Transcription
All right, so everybody back. Let me just first say, at the outset, we spent a month in part—what Jacob? Yeah, we spent a month in part because the assumption was we screwed up. And, you know, so half that time was, no, we didn't screw up; we did it right. I just also want to plan on interpreting these plots. This is data from March; it's a March average; it's at 1:00 p.m. local time. This is also for March. This is an average, I think, of two years—I can't read the 2008/2009. This is just for an average of March 2009, so it's pretty one-to-one. We did not screw up; we did the right thing.
Does anybody have any suggestions as to where we should look to try and understand why the model fell apart? Like what diagnostic might we want to look at? And the clue is in the stuff we talked about in the first lecture. And I don't—my hearing is terrible—so if you have an idea, you have to scream it. That's close. That now, as it turns out, the Kzz for these two calculations is the same, but it's related to that. Okay, so the first thing to look at, based on a discussion, is neutral composition. What happened to our atomic oxygen and molecular NOx oxygen, molecular nitrogen, between the two simulations? Did we do something? And the answer is yes. So here's—um, hours of the beginning of the calculation, 700 hours. I must be like a month or something. And here's—it's—it's a column integrated, but it's essentially a measure of O and to the top curve, and it varies, I guess, like a diurnal variation or something. The top curve shows it's basically constant. So at the end of the month, you're oscillating around 0.7. All right, in the idealized case, our new realistic one, after about five days it went down to a new equilibrium, and you see this equilibrium is much more variable because his real weather in there, but it's definitely a different space than the idealized run. So something we did changed the neutral composition. So at least the ionospheric response is consistent with our neutrals. So at least that much is correct.
So now the question is, what changed the neutral composition? Suggestion was Kzz. Now, as I say, it's constant between the two. The key thing to remember is that with these realistic models, you're resolving other things that are historically bound up in this old one decays easy parameterization. So what we looked at next was the vertical wind, and we took spectra of the vertical wind at the equator. And on the right is the old standard input wave number for either diurnal cycle, a semi-diurnal cycle, terdiurnal cycle. And on the left is the vertical wind spectrum for the case with the realistic meteorology driven by Bell Gaps. And the standard NCAR is using an idealized tidal model, which just has two cycles—once per day, twice per day. You see on the left there's a very rich spectrum. Oh, and the other thing is that the once per day is going as a wave number one, so it's one sine wave across a latitude circle; it's traveling westward with the Sun. That's called a migrating tide. The ones that go with negative wave number are going the opposite sense; they're called non-migrating tides. I'll talk more about that shortly. So you see that the vertical wind spectrum from the realistic case is a very rich spectrum; there's a lot more going on. We have just greater upwelling. So Kzz aside, we have something else going on that's pulling molecular-rich air up into the thermosphere and flushing the atomic oxygen out. So you can't use the same Kzz because now we're resolving more with the realistic weather model. Okay, and that's—so to compensate, we have to change the Kzz, and we have to lower it.
All right, this is—ah, I don't know if you can see all the plots, but what I'm trying to show is a number of model runs with different colors as a function of—the NCAR guys like to use these log pressure levels—and there's altitude scales on the right. This is roughly the hundred—hundred and ten kilometer region. And the point is is that the old models—the OL was too low. When we went to the case—the only case where we used the realistic meteorology and the Kzz, which was lower than the idealized case, we got this red curve. And this is some observations from SABER. It's not great, but this curve's the only one that even comes close to getting the mesopause atomic oxygen correct. You look at the O2, N2, you'll see that this red curve is different. Then it actually—this curve brings us back to the default case. So rich meteorology, location Z, brings us back to the high Kzz limited meteorology case. So the black and the red are supposed to be the same. And if you go down there, you see that our ionosphere—I'm gonna have to go right—he's now out. So again, the bottom boundary at 95 kilometers, the F2 peak 250 kilometers higher. So surreal. I think a good demonstration. And since the bottom boundary, this Kzz is all sorts of meteorology, which we'll discuss, bound up in it. It's a good example of sort of this lower to upper atmospheric coupling. So we had our original standard NCAR, we had the bad model, we have the improved model. It's maybe not as great, but we do a lot better. And you're beginning to see that we're still beginning to resolve some of this—the two ionospheric arcs in the improved model, which we were not doing at all on the bad model. Okay, so the summary is that including the more realistic meteorology obviates the need for an ad hoc high Kzz value, which again, this is something that people have been playing with for 50 years.
