Current atmospheric models underestimate the dirtiness of Arctic air

From the “we told you so” department, it seems they are just now catching up with observations.

Soot at the bottom of a meltpool in Greenland

Soot at the bottom of a meltpool in Greenland

Black carbon aerosols–particles of carbon that rise into the atmosphere when biomass, agricultural waste, and fossil fuels are burned in an incomplete way–are important for understanding climate change, as they absorb sunlight, leading to higher atmospheric temperatures, and can also coat Arctic snow with a darker layer, reducing its reflectivity and leading to increased melting. Unfortunately, current simulation models, which combine global climate models with aerosol transport models, consistently underestimate the amount of these aerosols in the Arctic compared to actual measurements during the spring and winter seasons, making it difficult to accurately assess the impact of these substances on the climate.

To find out if these inaccuracies could be mitigated, a team of scientists decided to use the Japanese K computer to perform fine-grained simulations of how black carbon aerosols are transported to and distributed in the Arctic region. By using smaller grids–with spacing of just a few kilometers rather than several tens of kilometers as in conventional current models–they were able to show that they could more realistically model the amount of black carbon aerosols, mitigating the underestimation in more coarse-grained models. Their finest model used 3.5 kilometer grids broken up vertically into 38 layers, so that it required 1.6 billion grids to cover the globe. The simulation, done on the 10-petaflop K computer, still required 17 hours to perform the two week simulation.

According to Yousuke Sato of the RIKEN Advanced Institute for Computational Science (AICS), “this research shows that powerful supercomputers, by performing more fine-grained simulations, can help us to model weather and climate patterns in a more realistic way. We have to note, however, that while our model reduced the underestimation, it did not completely eliminate it. Further generations of even more powerful computers will allow us to run simulations that may be able to make even more realistic simulations and help us to understand the mechanism through which these aerosols are transported.”

“It is also known,” continues Sato, “that current models do not realistically model the vertical distribution of the aerosols, and we believe that finer measurements could help there as well. Unfortunately there were no vertical measurements taken in November 2011, the time we chose to model, so we plan in the future to do simulations for time periods for which actual measurement data exist.”

###

The research, published in Scientific Reports, was carried out by AICS in collaboration with the University of Tokyo, the National Institute of Environmental Studies, Kyushu University, and the Japan Aerospace Exploration Agency.

Advertisements

68 thoughts on “Current atmospheric models underestimate the dirtiness of Arctic air

    • So just what is the scientific purpose to underestimating the dirtiness of Arctic air ?

      I can understand why one might measure the dirtiness of Arctic air, but why the hell, would you want to underestimate it with some X-box model ??

      Just asking !

      G

      • George – We’re hoping they will sell the underestimating model to the IRS so they can use it to underestimate the taxes we owe. Write your congress-critter and suggest that it’s a sure-fire good model to buy for the IRS. If we tell ’em they will have to buy the model before they can know what’s in the model, both Houses will authorize a purchase before the end of the week ;o)

      • I wonder, did they model November 2011 not knowing that the vertical measurements weren’t available or did they choose to model that timeframe because they didn’t exist. Seems like a massive waste of time to run models for a long period just to say “whoops we should have run it with vertical data too”. Must be seeking more grant money to rerun vertical information

      • “So just what is the scientific purpose to underestimating the dirtiness of Arctic air ?”
        An efficient policy to enable eugenicists to blame industrial pollution resulting from a population whose size is beyond the earth’s “carrying capacity” for the coal fly ash they’re dumping in the earth’s atmosphere.

    • Phillip Bratby May 25, 2016 at 9:40 am
      More settled science!

      Who said the science regarding black carbon particles was settled?

      It certainly wasn’t these scientists who are studying the effects of black carbon soot on the albedo of ice and how it is increasing ice melt.

      http://darksnowproject.org

      • It was the climate “scientists” who have been telling us for years that we already know everything we need to know about how the climate works.

      • The black carbon particles do NOT just fall down on the ice, like they just fell off the back of a turnip truck.

        The carbon particles are very neatly tucked inside each and every (nearly) water droplet, or snow crystal so they come down WITH the snow, not ON the snow.

