the Air Vent

Because the world needs another opinion

Projection , Regression

Posted by Jeff Id on December 30, 2010

You would never believe what the government will fund and peereviewlitrachur will approve of.

Title: “A scaling theory for the size distribution of emitted dust aerosols suggests that climate models underestimate the size of the global dust cycle”

regressomatic.

wow.

30 Responses to “Projection , Regression”

  1. Duster said

    So, now we know officially (the NAS says so) that small particles stay in the air longer than big ones; that the smaller the particle is, the more of a cooling effect it has, and that subaerial (sand actually bounces along under most wind conditions, rather than becoming suspended in the air; bounces are proportionate to wind speed and inversely so to grain size) transportation of sand causes tiny craters in dirt (as well as in the paint and glass on your vehicle) and therefore makes more dust. Every dust storm will have more of a cooling effect than a warming effect since there are a lot more small particles that will hang around, and, most far out, the iron transported out to sea is an important nutrient for algae that fixes carbon. That’s two cooling effects to one heating effect (if you think CO2 has any warming effect to speak of) and a net reduction in CO2 was well. Not to mention that the visible light bound up in the formation of complex carbohydrates by the algae is not going to be immediately re-radiated as infrared band radiation. Perhaps Kevin Trenberth’s missing energy is lurking at sea in the pelagic zone.

  2. Brian H said

    Why do I suspect this study is trying to throw dust in our eyes?

  3. M. Simon said

    The research, by National Center for Atmospheric Research (NCAR) scientist Jasper Kok, suggests there are several times more dust particles in the atmosphere than previously thought, since shattered dirt appears to produce an unexpectedly high number of large dust fragments.

    I dunno but wouldn’t it be easier just to take some measurements?

    In any case there is a computer model of what is happening and that can be used to calibrate (adjust?) the computer model of what is happening. We are very lucky.

  4. omanuel said

    NAS and government employees in the federal agencies that NAS controls have been “up to their ears” in deception for decades.

    In a UPI news report on the Iron Sun by Dan Whipple (6:50:07 PM EST July 17, 2002), David Hathaway, solar physics group leader at NASA’s Marshall Space Flight Center, said:

    “This is crackpot science. We’ve got information on the composition of the Sun from a variety of different sources … There’s no way its mostly iron. We would have known that a century ago.”

    Dr. Hathaway was apparently unaware that Fred Hoyle admited in his autobiography that he and other astrophysicists almost all believed that the Sun was mostly iron until the end of World War II.

    WUWT and Tallbloke’s Talkshop recently commented on the remarkably different attitude coming from NASA after 2002 when solar cycle 23 was ending:

    a.) Nov 12, 2003: “The Sun Goes Haywire – Solar maximum is years past, yet the sun has been remarkably active lately. Is the sunspot cycle broken?”

    http://science.nasa.gov/headlines/y2003/12nov_haywire.htm

    b.) Oct 18, 2004: “Something strange happened on the sun last week: all the sunspots vanished. This is a sign, say scientists, that solar minimum is coming sooner than expected.”

    http://science.nasa.gov/headlines/y2004/18oct_solarminimum.htm

    c.) May 5, 2005: “Solar Myth – With solar minimum near, the sun continues to be surprisingly active.”

    http://science.nasa.gov/headlines/y2005/05may_solarmyth.htm

    d.) Sept 15, 2005: “Solar Minimum Explodes – Solar minimum is looking strangely like Solar Max.”

    http://science.nasa.gov/headlines/y2005/15sep_solarminexplodes.htm

    etc

  5. Steve Fitzpatrick said

    Duster,
    “the smaller the particle is, the more of a cooling effect it has”
    The scattering profile of a particle depends on it’s size and refractive index, with size the dominant effect for most kinds of particles in air. The angular dependence of scattering (how much intensity is scattered at each angle) is very strongly dependent on size, with smaller particles (much smaller than the wavelength of light) scattering in all directions almost uniformly, and larger ones (comparable to or larger than the wavelength) mainly scattering in “forward” direction… that is, the scattered light is closer to the original direction of the light.

