New PNAS paper claims Arctic planetary albedo dropped significantly, yet recent CERES data shows no significant change

From PNAS:

Direct satellite observation reveals that the Arctic planetary albedo, a measure of reflectiveness, decreased from 0.52 to 0.48 between 1979 and 2011, a change in albedo that corresponds to a climate forcing 25% as large as that due to changes in atmospheric carbon dioxide concentrations over the same time period, according to this study.

But looking at more recent CERES measured data showing what is light is reflected into space by the posited changing albedo, neither Arctic nor Antarctic seem to have changed much at all; less than a watt per decade everywhere. For recent CERES data the trend lines look dead flat at any reasonable scale. See the graph:

arctic-antarctic-reflected-solar

The graph is from Willis Eschenbach, who will have more on this in subsequent post.

I’ve said in the past that black carbon is likely playing a role in Albedo change in the Arctic, but I’ll have to admit it is puzzling why the signal isn’t stronger in Willis’ graph above.

Compare the period above from 2000-2013 to the graph they cite in the paper:

Arctic-albedo-pistone-PNAS

Fig. 4. (A) Observed annual-mean clear-sky and all-sky planetary albedo for
the entire Arctic region. Solid lines are direct CERES observations, and
dashed lines are estimates derived from sea ice observations. The error bars
in the Bottom Left corner indicate the uncertainty in the pre-CERES clear-sky
and all-sky albedo values (Supporting Information). (B) All-sky albedo as in A
compared with annual-mean observed sea ice area (as a fraction of the
ocean in the Arctic region) and surface air temperature averaged over the
ocean in the Arctic region.

Note they claim a drop of clear sky albedo from about 36 to 33%, that may be true for their method, but the base CERES all sky reflected data suggests no trend at all. Tis a conundrum that may be due to the sea-ice method they chose:

The sea ice surface albedo is estimated using two steps. First, the clear-sky planetary albedo associated with 100% sea ice cover is computed from an ordinary least squares linear regression between albedo and sea ice cover for each month constrained to go through an ocean albedo of 0.175 (cf . ref. 25) at 0% sea ice cover. For this calculation, a region containing all ocean grid
cells between 80 and 90°N is used to reduce the extrapolation to 100% ice calculation and to focus on multiyear ice for comparison with the in situ observations. Next, surface albedo (ásfc) is calculated from this clear-sky planetary albedo (ács) based on a linear estimate, ásfc = (ács . a)/b, with empirically derived seasonally varying monthly parameter values of a and b adopted from a previous study (17).

The clear-sky albedo during 1979–1999 is computed from sea ice using a total least squares linear regression between 2000 and 2011 clear-sky albedos and sea ice. All-sky albedos during 1979–1999 are similarly computed from clear-sky albedos using a total least squares linear regression between the two albedos during 2000–2011. In both regressions, the quantities are
normalized by their uncertainties, and error bars are estimated based on the regression confidence intervals (details provided in Supporting Information).

The study:

Article #13-18201: “Observational determination of albedo decrease caused by vanishing Arctic sea ice,” by Kristina Pistone, Ian Eisenman, and Veerabhadran Ramanathan.

Abstract:

The decline of Arctic sea ice has been documented in over 30 y of satellite passive microwave observations. The resulting darkening of the Arctic and its amplification of global warming was hypothesized almost 50 y ago but has yet to be verified with direct observations. This study uses satellite radiation budget measurements along with satellite microwave sea ice data to document the Arctic-wide decrease in planetary albedo and its amplifying effect on the warming. The analysis reveals a striking relationship between planetary albedo and sea ice cover, quantities inferred from two independent satellite instruments. We find that the Arctic planetary albedo has decreased from 0.52 to 0.48 between 1979 and 2011, corresponding to an additional 6.4 ± 0.9 W/m2 of solar energy input into the Arctic Ocean region since 1979. Averaged over the globe, this albedo decrease corresponds to a forcing that is 25% as large as that due to the change in CO2 during this period, considerably larger than expectations from models and other less direct recent estimates. Changes in cloudiness appear to play a negligible role in observed Arctic darkening, thus reducing the possibility of Arctic cloud albedo feedbacks mitigating future Arctic warming.