        G

    • “More settled science!” indeed, here is some more of it:
      Global warming ain’t going to be as catastrophic as the climatism practitioners thought

      New cloud formation discovery may lessen warming forecast
      “A new discovery about how clouds form may scale back some of the more dire predictions about temperature increases caused by man-made global warming.
      ……….The new work shows that a combination of cosmic rays from space and gases emitted by trees also creates particles, and then clouds, without man-made pollution.”
      According to researcher Jasper Kirkby from CERN.
      http://phys.org/news/2016-05-cloud-formation-discovery-lessen.html

  1. “Their finest model used 3.5 kilometer grids broken up vertically into 38 layers, so that it required 1.6 billion grids to cover the globe”

    Exactly how do you go about ground truthing a model like that? Unless you can do so the model just becomes an exercise in parameter fitting.

    • So just how did the get all of those sensors into place to measure that many grid points and in so many altitudinous layers ??

      Seems like a pretty big project to me. Who is paying for all of those dirt sensors ??

      Well I know; it’s me that is; right ??

      g

    • Unfortunately there were no vertical measurements taken in November 2011, the time we chose to model, so we plan in the future to do simulations for time periods for which actual measurement data exist.”

      No , shit.

      • Why would you do that ? People would just compare the two and complain about the lack of fit.

        g

      • Why would you do that ? People would just compare the two and complain about the lack of fit.

        ISWYDT

      • I cannot decide which is more breathtaking.
        A. To perform a 17 hr simulation on a world-class supercomputer on a scenario where you have insufficient data to validate the results as opposed to other possible scenarios with more data.
        B. Admit you did A.

        Optimistically, perhaps peer review worked in this case and the reviewers insisted on B.

    • tomcourt. Good point. A simulation model such as this is essentially un-testable, thus it is unverifiable and un-validatable. Hence, it has no empirical import and is thus of no scientific value.

  2. Ermmm…. How about sampling the air, water, and ice for TSS, and a few other residuals? Models? Maybe some real data to plug in such as “volcanic ash with a Kamchatkan signature is here, lignite combustion ash was here, biomass here, the ratio’s were this here, etc and so forth.

    Good grief, I would have been crucified trying to defend a paper like this in some undergraduate classes.

  3. Presumably the soot particles fall on the snow, where melt flows bring them together in pools. Then those pools preferentially melt, so they fall right through the ice, and deposit the soot particles in the sediment on the sea floor.

    A self-cleaning process. The sort of thing which is totally ignored by the alarmists..

    • NO ! The soot particles are IN the snow.

      Each water droplet / snow crystal condensed ON the soot particle and then grew until it was too heavy to sustain its altitude, so it fell to earth as snow / sleet / rain / hail / whatever.

      G

  4. How many decades have Europeans favored diesel powered cars to deal with extreme gasoline tax rates? The same question applies to their auto exports and foreign car plants.

    • For years the locals in and around Chamonix and up through Zermatt have believed that glaciers were impacted much more by diesel soot than by temperature. As diesel fuels get cleaner many of the European glaciers have been on the rebound. I have cousins in Zurich who talk the talk of AGW simply because they work in the world’s most active market for trading carbon credits. Off the record they just roll their eyes.

    • Diesel motors are inherently more efficient since they run at a higher compression ratio. This was the traditional reason to favour them fiscally.

      In any case we should be using all fractions of the crude oil to exploit resources in the most useful way.

      Reducing over-congestion in major cities is adantageous to reduce health risks. It would make more sense to favour public transport with subsidies but then some free-marketeers like Thatcher smash schemes like “fairs fair” in London which provided a day pass unlimited travel for £1.

      Now instead we have a ‘congestion charge’ which means that the well-off get to use their cars ( individual freedom ) and working folks are obliged to use the now much dearer public transport.