    I do not know if this particular study presents any useful information, but I suspect that knowing the size distribution of atmospheric aerosol particles is something worthy of study, since the increase in albedo from atmospheric aerosols would seem to depend on the size of the particles.

  6. Carrick said

    I think it has been pretty well studied, SteveF. In the case of most aerosols, because of their size they predominantly scatter light backwards. The size matters because very larger particulates are easily washed out by precipitation (probably the source of Derek’s comment,, not sure). Whether the particulate absorbs water vapor (is hydroscopic) matters as does how effective it is a seed for clouds. Some of these aerosols absorb in long-wave regions that overlap gaps in the H2O spectrum and thus play a role as a greenhouse gas. There is also an interaction between some of the particulates (especially the sulfates) and the biosphere (they act like an “antifertilizer”).

    If they are light (so they make it up higher into the atmosphere) and break down under UV radiation, then there are even more interesting things that happen.

    It’s a complex and fascinating topic , IMO (starting with how the atmospheric boundary layer influences their initial diffusion into the upper portions of the troposphere).

  7. stan said

    Peer review is magic. [but see, http://breast-cancer-research.com/content/12/S4/S13 per the Bish]

  8. mrpkw said

    “regressomatic” by Ronco !!!

    But Wait !! There’s more !!!
    Oder now and we will send you a 2nd regressomatic for FREE !!!

  9. mrpkw said

    OK, beyond the comic relief, it is nice to read part of a paper about climate that did not mention “CO2″ and completely blame mankind.

    I know that a volcano can put up massive amounts of particles that can and do cool the planet, but how big of an event does it take to show an effect?

    Would the 1930’s American dust bowl have done anything to climate/temps?

  10. Steve Fitzpatrick said

    Carrick,

    I do not known the size distribution of aerosols in the atmosphere, but particles that are very much smaller than the wavelength of light scatter in all directions almost equally (at least according to Mie calculations, which I believe are correct). Larger particles (as I said earlier, comparable to the wavelength) scatter very strongly in the forward direction, with minor back-scatter compared to forward scatter. Near macroscopic particles (like cloud droplets) do back-scatter quite a lot of light, but I doubt many macroscopic aerosol particles survive in the atmosphere for very long, though I could be wrong about this. The references I have seen (they use the ‘angstrom ratio’) which suggests the particles are (on average) relatively large, and so scatter forward much more than backward.

    I suppose I could do the Mie calculations to estimate and generate the back-scattering profile for a range of sizes and refractive indexes, but I expect this is already well known. What is probably not well known is the size distribution and refractive index distribution.

  11. Steve Fitzpatrick said

    Carrick,

    I forgot one thing. Clouds reflect mostly via multiple scattering, which allows strongly forward-scattered light to get ‘turned around’ through multiple interactions with different particles. At very low concentrations (like aerosol particles), particles in the size range of cloud droplets scatter near 100% forward.

  12. Frank K. said

    What I find fascinating is the amount of money and effort that institutions like NCAR and UCAR put into press releases. Here’s the original press release:

    http://www2.ucar.edu/news/3510/broken-glass-yields-clues-climate-change

    In the upper right hand corner you see the media people involved:

    Contacts for This Release
    For Journalists

    David Hosansky, Head of Media Relations
    Rachael Drummond, Media Relations

    I suppose this is the way you get more government funding to feed the money-hungry climate research machine…

  13. Terry said

    Steve #10

    Particle size distribution of aerosols in the typical mixed layer has been well characterized for many years, especially in urban atmospheres, and of course it varies considerably from location to location. What is not so well known is the distribution for the upper layers. You are also right in that it is dominated by small particles, less than 1µm aerodynamic diameter and in terms of particle numbers (not weight) is dominated by very small particles less than 0.1µm, but again the actual size distribution is not well characterized on a global scale.