So now, for the rest of the talk, when I talk about what is more realistic meteorology, what exactly is in there? I mentioned something about tides. That's one class—planetary scale waves, large waves, gravity waves, small buoyancy waves. We'll talk about all three. I do want to just mention this—some of the structure of the Earth's ionosphere, which I know you've talked about—why there are the two separate stripes centered around the magnetic equator. I always like to show this picture because it's an NRL camera that was placed on the moon by the Apollo 16 astronauts, and it's a measure of—it's the far ultraviolet airglow, and it's a measure of the recombination of O+, and I think you've seen a version of this slide before, showing how an electric field can be generated from collisions between ion and neutrals in the lower thermosphere, the E region of the ionosphere, that can cause a drift in the presence of a perpendicular magnetic field. And then at high altitudes, you have diffusion down along the field lines, and you have the recombination. So this recombination releases the airglow that they saw on the lunar camera, and you see it's—it's about 20 north or south and the magnetic latitude away from the equator. So that's what the two stripes are. When I was looking at vertical wind, I was looking here, but you see that this is coupled to the ionosphere of peaks away from the equator. And this doesn't happen at all on Mars or Venus because Mars and Venus don't really have a magnetosphere. And this is supposed to be an artist's conception of the magnetic fields or the interaction of planetary atmospheres with the solar wind—Earth, Venus, Mars. And so you might say, at first glance, like Mars is singularly unexcited; it's just this little ball of rock with the solar wind blasting past it, whereas the Earth has got all this complicated magnetospheric physics. Now, this is why the question of planetary escape is interesting for Mars, because the fact that the solar wind can gain access to the Martian upper atmosphere is thought to be one mechanism for stripping off gas from the Martian atmosphere. And so that's one of the things MAVEN is going to look at.
Now it also turns out, if you look in a little more detail, that the Martian crust is partially magnetized, and that magnetization varies in certain places, and that mag—that magnetized crust can—can create what are called mini-magnetospheres. You see these little closed loops here, and that can actually cause a little bit of resistance to, for example, the solar wind hitting the Martian atmosphere. The cartoon here puts them in the southern hemisphere, and that's accurate because that's where there's more commonly—that's where the magnetism is—is higher. Satellites that have measured the electron density peak—when this is now latitude/longitude on Mars—as a measure of magnetic field—crustal magnetic field strength—there's supposed to be some orbital swaths going through here—this peak region. You see this spike in the Martian electron densities, and as we talked about, dissociative recombination is inverse with temperature. So if the electron density spikes up, the thought is that the temperature went up, that the dissociative recombination went down, and so that there's some sort of heating mechanism in these little mini-magnetospheres. And that's something that's going to be looked at by MAVEN more carefully. Why—what—I don't—I think this is—I don't know. I would have thought this was a measure of the—So, as I said, a reason for mentioning going back and talking about this little sort of electrodynamic stuff is—a more recent airglow images since the lunar observation have shown waves in the airglow. And this was a paper by Tom Emmel et al. And the waves in these—this airglow actually was probably the single thing that got the ICON mission funded by NASA. And the ICON mission's objective is to look at these waves. And he shows here—there's some dotted lines to link it to oscillations in temperature with the same wave number—let's see if this—And so the idea is that it's—um, tides. So we're gonna look at tides for a little bit. And I don't know if it's which slide it's on, but there is a chapter which isn't referenced in the—in—in the, like, the guide to the lectures, but it's in—in the text from several years ago by Walter Scheib that really goes into this and explains all this in depth. And I'm just sort of going to show some phenomena. He gives you the real derivations and explanations of all this dynamics, and so you should—you should look at that as a good reference.