      • Greg,
        Diesel engines are inherently more efficient since they are not throttled like gasoline engines. Compression ration has some impact, but pumping losses are more important. Engines are essentially air pumps, and a throttle body or carburetor causes the pistons to suck air in under vacuum except for wide open throttle. This under light load, gasoline engines are inherently less efficient than diesel engines. Add to that the fact that gasoline has about 115,000 btu/gal LHV compared to about 132,000 btu/gal for conventional diesel and you have a very big improvement in volumetric fuel economy. Due to the pumping losses, the Brake Specific Fuel Economy (lbs fuel/HP-h) is also much lower.
        Some of the efficient gains for diesel engines are lost when trying to meet strict emissions standards though. The diesel injection timing must be retarded to reduce NOx, but that increases PM emissions Thankfully catalysts and traps can reduce diesel emissions to those of gasoline engines but with FE losses. Overall, diesel fuel is still less expensive to produce than gasoline due to the fewer restrictions on diesel fuel and the fact that you don’t have to add ethanol to diesel fuel. Market forces in the past have pushed diesel fuel value above gasoline, but they are on parity now and in some places, diesel is less expensive than gasoline.

        Overall, diesel engines should provide 25-40% improvement in FE over a gasoline engine and are better suited for hybrid use. But they are more expensive to produce and that cost must be recovered in FE benfits or in towing capability (PU trucks, HD vehicles, etc).

      • If the Free Market were allowed to operate FREELY many businesses would leave London, New York and other congested cities and actually solve transportation issues. Permitting large cities to operate on National subsidies, mandating Unionization and locating Central Government in over-priced and over-crowded locations raises costs. Easy money since the non-urban peasants pay for stuff.
        Stop subsidizing fares and permit real competition. Uber will break the taxi-government monopoly and reduce fares a bit. Destroy the power of government to license monopolies and watch costs drop and service improve.

      • MarkW,

        Did you miss the bit where he mentioned throttle bodies. All petrol engines fuel injected or not, use a means of restricting airflow to throttle the engine. Carburation just adds the fuel in at that point. His point is still valid when comparing diesel to petrol engines.

        Regards

        Gnrnr

    • I was implicating soot pollution, but I guess I needed to spell that out more clearly. The same applies to the belated switch to unleaded gasoline in Europe.

    • The reason for diesel engines is that the thermodynamic cycle of the Diesel engine (Diesel Cycle) has a higher thermal efficiency that the Otto Cycle of the standard sparked gasoline engine.

      Neither one is as efficient as the theoretical Carnot cycle.

      In principle, the Stirling cycle engine, can achieve Carnot cycle efficiency, but Stirling engines apparently are hard to implement.

      A diesel cycle engine doesn’t actually need to be a very high compression ratio engine; that is a function of the fuel you plan to burn and its self ignition conditions. But of course, the use of a high compression ratio design does give a more powerful and efficient engine, because of higher Temperatures and pressures; which also gives higher bearing loads.

      But the thermodynamic efficiency of the diesel cycle is closer to the Carnot efficiency than an Otto Cycle can give.

      G

      • As near as I can recall, in the Carnot cycle, Heat is added to the working fluid (air ?) at some constant high temperature Th. Then the working fluid is aloud to expand adiabatically to some lower pressure, higher volume and lower temperature Tl.
        The waste heat is extracted from the expanded working fluid at a constant Tl, and then the working fluid is recompressed Adiabatically up to the higher pressure smaller volume condition at Th.

        Or something like that. So two isothermal strokes interleaved with two adiabatic strokes.

        Off hand I can’t reproduce the proof that that is the thermodynamic most efficient a heat engine can be.

        Steam engines follow a “Rankine Cycle”. Steam trains are Rankine cycle and pretty damn inefficient as I recall.

        I dunno what you call the thermodynamic cycle of a triple expansion marine steam engine. Very cool way to go.

        G

  5. Well that’s new. Researchers willing to test their model against observations AND admit there’s still a way to go. Will wonders never cease.

    • Well, in this case, they modelled it against observations we don’t have…I suppose its a start, but “still a way to go” seems to be an understatement.

  6. “It is also known,” continues Sato, “that current models do not realistically model the vertical distribution of the aerosols, and we believe that finer measurements could help there as well. Unfortunately there were no vertical measurements taken in November 2011, the time we chose to model, so we plan in the future to do simulations for time periods for which actual measurement data exist.”