  14. Kenneth Fritsch said

    I am not at all sure, that in all the many ways that governments waste our money, this one would be among the worst examples. I am attempting to comprehend what this research is all about. It appears that scientists have found evidence that dirt/soil particles formed by abrasive forces can produce more (smaller?) particles than previously estimated. These particles can potentially become airborne and there have an aerosol affect on the climate. I am admittedly getting a bit slow of mind but is the objection because it is judged that this measurement could be made more effectively in a more direct manner.

    If the above thesis is correct and the climate models do not account for the correct number and size of particles and yet get the climate “right” what would that say about the models? Does it mean that the aerosol effect is parameterized to whatever level fits the climate and thus this thesis has no bearing on the final result that the models produce? Or do indeed the models base the aerosol effect on some first principles that would require adjustments given that the atmospheric particles are misrepresented presently? In the end does it say something about the uncertainty of the model results based on the models not being sufficiently comprehensive?

    Could such research lead to an inexpensive means of seeding the atmosphere to cool it down if required? Would researchers ever dare use this rationale (potential beneficial man made climate effects) for raising research funds in today’s politically charged environment that frowns on anything man made or influenced?

  15. Terry said

    kenneth #14

    Yes folk have already postulated this in the past, based on the already known fundamentals of particle science. The idea was to release humungous amounts of SO2 to form SO4 aerosol in the upper atmosphere and thus effect a cooling. Didn’t really get past 1st base as a realistic option.

  16. Carrick said

    SteveF, here’s what they say about aerosol particles at NOAA (emphasis mine):

    Introduction
    • Aerosols are suspensions in air of microscopic solid or liquid particles.

    • They are emitted into the atmosphere both naturally (e.g., sea spray, volcanoes, forest fires, dust) and as a result of human activities (e.g., fossil fuel combustion, biomass/refuse burning, land use changes). They are also created and modified by chemical processes in the atmosphere.

    • Aerosols consist of many different chemicals. Their composition and size affects how they absorb or scatter light, whether they are liquids or solids, whether they are acidic, neutral, or basic, how much water vapor they absorb, how effectively they act to form cloud droplets, and how long they reside in the atmosphere.

    • The average lifetime of aerosols in the atmosphere is about a week. This lifetime is much too short for them to mix all around the Earth and hence aerosol concentrations are highest near their sources.

    • Aerosol particles profoundly affect Earth’s climate.They mostly reflect sunlight, and thus cool Earth’s surface. They therefore counter the effects of greenhouse gases, which absorb the heat escaping from Earth’s surface and thus heat Earth’s surface and lower atmosphere. The cooling by human-induced aerosols is roughly comparable to the heating by human-induced carbon dioxide. However, dark colored aerosols (soot) absorb sunlight and heat the atmosphere.

    • Aerosols also modify cloud amount, cloud distribution, and cloud properties such as cloud brightness. Aerosols may also influence precipitation amount, distribution, and frequency. Aerosols decrease visibility by scattering and absorbing sunlight. Aerosol deposition to the surface alters ecosystems by changing acid and nutrient balances.

    • The major aerosol types are sulfate, black carbon, organic carbon, mineral dust and sea salt.

    • The atmospheric Physics and Chemistry group is working on these different aspects of aerosols, and a summary of our research on aerosols is presented on this page.

    If the average particle size is 0.1µm, that works out to about 2x the wavelength of visible light. If the primary mechanism for scattering is Rayleigh scattering, the cross-section for that goes as (d/lambda)^4. So you’d really need to know the distribution of particles to know how much is forward scattered versus backward scattered.

    (This is ignoring absorption of light of course by the particles.)

  17. Steve Fitzpatrick said

    Carrick,

    I have spent much of the last 20 years thinking about the scattering of light by small particles; my business depends on getting it right. The NOAA site is targeted for general (non-technical) consumption.