There are tides. Ultimately, are linked to release—release of heat and with the rising and setting of the Sun. And to our most relevant, there's the absorption of sunlight by water vapor. There's also the latent heat release from tropical convection. And there's the absorption by ozone in the middle atmosphere. And the depth of the absorption, for example, ozone—it's a deeper, thicker absorbing layer—guides the vertical wavelength of the tide that's produced and the mode of propagation. So, as I said, there's the migrating tides, and that's from the solar—of the absorption of ozone, typically giving you a semi-diurnal variation. The water vapor heating giving you more of a diurnal variation. The semi-diurnal one tends to propagate up into the lower thermosphere more readily. Ah, and then there's non-migrating tides, which is the latent heat released, but these are locked to specific geographic regions because, you know, we have roughly three rainforests on the Earth—Amazon, Africa, Southeast Asia—and so that's sort of a natural wave 3 pattern. And that's thought to be linked to the—possibly the wave 3 pattern we're seeing in the tides in the lower thermosphere. That picture—the airglow showed wave 4, but that's an artifact of the sampling. It takes a day to get the samples, and then—so you're gonna get 4—4 oscillations in the time it takes to make the samples. These tides grow with amplitude exponentially due to the decreasing background density. So this is just a figure showing an attempt to use—this is not the WACCM whole atmosphere model; it's a Japanese model called GAIA. And they really tried to take the rainfall pattern—this is in millimeters per day—from the three wealth regions of convection and link it to temperature oscillations, wind oscillations, and ultimately electron density fluctuations in the thermosphere. They were only partly successful. The thought is is that it's probably not so simple as simply this wave 3 just mapping itself onto the lower thermosphere, and that these interactions between different waves and it's—it's a lot of this—some nonlinearity involved. But there's definitely something to—something to this. On Mars, remember I said that a lot of the math is, you know, the—the Coriolis is saying the seasons are the same, so a lot of the math is the same for a lot of these wave activities—be the wave modes and all that—but the forcing is different. So that's really what—what distinguishes the two. Ah, the absorption by dust, which can be a deep layer in the Martian middle atmosphere, kind of serves the role that ozone does on the Earth. Mars doesn't really have the O's—a little bit of ozone—doesn't have the ozone layer that Earth does, or it doesn't have the dust layer that Mars sometimes has. Both of them generate this semi-diurnal oscillation in the lower thermosphere. Then Mars doesn't have the latent heat release, but it does have a very strong diurnal heat cycle in the lower atmosphere, and that superimposed with the very large topography is what drives a non-migrating tides. In other words, they don't follow the Sun necessarily, and they have their own wave—I think it's more like a wave 3 pattern in the Martian lower thermosphere. So the aerobraking data show a lot of these oscillations pretty clearly, and there's a fairly rich literature looking at this—is longitude, this is mass density for four different altitudes, looking at these, and this is the Mars Global Surveyor data. So they had, you know, all these missions come in and to circularize the orbit, they have an aerobraking period. And so there's a fairly decent literature on decomposing these—the wave modes and trying to link it to the surface forcing on Mars. On the right, you see again—Mars doesn't have all this electrodynamic variability that the Earth does. What you see is the height of the ionosphere moving up and down, and again, it seems like there's some wave mode linked to that—that the thought is it's linked to tidal forcing from the lower atmosphere.