The paper:

pnas.201318201 (PDF)

About these ads

37 thoughts on “New PNAS paper claims Arctic planetary albedo dropped significantly, yet recent CERES data shows no significant change

  1. We find that the Arctic planetary albedo has decreased from 0.52 to 0.48 between 1979 and 2011, corresponding to an additional 6.4 ± 0.9 W/m2 of solar energy input into the Arctic Ocean region since 1979.

    I’m curious how this was calculated. Was the low angle of incoming solar energy, that most of the day the Sun is not overhead any specific longitudal line, and that during the rest of the day when it was clear out open water would be exposed to a very cold Tsky all accounted for?

  2. What actually matters is global cloudiness and NOT Ant/Arctic ice or Ant/Arctic cloudiness.

    See the Earthshine project. Global cloudiness decreased during the late 20th century warming spell and has been increasing since around 2000 when the jets and climate zones stopped shifting towards the poles and started moving back again.

    Consider the more meridional / requatorward jets of recent years and the resultant longer lines of air mass mixing with resultant increased global cloudiness.

    Apart from that, more open (albeit darker) water in the Arctic Ocean actually means faster cooling from that open water which loses heat to the air faster than does water covered by a layer of ice.

    The Arctic Ocean is warmed not by the weak insolation from high latitude sunlight but by warm water flowing from the equatorial regions via the Gulf Stream past Svalbard.

    The more of that water is exposed to the air the faster the energy loss for the planet as a whole because it is always at a higher temperature than could have been achieved by insolation alone.

  3. Go to the PNAS site and look at the articles and authors. http:www.pnas.org Look at “PNAS in the News”. Some very good articles I am sure but so many alarmist style papers on so many different subjects. Is this a modern trend in science to get noticed? Say PNAS five times very fast.

  4. Is the albedo decrease the cause of decreased sea ice or is the decreasing sea ice the cause of the albedo decrease? I don’t agree with their assumption.

  5. What mechanism causes the jet stream to veer south? Does the warmist explanation make sense?

    http://news.yahoo.com/jet-stream-shift-could-prompt-harsher-winters-scientists-132931128.html?bcmt=1392583343783-2fca2058-3f01-42f1-834e-c6c4fff0a483&bcmt_s=e#ugccmt-container

    [Chicago (AFP) – A warmer Arctic could permanently affect the pattern of the high-altitude polar jet stream, resulting in longer and colder winters over North America and northern Europe, US scientists say.]

  6. Presumably the Ceres measures light leaving the Arctic vertically from the surface. Since much of the light from the higher latitudes is reflected because of the low angle of incidence, this would give a collar effect because the reflected light would be picked up as having come from lower latitudes. Am I completely wrong about this? Surely, even if the Arctic was all water, 90% of the incident light above 80N would be reflected anyway, perhaps less in summer and more in winter. Indeed, during the Arctic winter this area gets no incident light. Light with an angle of incidence of 80degrees, 45% of it is reflected. Ice with a refractive index slightly lower than water actually (and counterintuitively) has a lower percent of light reflected than water at the same angle of incidence – WUWT?.

    http://www.physicsclassroom.com/getattachment/reasoning/refraction/src41.pdf

  7. This is quite simple. They are calculating the reflected in solar radiation using reduced albedo in a mathematical model with the change in observed ice extent. CERES is taking a direct reading of reflected solar radiation. The difference between the two is like finding out what your image in a mirror should look like based on calculations from the amount of reflecting material on the back instead of simply looking in the mirror.

  8. Methods
    In this work the Arctic is defined as the area north of 60°N

    Now, 60°N runs south of Iceland, across the entrance of the Gulf of Bothnia and at the southern tip of Greenland. Neither one can be seen in Fig. 1. of the paper, those maps seem to be cropped at the Arctic Circle (66.5622°N).

    This kind of sloppiness is impermissible in science, those who can do that are susceptible to far worse crimes against reason, not to mention reviewers generously skipping this blunder.