    “If the average particle size is 0.1µm, that works out to about 2x the wavelength of visible light.”

    If the average size is 0.1 micron, that is about 20% of the wavelength of light and is in the transition region between Rayleigh scattering (completely random in all directions) and directional scattering (with a strong forward directional effect). 0.1 micron has relatively strong scattering at high angles, but is still weighted heavily in the forward direction. BTW, 0.1 micron seems very small based on what I have seen with published “Angstrom ratios”…. closer to 1 micron seems more reasonable; and would in fact be ~2X the average wavelength of visible light… and with extremely strong forward directional scattering.

    “They mostly reflect sunlight, and thus cool Earth’s surface.”

    Well, that depends on what they are and what their size and concentration is. Absorbing particles (black carbon/carbon black) most certainly do not reflect much light; they strongly absorb at all wavelenghts, independent of particle size. Hydroscopic sulfate aerosols are spherical droplets which absorb little, but which may (depending on size) scatter strongly anywhere from low to high scattering angles.

    “Aerosols also modify cloud amount, cloud distribution, and cloud properties such as cloud brightness.”

    Completely true… if the aerosols substantially reduce the average diameter of cloud droplets, and so increase the fraction of high angle scattering.

    The reality is that optically thick clouds some fraction of incident sunlight, transmit a (modest) fraction, which shows up as ‘diffuse’ light at the surface, and absorb a fraction. Once light enters an optically thick cloud, it is scattered many times, and so becomes ‘randomized’ in direction, bouncing about until it either escapes or is absorbed. The net path length of light in a dense cloud is VERY long due to multiple scattering, and so even though water is mainly ‘transparent’ in the optical region, the light inside a cloud has such a convoluted path that much of the light passes thought a substantial thickness of liquid water (or ice), and most wavelengths, especially longer ones, end up being absorbed to a large extent because the light passes through so much liquid/solid water.. light passing through a dense cloud is very much like light passing through a substantial layer of liquid water. (Remember that below ~100 meters, the ocean is pretty dark!) Near infrared and red wavelengths are most strongly absorbed, so the base of thick clouds have a blue-violet tint… the red part of the spectrum has been mostly absorbed. Near ultra-violet is hardly absorbed at all, so you can still get a bad sunburn, even on a cloudy day. (Side note: I never realized this color shift was there until I watched a talented artist in Holland painting clouds in the sky… he could see the color shift that I had never noted.)

    Thin clouds (cirrus) scatter very little light, and what they scatter is mainly in the forward direction; they do not reduce net surface intensity of the sun very much. On the other hand, thin clouds are strong infrared absorbers, you can think of them as ‘black carbon’ for the infrared.

    So the net is this: Yes, aerosols are important. How important depends on their size, their optical properties, and how they change (or do not change) the optical properties of clouds. Thick clouds are reasonably strong absorbers of solar energy, and lead to both substantial reflection of sunlight into space and absorption of solar energy well above the surface. Thick clouds are (of course) also strong absorbers of infrared, and so also inhibit loss of infrared to space.

    Overall, it is very complicated, and it seems to me unlikely the models have it right. The NOAA web page does not do the subject justice.

    If you want a good starter on scattering: “Absorption and Scattering of Light by Small Particles” C.F Bohrenm, D. R. Huffman. A classic.

  18. Steve Fitzpatrick said

    Sorry, that is C. F. Bohren and D. R. Huffman.

  19. Steve Fitzpatrick said

    Sorry again, that was supposed to be “hygroscopic” not “hydroscopic”.

  20. Kenneth Fritsch said

    Yes folk have already postulated this in the past, based on the already known fundamentals of particle science. The idea was to release humungous amounts of SO2 to form SO4 aerosol in the upper atmosphere and thus effect a cooling. Didn’t really get past 1st base as a realistic option.