All right, so that's tides. Wave number 2 or planetary waves, also known as Rossby waves—large-scale. I used to think—I usually think of it for explaining to people like it's—well, it's the reason that the jet stream, you know, varies from north to south as you go around the globe. So it may be, you know, one—one sine wave, two sine waves in the troposphere. You can get up to about seven in the middle atmosphere; it's usually one or two. The key thing is that the restoring force here is actually the change in the Coriolis as a function of latitude. And so this again is where the rotation serves to generate this variation—Coriolis. And you can even see this in—spinning tabletop fluids—that the—the—if you—if you set the wave—if you—if you—if you have a tracing, you generate a wave in a tabletop fluid, the oscillation amplitude will depend on how quickly it's spinning, because that controls—that's the restoring force. Now the mode of the—the propagation of these things is to the West. That has to do with conservation of vorticity again. Okay, see here's the reference to Walter's tides book. So as a wave is moved to the north where the Coriolis is greater, to conserve vorticity, it'll spin down a little bit, and so you'll get a circulation anomaly—circulation like this. And then if it's moved to the south, it'll tend to spin up. The combination of the motion here going southward and a motion here going north wind will tend to push this wave to the left. And so Rossby waves propagate westward. Now I don't have a pretty picture of these Rossby waves, but this is a useful plot to see because it's a category plots that you often see when people talk about waves and their models; they're called Haurwitz plots, and there's an umlaut on one of those, so I'm probably pronouncing it wrong. But typically there are time—and the label here is Mars orbital solar longitude—in other words, what—where it is in the orbit around the Sun. All right, and so roughly 360 degrees in orbit; each degree is 2 days because the two-year year. First is geographic longitude, and this is thermal emission—this is sort of near the surface. And the idea is that if you look at a single longitude, you'll see a wave. And the tilt of these—so that gives you a sense of the propagation. Others, if the waves were stationary, there—there would be sort of just perfectly horizontal. So anyway, Haurwitz plots are something that show—show up a lot. I realize it's not the clearest figure. There are two types of planetary waves. There were sort of the stationary waves, which, for example, in the Earth's troposphere, ultimately triggered by land-sea contrasts in the winter. It's warmer over oceans than over land, and so that triggers a wave. Then there's sort of traveling waves, which are often more considered to be instabilities. And so I do have a pretty picture of a traveling wave, which we've gotten very interested in lately. This is from NOGAPS, and so it's the mesosphere and its water vapor. Okay, parts per million by volume. So water vapor in the middle atmosphere is a trace constituent. This is near 80 kilometers, and this is for January, so this is summer. And we know that summer—this water—the summer mesosphere is wet, and is upwelling makes it wet. Okay, so oscillations in the north-south wind will periodically cause that moist air to slosh northward and southward, and that's why you're seeing this wave 3 pattern. If I showed you this as a movie, you'd see these things moving to the left, which is westward propagation because this is a Rossby wave. All right, so the period here and the speed of this proximate propagation roughly gives you about a two-day period. So this is not a fast-moving wave; it's about two days from one peak back to the second peak. This is the same thing now. This is the stratopause again; this is water, and you can see here at the equator where there's a nice gradient again—north/south oscillations pulling in dry and wet air north and south. Okay, so this is a fairly well-known phenomenon in the middle atmosphere that you—you have these things; they tend to show up in summer more so in the south than in the north for—or, you know, arcane reasons related to the middle atmospheric circulation. One thing that's gotten a lot of interest lately, though, and this is a paper that was sort of ignored for like 20 years, is that if you look at the ionosphere, all right, and this is a measure of the peak ionosphere density for—this is actually March, which is sort of an unusual time of year to see these waves—you see that there is this two-day oscillation in the peak ionospheric electron density. Moreover, when the oscillation kicks in, the overall density is lower. So the question then is, how can we link this mesosphere wave phenomena with the ionosphere? Well, we go back to the same kind of fundamental diagnostic that I've been talking about all along. We want to go back and look at what might the neutral composition be doing, what might the O and to be doing as a function of this wave activity. So on the left is my calculation from NOGAPS linked to the TIEGCM showing—narudy onyl wave amplitude. On the right is a calculation from—it's just recently published—I'm gonna say Ciao, you weigh it al. And so the left calculation shows that from NOGAPS we can—we can take the two-day wave and map it up into the lower thermosphere, and even in the tropics there's a two-day oscillation. Remember that's where our vertical wind affected the entire ionosphere. And what he showed is that for periods when the two-day wave is active that you—this oscillation will tend to bring up more molecular nitrogen at the expense of atomic oxygen, and that's consistent with lowering the height of the electron density. So it's another example of a middle atmospheric phenomena linked to tropical upwelling and composition. So this is—this is the same figure—um, this is from the NCAR—yeah, you weigh in Wang here—there's the reference—the NCAR time GCM reducing the electron density for the same reason that Kzz did in our simple little 1D toy model. For stationary planetary waves, which I talked about—land-sea contrast—probably the one that's upper atmospheric people have heard of—it's the most famous—is associated with stratospheric warming—z. So in winter—ah, the lower stratosphere—it gets very cold, and so the air just spins around—this is the polar vortex that all the weather people are making a big stink about this year. It's really a lower stratospheric upper tropospheric phenomenon, but this polar—circumpolar vortex will often be disturbed, and you can have a wave interact with the vortex. There are sort of two ways that can happen. This is a wave one displacement of the vortex where it's shifted off the pole.