    Now that is refreshingly honest.
    But .. disturbing .
    Simulation first.
    Data ?
    I used to think science required you to start from the data, then proceed to the modelling.
    So is this just an accident that no data ,actual measurements, exists?

    • I infer that they modeled in 3D but only have 2D measurement data for comparison. In the future they will compare their 3D model to 3D measurement data.

    • You do need to start with the data, which is produced by the system being examined. Then you need to develop a hypothesis about how the system produced the data. Then you need to develop an algorithm to reflect the hypothesis. Then you need to implement the algorithm in a computer program. The data is generally divided into input data and output data. The input and operation of the algorithm produce the output. Then you compare the actual system output to the simulation output. If they do not reasonably match, something is wrong. It appears to me that in climate science we do not have data measurement right yet. Everything else is premature and a waste of time and money. We first need to work on methods to measure climate data that everyone can agree is reasonably accurate and represents real system output.

  7. There has been early melting in the Beaufort strait this year and we have had volcanic activity in Alaska, is there a connection? Anyone have experience in this type of interaction.

    • Doubt it.

      The volcanic activity has not be big enough to be reported on, so is probably poultry eruptions that would not have a measurable impact. To understand Arctic melting needs much larger drivers. Atmospheric and oceanic circulation IMO.

    • Bob Boder at 9:47 referred to early melting in Beaufort Strait (Sea?)
      I have not seen that there is early melting in the Beaufort Sea this year. What has occurred in this Sea is that the Beaufort Gyre has broken up the ice and moved it clockwise around the North Pole. For the most part I don’t think the ice has melted. I have watched it over the last month and it is quite spectacular. A time lapse video of it is available here http://www.ec.gc.ca/glaces-ice/default.asp?Lang=En&n=602ED529-1.

  8. The standard line is that the soot is from Chinese coal fired power plants, and preferentially the older, dirtier ones. Maybe the standard line is true enough, maybe the story is more complex.

    Wouldn’t you love to have a few grams of material from the bottom of that pool? Do some trace element analysis, some isotope ratios, and a story might emerge. Who knows, maybe some accurate source maps could even help the modelers. (I need a grant)

    • Even eco-activist Dr Jason Box admits that most of the dark material on the Greenland on the Greenland Ice Sheet is from cyanobacteria and blue-green algae.

      These Japanese folks have fallen at the first hurdle: the darkening is probably not all from aerosols so there may be no underestimation at all. Sunlight + water = life (cyanobacteria + blue-green algae)

      The blog on the http://www.promice.org website has recently disappeared but the inconvenient information is still available on archive.org. Click on the Archive for August 2013 on the right hand side of link below, then look for the caption of the second photo down on the resulting page:

      https://web.archive.org/web/20160323212944/http://promice.org/Blogs.html

      The caption reads:

      boots on the ice offer a close look (and to sample) impurities concentrating at the surface. The fact is, much of this dark material is from cyanobacteria and blue-green algae. Photo J. Box.

  9. From the article: “so we plan in the future to do simulations for time periods for which actual measurement data exist.”

    Excellent idea. Using actual measurements is good. Real world! Old School!

  10. It’s just the latest from their giant Red Herring department. They don’t have a clue what drives climate (though they think they do), but – “hey look, – Squirrel!”

  11. Up above, Dodgy Geezer motes that the melt pools are how the ice clears itself. It is also clear that a lot of soot gets buried. Looking at the back wall of the pool, at least four or five distinct soot bands are clearly visible. (Open the picture in a new window to embiggen.)
    Just wait, in a few hundred years, people will be drilling cores in that ice and using those soot bands as a proxy for something.

  12. If they’ve underestimated the dirtiness of the Arctic atmosphere then that also means they’ve over estimated the effect of raised temperatures on melting ice.

  13. Trivia Question Time: How many not-clean diesel vehicles were added to the Chinese market over the past 30 years?

    • Diesels all over Asia are filthy, huge black clouds of soot every time the engine is revved. Diesels on ocean freighters are also extremely foul. Diesel Particulate Filters are found on most Diesel cars now. Wildfires and coal-seam fires also put huge amounts of soot into the atmosphere. In China they burn brown coal, and many power plants have no baghouses to capture the soot. Lots of Arctic melting is caused by this, nothing to do with CO2…

  14. I have heard that a single tree that burns in a forest fire puts more carbon into the atmosphere that an auto does in a full year. The Fort McMurray fire along with others burning in northern Canada are dumping so much into the atmosphere that they will mask the anthropogenic contribution this year. Or maybe the climatists will claim it all for man ;>)

  15. so we plan in the future to do simulations for time periods for which actual measurement data exist

    • Exactly my thought. If you know your model is consistently underestimating the real data, why persist with the model?

      • By understanding why your model consistently underestimates the real data, you may be able to learn something about how the real world works.

  16. I will go OT for a moment to share this developing story under the tag ‘underestimate’.

    French Nuclear workers to go on strike at 19 nuclear plants in France tomorrow.

    http://news.trust.org/item/20160525173613-0qrhj

    They may have to import power from other regions. (all those wind turbines).

    Here are the wind forecasts:

    Germany

    http://www.windfinder.com/weather-maps/forecast/germany#6/51.330/10.453

    France:

    http://www.windfinder.com/weather-maps/forecast/france#5/46.449/2.210

    France also is the largest net electricity exporter in the world…

    http://www.world-nuclear.org/information-library/country-profiles/countries-a-f/france.aspx

  17. Make no mistake: the agenda is to blame fossil fuels for the soot and dirt, even though they would be only one source, and more likely to occur in developing countries with fewer controls on pollution. The argument that we should cut down on fossil fuels because of black carbon is just as foolish as saying we should do so because of “carbon”.

    • They still have a long way to go.
      Clouds/rain have a lot to do with how fast particles get washed out of the atmosphere. The resolution will have to get down to well under 1 km before they have even a chance of getting clouds right (Not a big chance, but better than they have now.)

  18. ‘We have to note, however, that while our model reduced the underestimation, it did not completely eliminate it. Further generations of even more powerful computers will allow us to run simulations that may be able to make even more realistic simulations.’

    I speak “climate science.” Let me translate:

    “All previous models were junk. Our new, improved one isn’t so good. It will be better in the future.”

  19. Geez, models!!! We’ve been scooping air for sampling for half a century at least and measuring particulate matter. Tow a filter frame in a grid over Greenland and calculate the real numbers!!! What do the Geophysical Year boys and girls do in the Arctic? They run around on icebreakers for two years to see what is happening to the ice!!! I think experimental science is close to dead. Apparently all the experiments done on “acidification” of sea water and its effects are bordering on useless because of a lack of skill and intuition on designing experiments.

  20. MarkW,
    Carburetor or Fuel Injection, both require throttling air to get proper air/fuel ratio and control power output. Diesel engines control fuel flow to control power. GDI is an attempt to reduce throttling losses with some success but with more NOx emissions and increased catalyst requirements to reduce emissions.

  21. “Black carbon aerosols–particles of carbon that rise into the atmosphere when biomass, agricultural waste, and fossil fuels…..”
    Left out wild fires.

  22. When I consider the atomic level, I ask how do atoms emit a photon and cool. In the macro world a coarse surface far greater emissivity than a smooth surface. Air does not even have a surface. So spinning atoms with the force of their molecular momentum some how come together and interacting electric and magnetic fields produce a photon. But what effect does a huge by comparison aerosol have in creating a Nucleation site for the formation of photons. Or a Nucleation site for forming an ice crystal.

    So what implact on warming do these Nucleation sites have in an artic of 24×7 hours of winter night time.
    And in the world of emissivity 1 is given to a perfect black body and ice is at .97.

    I wonder if 90+ percent of all atmospheric black body radiation is emitted by aerosols

    A different process of transformation showing what a Nucleation site mean in a state transformaton.

  23. My grandmother (Finland) taught around 1965 (I was 6-7y old) that use ash on snow and the Sun will vapor snow fast. The ash is lighter than snow and it stays on surface of the melted water drops on top of surface of snow. Even the ash is more gray than black, this is an efficient way to clean snow in Spring when Sun is shining.

Comments are closed.