    A chemical aerosol would a bit more tricky than getting some fine particles of clay up there.

  21. [...] This post was mentioned on Twitter by Science Blog News, C Jenkins. C Jenkins said: Projection , Regression http://goo.gl/fb/Nn4uC [...]

  22. Carrick said

    Right you are on 0.1µm…. quick math in the head doesn’t always pan out.

    I have spent much of the last 20 years thinking about the scattering of light by small particles; my business depends on getting it right. The NOAA site is targeted for general (non-technical) consumption.

    Given that there are scientists listed on the page, it was likely written by them. I’m skeptical that it is 100% at odds with what they believe to be true.

    BTW, 0.1 micron seems very small based on what I have seen with published “Angstrom ratios”…. closer to 1 micron seems more reasonable;

    I’ve found a reference since then.

    and would in fact be ~2X the average wavelength of visible light… and with extremely strong forward directional scattering.

    I’d like to see the math behind that…having a diameter larger than the wavelength of light creating dominantly forward scattering seems back-asswards to me. If I get a chance, I’ll look up that reference, but quickly..why doesn’t Rayleigh scattering theory apply here?

    Well, that depends on what they are and what their size and concentration is. Absorbing particles (black carbon/carbon black) most certainly do not reflect much light; they strongly absorb at all wavelenghts, independent of particle size. Hydroscopic sulfate aerosols are spherical droplets which absorb little, but which may (depending on size) scatter strongly anywhere from low to high scattering angles.

    Black lived soot is not very long lived. When it lands on snow it actually reduces the net albedo of course. (I made the same mistake of using “hydroscopic” rather than “hygroscopic”.) From what I understand, sulfates are a dominant player as a climate driver, so this is consistent with my understanding too.

  23. Terry said

    Steve # 17

    Sorry, I wasnt aware of your particle science knowledge. Re the literature on sizing, I find it quite sketchy. Also some data, and perhaps what you allude to w.r.t. 1µm is reported as mass fraction, not number density. I also note that some of the recent research does not go down into the sub-micron region either, which from my understanding, misses the bulk of the particles (number basis). However, if of course you are interested in directional scattering and reflection then that is another matter. There is some quite good literature going back to mid 1970’s or 80’s if I recall, that quantified upper air particle size distribution. I will have a look for it.

  24. Steve Fitzpatrick said

    Carrick #22,

    Very fine aerosols (well under 100 nm) are in the Rayleigh region, and do in fact scatter light in all directions, but with the minimum scattering intensity at 90 degrees to the incident direction. The highest scattering intensity for these very small particles is directly forward (near the original direction) and directly backward, and those intensities are almost equal. There is a smooth curve between the peaks of scattering intensity and the minimum at 90 degrees, that minimum is a little more than half the maximum. The thing to keep in mind though is that the scattering cross-section (essentially the ratio of effective scattering ‘area’ to physical cross-sectional area of the particles) falls rapidly as particle size falls; a reduction of a factor of two in size in the Rayleigh region reduces the total scattering per unit weight of particles by a factor of roughly 16. So while the number and even area distributions may be dominated by extrememly small particles, they most certainly do not dominate the total scattering. Page 9 of the PowerPoint presentation you found shows a graph of net scattering efficiency as a function of size for ‘sulfate’ aerosols. I can tell from the shape of the curve that this is the result of a Mie scattering calculation. What the plot does not show is the very strong directional component of the scattering…. which is dominated by forward scattering for particles over 100 nm. For analysis of how much light sulfate particles reflect back into space, this graph is not terribly informative. You would need to known the angular intensity shape of scattering for each size, an accurate size distribution, and then integrate over the day with changing solar angles to estimate how much light (on average) gets scattered back into space. Very messy. And having clouds also present in atmosphere make the calculation essentially impossible in the absence of an assumed distribution of clouds during a typical day.

    Easier I think to just measure the albedo from space.