And there's actually a little wave breaking here, or you can have a wave too, which is a split in the vortex. So wave to me. If you went around a latitude circle, you'd see two peaks again. The middle atmosphere people have been all over this for decades, but it's only been in the last five years or so.
Alright, so this is just more middle atmosphere with no gaps that really, you know, extended the weather up to 95 kilometers. We've looked at all sorts of effects on, for example, the temperature. When when the planetary wave comes in, it's mixing in air from lower latitudes, which is warmer than this cold vortex. So the temperature at the pole goes up, and that's why it's called a warming. And at a certain altitude, it's really a decrease in the stratopause altitude. So we simulated a lot of this, the changes on wind, the planetary wave amplitudes, but you know, this is a useful plot because in all these stratospheric planetary waves that are causing the stratospheric void pull a polar vortex to change, and the stratospheric warming's generally thought to be filtered out before they get to the thermosphere. It really was thought to be a middle atmospheric phenomenon, so it was sort of surprising that during one of these warming's, they saw some pretty pronounced ionospheric effects.
And so the questions lately is like how to explain that. This is data from Jicamarca, so it's the equator; it's not even the pole, it's the equator, and it's the vertical drift, which has a nice, well, sort of understood diurnal variation. This is the control days, and then during the period of the warming, you see this diurnal variation is completely disrupted. Well, when we think diurnal variation, we should think tide. So we've got planetary wave, which is giving the stratospheric warming affecting a tide in the tropics, and that's sort of where the the cutting edges is, try and link those two, and and and this is not only neutralized; this is the electrodynamic tide, so it's changing the dynamo. A year later, they looked at electron density, and they also saw some dramatic effects in the diurnal variation. So on the the top is the total electron content for certain times of the day from a satellite data, and again this is showing the Jicamarca radar observatory. And there's more of this peak in—let's see if I get this right—local time, and this is 16 local time; this is the universal times are up there, so it's the local time of Jicamarca. So morning there is, during a big stratospheric warming, the afternoon electron density goes away completely perturbed. The bottom plot shows this vertical ion drift, which is, you see, for the day of the warming completely different than climatology.
So the models have been looking at this, and there are some ideas there, getting some wave-tide interactions; there's this temporal arousal, and but what he shows that this is the semi-diurnal high-amplitude decreasing during a warming; this is a terdiurnal tide, three per day, that seems to be dramatically increasing. Not every model gets this, but they get something; they get some tidal interactions with the warming. So that seems to be the area of nonlinear interactions between a planetary wave and a tide in the mesosphere, and then the tides can propagate up into the thermosphere even if the planetary waves themselves can't. Yeah, they've done some intercomparisons, and everybody sort of has their own model that seems to be enhanced; seems to be how you handle some of the small-scale gravity waves because that controls how your what your winds are in the middle atmosphere, and it seems like that's the the key thing for getting the right tidal mode to propagate.
So this will give us a chance to turn to the last wave, which is gravity waves, which is the ones that really are the big uncertainty. These are vertically propagating waves where buoyancy is the restoring force, and you know, if there's just a sort of a small conservative oscillation that's not too exciting, but they grow with height, and as they get to a certain height they get unstable and a little return, and you can just think of waves washing up on a beach; that's what transports it. Can it crash against as the wind flow and change the the angular momentum budget? It can also transport constituents. These are too small to or to wrap it to easily resolve, so this is where the residual Kzz parameterizations are all focused on. Ah, there were three sources, and these sources actually work on Mars as well. There is this is a longitude of and and it's a units of like milliPascal, is a measure of gravity of the flux of the vertical flux from small-scale waves, and this is I think for July, so this is the Indian monsoon, so convection. Now this is not latent heat release; this is mechanical; this is convection, cumulus towers slamming into the stable stratosphere and generating ripples that propagate up to the mesosphere. Mars has convection; doesn't have latent heat release, but Mars's convection there is topography. So the winter winds hitting the Andes got to get over the Andes, and that triggers mountain waves that can propagate up. We'll call that on Mars the mountains are three times higher than on Earth, so you can see where gravity waves on Mars are gonna be three time could be three times bigger, all else being equal, than Earth. And then harder to see this, but shear in the like baroclinic instabilities in the jet stream, which is something that is a governs a lot of tropospheric weather, and that also happens in the Martian atmosphere. So that's the source.