  9. Mike Tremblay says:
    February 17, 2014 at 1:35 pm

    I’m not sure your assessment is correct. The study seems to have calculated the linear relationship between the available CERES data (2000-2011) and ice cover. They have then used that relationship to estimate the change in albedo between 1979 and 1999 (i.e. the pre-CERES period). It doesn’t seem an unreasonable approach. The paper states

    Thus, we estimate the magnitude of the longterm increase in solar energy into the Arctic using the observed pre-CERES (1979–1999) ice cover paired with a linear approximation of the relationship between albedo and ice constructed from observations during the CERES period (2000–2011) for each month and each region, as illustrated in Fig. S6

  10. The sea ice surface albedo is estimated using two steps. First, the clear-sky planetary albedo associated with 100% sea ice cover is computed from an ordinary least squares linear regression between albedo and sea ice cover for each month constrained to go through an ocean albedo of 0.175 (cf . ref. 25) at 0% sea ice cover. For this calculation, a region containing all ocean grid
    cells between 80 and 90°N is used to reduce the extrapolation to 100% ice calculation and to focus on multiyear ice for comparison with the in situ observations.

    Nope. I have to disagree. Those values are NOT what the measured albedos actually are.

    Dr Curry’s measured actual, real-world Arctic sea ice albedo begins at 0.822 in January, stay near-constant until early May, decreases until day-of-year 206 ay 0.46 (best fit, lowest sea ice albedo = 0.38) then rises until day-of-year 275 back to 0.823. It stays at 0.82 until the end-of-year on day = 365.

    Actual real-world measured ocean albedo is available from several different open ocean (real waves and winds!), and varies based NOT on any assumed values of anything, but can be specified (under these conditions of the study = clear day, arctic conditions, direct radiation values). The open ocean albedo varies specifically according to solar elevation angle (SEA) and it varies EVERY HOUR OF THE DAY according to very specific deeper calculations of declination number (DAY variation), hour-of-day variation (HRA), and latitude.

    Further, the actual solar radiation at top-of-atmosphere varies significantly but predictably as a direct function of day-of-year and yearly average TSI value.

    On any given day of year, at any chosen latitude, I have a specific spreadsheet available to any who wants it calculating the actual direct and diffuse radiation at TOA, expected radiation on a perpendicular surface and on a flat (horizontal) surface for all 24 hours of the day, and the expected absorbed and reflected direct and diffuse radaition based on that day’s sea ice albedo and that hour’s solar elevation angle. All are invited to offer corrections or extensions, revisions or ridicule – as long as each has a verifiable measured data contradicting any function. (For example, there are several different ways to calculate air mass, and several other ways to assign a clear-sky attenuation coefficient.

    Hint: A “monthly” value is meaningless unless it is based on a hour-to-hour changing value – at the actual altitude where the ocean and sea ice actually are.

    Hint: Actual measured ocean albedo against direct radiation on clear days vary hour by hour from 0.40 (or higher!) at SEA under 6 degrees down to 0.035 (at SEA’s over 40 degrees). Open ocean direct radiation albedo does vary with wind speed, lowering as wind speed increases and lead size increases – neither effect is mentioned in their “average” 0.17 value.

    Diffuse radiation does NOT vary with solar elevation angles, and varies only slightly with moderate wind speeds, but stays right at the “common” average of 0.065

    The proportion of diffuse radiation received to theoretical-clear-sky direct radiation DOES vary significantly with solar elevation angle, but understand that there is ALWAYS “some” diffuse radiation present even during clear skies, but there may be no direct radiation present at all under cloudy skies. In mid-summer in the Arctic, there may be only one day in seven that actually sees “clear skies” direct radiation! The rest of those six days during the summer, 45-70 percent of the “potential” radiation is reflected away by the tops of the clouds and attenuated by the atmosphere, and only that minute 30% of what arrives at the TOA is left over to actually hit the surface.

    Hint: The location of all Antarctic sea ice varies between 59.2 south latitude in September last year to 70 south latitude in late February. This creates a very, very significant different effect from the Arctic sea ice which is retreating from 70 north in February to 80 north in September. Sea ice albedo (Antarctic sea ice is almost entirely first year clean ice, arctic sea ice contains many million km’s of secondeyar, third year, and even fourth year ice. All other variables change as well: solar exposure times, and top-of-atmosphere radiation values, and ocean albedo are changing as well.

    Now, did the writers actually make every one of these very important changes to every variable? They did not say so. You hope they would. But they didn’t say they made any of them.

  11. If we were to seriously look at all the stuff out there, the climate isn’t going to change that much in the coming century in terms of the over all Earth. We can expect Glaciers to begin melting around 2035.

    Hurricane activity will pick up around 2040. About the end of the century water tables will begin to rise as we enter the 400 th year of post Mini-Ice Age. Study the Vikings.

