actually understands the physics of cloud formation finds it obvious that low altitude heaps of cloud can never exist and therefore Stratocumulus isn't ambiguous at all". But no, Cumuliform clouds can exist at low altitude, this is in fact common in the Summer time. What are those clouds called? Cumulus humilis. Standard abbreviation? No, it's not Ch, what, you thought *all* the standard abbreviations would be two letters long? You still expect consistency? They are "Cu hum". Are there any other kind of "humilis" clouds in existence? There are not. Okay, I have done some more reading, Cumulus humilis is a "species" of the "genus" Cumulus, which is why its abbreviation must begin with "Cu"; the cloud system is trying to imitate the Linnaean binomial naming system for living things, in the desperate hope of acquiring a scrap of scientific credibility by association. I'm not really sure what a Cumulus (Cu) cloud is. Obviously it is Cumuliform. It can't be the category "All Cumuliform clouds regardless of altitude", because high altitude Cumuliform clouds have the decency to conform to the system and are called Cirrocumulus (Cc). Oh, wait, no, sorry, that's wrong, Cirrocumulus clouds are not Cumuliform, they are Stratocumuliform. They *should* have been named "Cirrostratocumuliform", but apparently that form gets ambiguously abbreviated just like "Stratostratocumulus" does, even though it *doesn't* have silly sounding reduplication in it. You know what, for the third and perhaps final time I wash my hands of this dog's breakfast of a classification and resolve to simply observe and appreciate and try to learn to read clouds in the absence of names, which presumably is how it's been done for most of human history anyway.

I take back my earlier assertion that as a geek I struggle to let go of an instinct to impose a rigid and consistent systematic structure onto natural phenomenon which do not actually fit into one. I mean, I probably do struggle with that, but that is definitely not what is happening here, that is not the reason that every time I try to learn this system I end up refusing for the sake of my own intellectual dignity. What has happened is that in fact somebody *else* has tried to impose such a system, has done an awful job of it, and I can very quickly see how awful it is (yep, I guess I am also taking back my assertion that I know it's silly to blame my difficulties on the system itself). It's not that it doesn't correctly "carve nature at the joints" and identify cloud types of meteorological significance. It's just that it adds very little value over giving each of the types its own idiosyncratic name and in fact I would argue it makes learning your way around clouds more difficult than a totally unstructured naming scheme would. The whole 5x3 two-part naming scheme misleads the student into expecting consistency which does not exist. There are only ten commonly occurring cloud types in the first place, so only 66% of the 5x3 names the system generates could possibly ever find use, and learning 5+3=8 names is only very slightly easier than learning 10 names, so the whole thing is a dubious proposition right off the bat. And two of those ten forms are not confined to a single altitude but have substantial vertical extent (those would be Cumulonimbus (Cb) and Nimbostratus (Ns) - note with wonder and admiration that even the two "odd clouds out" are not named consistently, with "nimb" acting as a suffix in Cumulonimbus but as a prefix in Nimbostratus, why on Earth not?), which means that only 8 of the 15 names are actually needed. Even if it weren't for confusing similarities between "stratiform" and "strato/stratus" and "cirriform" and "cirro/cirrus", even if we really did have "type 1A" and "type 3C", the system would actually still be over-engineered and hard to justify over "type A" through "type J".

And yet, it has not been reformed for over 200 years. Does everybody who needs to use this system professionally hate their life? That seems unlikely. Maybe the real problem is with my approach. I am attempting to attain a thorough theoretical understanding of cloud classification indoors in front of a screen before going outside to apply my mastery. Maybe going outside with ten labelled photographs on a "good cloud day" and trying to match things up would yield much better results. "Stratocumlus" induces rage if you know that it is supposed to be part of a system where "strato-" is a prefix for altitude and both "stratocumuliform" and "cumuliform" are kinds of shape, but if you *don't* know that, if you just know "there's a kind of cloud called Stratocumulus, here are some photos of typical examples", there's nothing to be mad about. The existence of the system is still heavily implied by the names, of course, I'd eventually try to figure it out...but maybe then the urge to rage quit wouldn't come until *after* I'd actually learned something.

Cloud spotting blues

Replies