The thing that's interesting is that these waves will grow, and one way to express that is we are to write the vertical flux of zonal momentum as due to these terms. Okay, so the idea is that this guy on the left is constant; this is the density; we know that goes down. Okay, so if this is going down and these guys got to go up to compensate, this is the stability, the Brunt-Väisälä frequency to zeroth order; we'll call that constant; this is the zonal wind and the phase speed of the wave; these this could vary, and if this gets close to zero because the wind the the the wind is near the phase speed of the wave, then this amplitude goes way up, and that's where the wave will will get filtered out. But all else being equal, if this goes down, the amplitude has to go up, so this variance increases as one over Rho. So when you keep that one over Rho thing in mind because if you are measuring graph if you have a measure of gravity waves that's what you want to gravity wave variance from what, let's say, density fluctuations or whatever it is, and you have it as a function of altitude, you want to look at that and see is it increasing by one over Rho? Because if it's not, then there's probably dissipation, it's, and that's what they found on Mars.
So this is the accelerometer and accelerometer data from Mars Odyssey; it's a paper by Fritz et al. This is density variances, so this is going from 100 to 140 kilometers, and if you look at this region here, the variance is increasing by about a factor of 5 or 10, but this is over about 5 scale heights, e to the minus 5 says it should have gone up by a factor of 150. So yeah, it went up, but it only went up by one thirtieth of what it should. From that difference, Fritz et al inferred significant dissipation. Long story short, you compare that to Earth; you you express that as sort of an acceleration, meters per second per day; Earth is somewhere is between, depending on the sign, 100 is probably the max, and then there's a model on Mars, and it's like 3 to 600. So this then may be the reason that the homopause on Mars is higher than on the Earth because the gravity wave fluxes on Mars are several times, many times larger than on the Earth, and that's really coming back to again the large topography on Mars. So we can link the large topography, the gravity waves to the homopause altitude, and that governs the thermosphere composition everywhere is as you've seen from our Kzz experiments.
So I think that takes this full circle where we really now have considered everything that—not everything—but we've got sort of a least crude outline of a picture of ground-to-space coupling. We've looked at the three waves; they're all present on both planets, but the forcing is different in general. The effects of gravity waves and getting to it on but tides as well is greater on Mars, most likely due to the topography. Martian data is limited to the aerobraking, although they did, and we're gonna get some composition of meteorological data, but the aerobraking can show us a lot; they can show us gravity wave variability, and they can show us tidal variability. Ah, there's some factors unique to each planet, so we can go back, and I hope we now have filled in at least conceptually some of these differences. We talked a lot about scale heights, okay, get into this explicitly except of the fact that the dust storms really occur are in some sense related to the large eccentricity of the Martian orbit. We talked about the Coriolis, I took two, magnetic fields, go back, heating efficiency. So I think we've covered pretty much all of these at some level. So what I'd like to leave you is some food for thought: play God to amuse yourself, because Gods get bored. This is you're one of these like Greek gods who like a row; he's causing mischief, right? You decide to chop off half of the tops of all the big Martian volcanoes; what might that do to the Martian thermosphere, the Martian ionosphere, and to the spacecraft orbiting Mars? So I'll stop there. [Applause] I didn't hear, so I was there a question or you everyone grow three four? So we're talking about your weight; they're there any evidence of classroom a puppy or well that's another question is a conference if there is evidence the thermosphere of gravity waves? Okay, certainly the awarded generator song última hora; there's still a suggestion that so the treasure gravity waves breaking the mesopause, which for the thermosphere is maybe enough because it changes the mixing in the homopause on the lap, but then is this thought that there's actually gravity wave breaking it secondarily; the modeling connection actually opt into the thermosphere on footage like a secondary effect, like a two-step process. But then there's a question if some people say these secondary waiters they break are they important? No, madam, I have won the issue very vicious so the you know wait purely a here. No Frank, just a few. [Applause]