    Too much depends on the Antarctica in terms of what goes on in the Arctic. We need to look south first. That is the Cause and Affect in the Atlantic Conveyor Belt.

    The hard part will being a human in a moderately populated world. The Earth will need a mini Ice Age for two centuries just to get the Population back to a point that the people left have a food source on which to survive.

    Got to be careful what we wish for.

    Paul

  12. John Finn says:
    February 17, 2014 at 2:19 pm
    ————————————————————————————————————————–
    You make a valid statement, but that’s not the point. They are calculating a value rather than looking at the direct readings – it’s like believing a weather report when all you have to do is look out the window.

  13. “They have then used that relationship to estimate the change in albedo between 1979 and 1999 (i.e. the pre-CERES period). It doesn’t seem an unreasonable approach. The paper states”

    yes, and I would suggest that one would want to cross check with the newly released CDRs
    for Albedo. There are two series going back to the early 80’s, monthly data, independent of CERES.

    They appear to have looked at some of the source data for one of these CDRs

  14. Mike Tremblay says:
    February 17, 2014 at 2:31 pm

    It is worse than that: They are describing the effect of a storm and predicting future flooding in Michigan (latitude 44) based on an “average” of a series of monthly weather reports for Key West Florida (latitude 24)!

  15. Mike

    They are calculating a value rather than looking at the direct readings

    They don’t have the direct readings for the 1979-99 period, That’s why they are using a linear approximation to estimate them.

  16. I built an Albedo calculation model.

    The average annual Albedo from 60N to 90N area-weighted is 0.48.

    80-90N is 0.67 and covers 0.76% of the Earth’s surface area
    70-80N is 0.57 and covers 2.256% of the surface
    60-70N is 0.46 and covers 3.383% of the surface

    So Arctic Albedo did not fall from 0.52 to 0.48 – it always was 0.48.

    If you plug all the numbers for all latitudes into the calculation model, one gets a global Albedo of 0.2983 which is the accepted number (from Trenberth etc.)

    The source for the Albedo values is “Earth’s Thermal Environment” which is the go to source to control satellites (and one needs to have very accurate measurements of radiation to keep satellites in the proper orbit).

    http://www.tak2000.com/data/planets/earth.htm

  17. @Steven Mosher-Okay what the heck are these CDRs of which you speak?

    You know, it sort of seems reasonable to use clear sky upward shortwave fluxes-which much come from surface reflection-regressed on the local ice coverage to estimate the effect of sea ice on this going back further. sort of. But it’s kind of complete BS to not properly quantify how large this actually is as a global forcing. The answer is that because it’s over a tiny portion of the globe that receives almost no sunlight to begin with, it is negligible. It might be relatively important locally but in global terms it doesn’t amount to much of anything.

    They also should have estimated the effect for Antarctic sea ice, too.

    Anyway, their claim is ~0.19775 W/m^2 (that’s 25% of the CO2 increase) over 1979-2011, something like ~0.006 W/m^2 per year. A forcing growing at that rate would take a little under 600 years to be the equivalent of a doubling of CO2. And you know something, that linear trend can’t continue indefinitely, it’s bounded by how much ice there actually is.

    Seems to be much ado about nothing.

  18. PNAS is little more than a gray journals for buddies to get their other buddies to sign off on their “research”.

  19. Andy: Willis’s graph isn’t a very useful way of addressing the information in this paper. The paper deals with albedo during March-Sept; Willis’s graph and trend covers the whole year. (I’m not sure what the meaning of albedo is when the sun isn’t above the horizon.) The paper finds the biggest decrease in albedo in September, when albedo on Willis’s graph is dropping. One can’t see by eye how much albedo has changed in September on such a graph. The only trend that is visible by eye is in peak albedo, which is clearly decreasing. It will be interesting to hear what Willis has to say. (I was unable to see the original article or the supplemental material at the PNAS website.)

  20. We find that the Arctic planetary albedo has decreased from 0.52 to 0.48 between 1979 and 2011, corresponding to an additional 6.4 ± 0.9 W/m2 of solar energy input into the Arctic Ocean region since 1979. Averaged over the globe…

    Are they serious? Every datum says Arctic trends are not global.