    “I’d like to see the math behind that…having a diameter larger than the wavelength of light creating dominantly forward scattering seems back-asswards to me.”

    It is counter-intuitive before you see what the Mei scattering calculations yield. I can’t show you the calculations directly, but I can quote a few words from Bohren and Huffman, describing the scattering from a 0.26 micron water droplet suspended in air: “The scattered irradiance in the forward direction is more that 100 times greater than in the backward direction; such directional asymmetry only becomes more pronounced as the particle size increases, to the point that it is of little value to display scattering diagrams in a linear fashion.” In other words, for bigger droplets, you can’t even see the back scattered intensity in the linear diagram… you need a log diagram for the back-scattered light intensity to be visible in the diagram. They go on to say “Our intent in showing this one linear plot is to emphasize the predominance of forward scattering even for rather small spheres – a 0.26 micron water droplet is so small as to be unheard of in clouds.”

    Driving into the setting sun is difficult mainly because all the particles in the air are scattering light a low forward angles, so even if you do not have the sun directly in your eyes, there is terrible glare from particles suspended in the air… not to mention particles on your windshield. Turn 180 degrees, and the air looks perfectly clear… since there is almost no back scatter.

    Now there is an ever increasing average angle of scattering as the particle size goes down, so if atmospheric aerosols significantly reduce the average droplet size (or crystal size) in clouds, then sunlight entering the cloud will get “turned-around” by multiple scattering more quickly, and so will reflect back into space at a higher rate (the longer light spends in the cloud, the more of it is going to be absorbed). While I have never seen the spectrum of light reflected from clouds, my guess is that it is shifted somewhat to the blue due to selective absorption of longer visible wavelengths, and because the droplets in clouds are “smaller” from the point of view of longer wavelength light… and so scatter at smaller average angles. Infrared light from the sun is ~100% absorbed by clouds… the portion of the solar spectrum which can possibly be reflected back into space by clouds is limited.

  25. Steve Fitzpatrick said

    Terry #23,
    “I wasnt aware of your particle science knowledge.”

    I am the co-founder of a company that makes very high resolution instruments for measuring the size distribution of particles in the ~30 micron to ~10 nm size range. Accurate calculations of how light is scattered by small particles is part of the technology.

  26. Jeff Id said

    Steve,

    It would be interesting to see a writeup of the basics. I’ve got some experience with particulate absorption and reflection from a project in which we used these properties to measure the emissions of a future diesel M1 tank engine. If you have the interest, there are a bunch of people who would like to read it.

  27. Carrick said

    SteveF, thanks for the comments.

    In the mean time, I’ve had a chance to think a bit more .

    Rayleigh scattering theory applies when the wavelength is large compared to the size of the object. When the wavelength is comparable, you use Mei scattering theory, as you allude to above.

    There are various codes here.

    There’s even a calculator online here.

  28. Steve Fitzpatrick said

    Jeff #26,

    Thanks for the offer. I will think about how the basics might be summarized. The problem is two-fold: 1) the basic calculations (if you want to go to any level of detail) are not simple; the simplest solutions are sums of Bessel functions, and 2) the field is very broad, so a creating short summary that is meaningful presents a challenge. Maybe limiting it to a discussion of atmospheric aerosols would be best. Are you sure there would be interest?

  29. Steve Fitzpatrick said

    Carrick #27,

    I had never seen those on-line resources.

    Mie theory is the true solution… for all sizes. Rayleigh’s approximation to the Mie solution is adequate for very small particles. With the development of inexpensive (and fast) computing power, there is little reason to not just use Mie calculations in almost every case, even for very small particles.

  30. Jeff Id said

    Steve,

    I know I would enjoy a technical post on aerosol scattering. Many here enjoy the math and have diverse backgrounds so I think you would have some interest, especially if the context is tied to atmospheric aerosols.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
Follow

Get every new post delivered to your Inbox.

Join 140 other followers

%d bloggers like this: