Slashdot Mirror
Airplanes May Affect Weather Patterns
blankmange writes "Wired is carrying an interesting piece: '...for three days starting last Sept. 11, meteorological researchers were presented with just such an opportunity when the FAA grounded commercial flights nationwide for three days following the terrorist air attacks. And now it has emerged that the American climate was indeed noticeably different during those three days without air travel.' Seems that what we do on the planet may have more effect than we may ever know."
This really isn't anything new.
Besides the fact that 3 days is too little to check, it has been studied and theorized how butterflies in Africa, or bats in Brazil can cause chaotic effects on the weather.
Mountains, trees, ducks, all cause effects on the weather. There are a lot more birds and bats in this world than planes. They probably have a larger effect on the weather than planes do.
If we really wanted to see how planes effect weather patterns, we can always try to go to another planet (Mars, is a good choice) and study their weather. Mind you, trying to get air planes to Mars would be a real pain to test this theory.
How about hydrogen powered jets? The exhaust would be water vapor.
It's my understanding that contrails aren't smoke from the jets, but are ice crystals formed in the chaotic vorticies of air spinning off wings.
That is, the high-speed wing creates even higher-speed whirlwinds, and the moisture in the air, when caught in these whirlwinds, freezes, leaving an opaque white trail.
I'm guessing that these "expand" into large cloud banks less by spreading and thinning than by catalyzing the creation of more clouds in adjoining air -- the frozen moisture cools the air around the contrails, and can cool it just enough to allow more clouds to form, etc., etc.
Of course, I'm just pulling all this out of my butt -- will a real meterologist please stand up?
Around here, it's been noted that the temperature differential of the city causes some storms to be deflected slightly as they go past.
I live in Amarillo, TX. While we're certainly nowhere near the size of NYC, we do have an interesting condition that exaggerates this effect.
Amarillo (pop ~= 300k), you must understand, is an island of glass, concrete, and asphalt surrounded by nothing but flat prarie and even flatter pasture for a distance of at least 150 miles in any direction. The only major body of water any where near the city is more than 50 miles away.
Therefore, it is noticably hotter and windier inside the city than outside. While we do have severe weather being on the lower end of tornado alley, it tends to 'part' around the city. The last damaging tornado we had (You may have seen it on CNN since it killed a couple people) was listed as striking 'near' Amarillo. What really happened was that the tornado ran through the tiny burg of Happy, TX, which is very nearly forty miles to our south. Then, when it started aproaching the city of Amarillo, it veered off and started heading Southeast rather than Northeast. The bulk of the storm itself did the same thing... Head northeast until it came up against Amarillo, and then push off to the southeasth.
The next Slashdot story will be ready soon, but subscribers can beat the rush and slashdot the links early!
... and a very poorly understood one, at that. Urban areas do tend to create what one might consider shallow, localized fronts.
It's well known that any atmospheric boundary can have effects on a storm, both in motion and intensity. It's also well-known that urban areas do have localized differences compared to their environment.
But it's not well-known what the overall effects might be. Severe thunderstorms are creatures that require moisture and the energy condensing it provides. In an urban heat island situation, the temperature is often higher, but the amount of moisture is only marginally higher. This actually tends to reduce the CAPE (Convective Available Potential Energy) of the urban area compared to its immediate environment. Also, the boundary of the urban heat island may act as a sort of "guide" for the storm to propogate on, much like fronts and outflow boundaries (cold, moist air from previous thunderstorms) often act.
But, the storm is also affected by the upper air steering flow. There's a certain slaving between the upper and lower levels that is beneficial to storm development and intensification. An urban heat island may modify this enough to make a difference... that much is possible.
However, in the case of your storm switching direction, that's actually a pretty common feature of tornadic storms, i.e. to change direction and move to the right of their previous movement vector. The dynamics of right-moving supercells (as they are called) are fairly well understood and widely accepted. So, to say that the storm turned strictly because of the city, in this case, is a little hard to believe. But, it's something that bears watching.
And I have noticed that "effect" of cities before. It does seem to be relatively common, as a sort of informal survey... but I'm not aware of any study that's fully addressed it yet.
-Jellisky
First of all, my biggest pet peeve: the difference between weather and climate. Yes, believe it or not, climate and weather are two very different terms. They should not be used interchangably, thankyouverymuch. Three days does not a climate make.
Now, to continue on that thought, it's interesting to think about the consequences of this "study". (The quotation marks are supposed to be there.) I cannot say what I really think of this "study" since I haven't read their results, but I cannot be that easily convinced that three days worth of data compared to years worth of data has any possible statistical significance, especially in something like diurnal ranges, which are 15-30 degrees C anyways. Show me three months and I might be convinced of a trend. I can name at least 10 different three-day weather features that could cause such a blip and that's just off the top of my head.
Next, IF this is true, then it only highlights something which has bothered me about climate modeling from the frickin' beginning: the role of clouds and how terribly they are handled in these (and all) models. Of course, this isn't the only problem I have with these modeling studies, but we won't enter that debate right now.
The type of cloud present has an effect on the net change in radiative flux. Deep, thick clouds (like cumulus) have a net positive change while thin clouds (like cirrus) have a net negative change. The thing, though, is that in balance calculations like these, there tend to be two effects, which are approximately equal and opposite in sign. So, you end up wondering how much of it is really real. (For example, the two terms might be 220 and -218, leading to 2 change... but if you're off by a little bit, those numbers might actually be 219 and -220, leading to a -1.) This is further compounded by the way models handle clouds, which is often routinely terrible (with respect to resolution, the actual physics involved in the cloud which can affect all the results, and many other factors).
To further put all this in perspective, let's assume the albedo (the amount of solar radiation reflected back to space by the earth which is largely a function of the cloudiness) of the earth increases by 1%. (It's currently around 30% in a climatological sense... even that number has an error bar of measurement around +/- 3%.) On average, that would mean that the earth would get 3.4 W/m^2 less radiation. (Daily and spacial average of solar radiation is about 340 W/m^2, again largish errors on this measurement.) This number is comparable to the change by doubling CO2 (about 4 W/m^2) and, as you can notice, opposite in sign. Of course, there's a huge extra batch of physics here that isn't even being considered like the change of the absorption of IR radiation from the earth by the clouds or the release of latent heat by the clouds or the feedback between warmer surface temperatures and clouds (which is barely understood since it's almost as complex a problem as the original)... Kind of makes your head spin to think about all these effects, doesn't it? And all of them are about the same order of magnitude by itself, i.e. about 0.5-5 W/m^2, both positive and negative. Let's also not forget that local effects, like all those new urban heat islands that have popped up around all our temperature recording stations that could very well explain that temperature rise in the last century or whatever, and that these effects are not put into these models...
Complex problem? You bet. Possible to understand? Eventually, I don't see why not. But, we can't sit back and keep using these antiquated ideas in these state-of-the-art models. As the old saying goes, "Garbage in, Garbage out." The effects of these contrails may be important, yes. I cannot debate that. However, to claim that off of whatever insignificant sample this is, or using any of the ideas we currently have, is ludicrous at best. Any imbecile with a computer nowadays can run a correlation analysis on data. But, to interpret it and explain WHY things are happening that way... that's the vital connection between statistical tomfoolery and real science. Then, to explain the dynamics and theory behind it all... that's the step to making a full fledged theory.
-Jellisky