  21. Paul Pierett says:
    February 17, 2014 at 2:25 pm

    “The hard part will being a human in a moderately populated world. The Earth will need a mini Ice Age for two centuries just to get the Population back to a point that the people left have a food source on which to survive.”
    ——————————————–
    What’s up with that? What’s that supposed to mean? I think you are sitting on the wrong side of the Malthusian fence, Paul. Not many people here fall for the mainstream Lefty nonsense that food production is inevitably doomed to fall with “climate change” (in whichever temperature direction) and population bound to rise towards 10 billion.

    In fact with proper application of best agricultural practices, which are far from being in use today, we could easily feed 10 billion though any fool know world pop will never pass 9 billion [as per UN lowest estimate] and that capacity is irrelevant anyway as it is the economic policies of the ruling elite which determine the price of wheat and pork bellies, the “who gets what”.

    Apologies for being off-topic but this kind of defeatist misinformation needs combating at every instance.

  22. And what do they suppose the forcing from CO2 was over the period? I’m guessing the 40% of the greenhouse effect attributed to it in the models. Loaded for bear, this gas comprising less than 2% of the greenhouse gasses by volume.

  23. @gymnosperm-The forcing from CO2 increased from 1.027 W/m^2 in 1979 (relative to 1750) to 1.818 W/m^2 in 2011. Note that the total forcing from various greenhouse gases increased from 1.712 W/m^2 to 2.873 W/m^2.

  24. Is the primary effect of black carbon on arctic ice a change in albedo? I would have thought a chemical effect would be stronger.

  25. Paul Pierett says:@ February 17, 2014 at 2:25 pm

    Daniel says: replys @ February 17, 2014 at 6:00 pm
    …. is irrelevant anyway as it is the economic policies of the ruling elite which determine the price of wheat and pork bellies, the “who gets what”.

    Apologies for being off-topic but this kind of defeatist misinformation needs combating at every instance.
    >>>>>>>>>>>>>
    Thanks for saving me the trouble.
    I detailed how the ruling elite are making sure of starvation and volatile food prices HERE.

    You should read it Paul. It shows you just how despicable these people you listen to are.

  26. Enough has been said. Trembley nailed the issue. More wack a mole from the increasingly desperate AGW crowd. They cannot deny the pause, so now they use models to speculate on esoterics in under observed (except by satellite) regions.

  27. Anyone here listen to CBS News tonight to hear more of Jennifer Francis’s BS about how the Arctic is getting warmer? (Jennifer Francis, who was picked apart by Drs, Spencer and Pielke, Jr. at the Senate Energy and Environmen5t Committee hearing last year.). Of course the MSM lapdogs and ass-kissers and sycophants and idolaters did their thing, as usual.

  28. @DR

    A lot of what I read there would lead me to believe you’re right. The PR angle is there also. It may eventually earn a new name:
    Probably
    Not
    Actual
    Science

  29. just a reader. Sorry to be off topic, but I don’t know where else to put this comment. here goes, Ive been watching the arctic temperature chart every day and something is really bothering me, ahh screw the tip toe talk.! Why is 80 north having the warmest winter sense 1958? ( greenies this question is not for you, I don’t want a lecture on co2.

  30. snow says:
    February 17, 2014 at 11:42 pm

    just a reader. Sorry to be off topic, but I don’t know where else to put this comment. here goes, Ive been watching the arctic temperature chart every day and something is really bothering me…
    >>>>>>>>>>>>>>>>>>>>
    Polar vortex and the wandering jet stream.

    If cold air is drawn south it gets replaced by warmer air from the lower latitudes moving north. You can sort of see that in this image

    If you look at 2013 (click on left listing – link ) you can see the summer temp was quite stable and below average. This is when you have ‘open water’ to help stabilize the temperature and a reasonably zonal jet stream. In the winter the ocean is pretty much iced over and the temperature is dependent on the winds/vortex interaction. Click through other years and you can see how wildly the winter temperatures swing while the summer temperatures remain fairly constant.

    A halfway decent description of the polar vortex:

    http://climatechange.cornell.edu/what-is-a-polar-vortex/

    A decent description of the Jet stream and it’s heat transfer function

    http://www.decodedscience.com/zonal-meridional-flow-transfer-energy-atmosphere/41385

    You might find these interesting:
    Researchers Pinpoint 1,500-Year Cycle in Arctic Atmospheric Pattern (This is evidence of Bond events)

    The 1,800-year oceanic tidal cycle: A possible cause of rapid climate change

Comments are closed.