RealClimate logo

Unforced Variations: Apr 2020

Filed under: — group @ 1 April 2020

Open thread for climate science topics.

62 Responses to “Unforced Variations: Apr 2020”

  1. 1
    MA Rodger says:

    UAH TLT has posted with a March anomaly of +0.48ºC, a drop down below both the Jan & Feb anomalies (+0.56ºC & +0.76ºC respectively) but this is still the 3rd warmest March on record and the warmest non-El Niño March by some way.
    The warmest Marchs in UAH now run 2016 (+0.77ºC), 2010 (+0.51ºC), 2020 (+0.48ºC), 1998 (+0.47ºC), 2004 (+0.35ºC), 2019 (+0.35ºC), 2017 (+0.31ºC), 2018 (+0.28ºC) & 2007 (+0.28ºC).
    March 2020 sits =21st in the UAH TLT all-month anomaly record.
    Now a quarter-way through, the start to 2020 averages +0.60ºC, 2nd warmest on record after El-Niño-boosted 2016 (+0.73ºC) with 3rd & 4th spots also El-Niño-boosted, namely 1998 (+0.53ºC) & 2010 (+0.49ºC). The previous warmest un-El-Niño-boosted start to the year was 2017 down at +0.39ºC so the start of 2020 still has a claim to being “scorchyissimo!!!”

  2. 2
    sidd says:

    Is there a simple reason that climate models exhibit a double ITCZ but reality doesn not ?


  3. 3

    s 2: Is there a simple reason that climate models exhibit a double ITCZ but reality doesn [sic] not ?

    BPL: Not all climate models exhibit a double ITCZ.

  4. 4
    sidd says:

    Re: double ITCZ

    In my experience, most do. I’d like a link to some that do not. But my question remains, is there a ‘simple’ reason for the double ITCZ in models?


  5. 5
    mike says:

    Ugly and noisy:

    Daily CO2

    Apr. 1, 2020: 415.81 ppm
    Apr. 1, 2019: 411.69 ppm

    waiting to see March monthly numbers


  6. 6
    MA Rodger says:

    sidd @4,
    You ask if there “is a ‘simple’ reason for the double ITCZ in models” and I think you’ll find the answer is “No.”

    There is a CarbonBrief article from Dec 2018 on the workings (or not) of climate models and the section ‘What are the main limitations in climate modelling at the moment?’ discusses Clouds, ITCZ and Jetstreams, providing ample links for the curious. The article quotes Dr Baoqiang Xiang saying “The causes of the double ITCZ in models are complex.”
    The CarbonBrief article, citing Lin (2007), also states “Most GCMs show some degree of the double ITCZ issue,” which means there are “some that do not” although Tian & Dong (2020) [abstract] says of CMIP3, CMIP5 and CMIP6 moidels “We find that the double‐ITCZ bias with a big inter‐model spread persists in all CMIP models.”

  7. 7

    sidd says:

    “Is there a simple reason that climate models exhibit a double ITCZ but reality doesn not ?”

    There are several strange behaviors along the equator that are not well understood, likely because it shows edge state characteristics of a topological insulator.

    GRL 3/28/2020
    “The double‐Intertropical Convergence Zone (ITCZ) bias is one of the most outstanding errors in all previous generations of climate models that may reduce the reliability of future climate prediction based on models. The models have two ITCZs (i.e., zonally elongated narrow belt of high precipitation) in both hemispheres over the equatorial central and eastern Pacific and Atlantic instead of one ITCZ over the northern hemisphere in observations except for a short period in March and April. “

  8. 8
    MA Rodger says:

    RSS TLT has posted with a March anomaly of +0.84ºC, a drop down below both the Jan & Feb anomalies (+0.88ºC & +1.01ºC respectively). All latitudes showed a drop in anomaly, (including ‘North mid-latitude’ which at +1.4995ºC is no longer off-the-chart as it was last month). March 2020 is the 2nd warmest March on the RSS record (3rd in UAH).
    The warmest Marchs in RSS now run 2016 (+1.10ºC), 2020 (+0.84ºC), 2010 (+0.76ºC), 2019 (+0.72ºC), 2017 (+0.62ºC), 1998 (+0.60ºC), 2018 (+0.57ºC), 2004 (+0.56ºC) & 2015 (+0.51ºC).
    March 2020 sits 11th in the RSS all-month anomaly record (=21st in the UAH TLT).
    Now a quarter-way through the year, the start to 2020 averages +0.91ºC, 2nd warmest on record after El-Niño-boosted 2016 (+1.08ºC) with 3rd spot El-Niño-boosted 2010 (+0.72ºC), 4th 2019 (+0.69ºC) and 5th 2017 (+0.66ºC), followed by 1998 (+0.65ºC), 2007 (+0.57ºC), 2015 (+0.55ºC) & 2018 (+0.55ºC).

  9. 9

    One of the likely reasons that models create the double-ITCZ is because spherical harmonics symmetry dictates a tripole pattern if the region in between comprises the ENSO dipole.

    “Most current general circulation models (GCMs) still suffer from the double intertropical convergence zone 25 (ITCZ) problem (Mechoso et al, 1995; Dai, 2006). They fail to simulate a single ITCZ north of the equator year-round. Instead, they produce a second maximum of precipitation south of the equator in the Pacific and Atlantic oceans during at least half of the year, whereas it is only observed in the eastern Pacific during boreal spring (Hubert et al, 1969; Zhang, 2001).”

    The behavior of ENSO requires only a single nodal crossing to make it consistent with the single spring predictability barrier (i.e. none for fall), therefore the tripole pattern may be weakened for one hemisphere.

    This is in contrast to the upper atmosphere equatorial QBO, which does show a double nodal crossing (i.e. semi-annual predictability barriers) and also clear hemispherical side-lobes that sit astride the QBO monopole (an index=0 dipole).

    So I would say that the oceanic models are not asymmetric enough in their structure, but the upper atmospheric model is the correct symmetry (which makes sense as the upper atmosphere has a uniform mirror symmetry).

  10. 10
    sidd says:

    Re: Double ITCZ

    Mr. Pukite, thanks for the link to the carbonbrief page.

    Here are some papers:

    doi: 10.1175/JCLI4272.1

    A paper by Lin from 2007 identifying three biases in models leading to double ITCZ and incorrect precipitation distribution:

    “the zonal sea surface temperature (SST) gradient–trade wind feedback (or Bjerknes
    feedback), the SST–surface latent heat flux (LHF) feedback, and the SST–surface shortwave flux (SWF) feedback”

    Lin mentions that GISS-ER does not exhibit the double ITCZ possibly because of the ocean model used, but that other models mostly do.

    doi: 10.1073/pnas.1213302110

    Hwang and Frierson (2013 ascribing the problem to “cloud biases over the Southern Ocean”

    doi: 10.1175/JCLI-D-15-0328.1

    Bischoff and Schneider (2016) trace it to the energy balance near the equator and use an aquaplanet model to explore the conditions under which a double ITCZ appears. Conceptually the most simple, but a very idealized simulation. On the other hand i quite like aquaplanet simulations …

    All papers are open access, chek them out.


  11. 11

    Sidd says:

    “On the other hand i quite like aquaplanet simulations “

    Thanks for the links. I can understand why the double-ITCZ would show up in models, as the hemispherical “aquaplanet” symmetry would imply it a first-order effect. But as this paper you link to implies, there is likely an asymmetry that would place one flow as a primary first-order effect and the opposite as secondary.
    “Previous studies have shown that the zonal-mean ITCZ displacement off the equator is negatively correlated with the energy flux across the equator; when the ITCZ lies in the Northern Hemisphere, energy flows southward across the equator, and vice versa. The hemisphere that exports energy across the equator is the hemisphere with more net energy input, and it is usually the warmer hemisphere.”

    When I was modeling ENSO vs QBO, it took me a while to realize how strongly the hemispherical asymmetry applies to ENSO in contrast to QBO. In the future it may be that the concept of a “hemispheric tide” is used to attribute the forcing for the flow of energy flux across the equator. The tidal clue is key — lots of things fall into place after that. Thinking about this a lot recently, see a blog post I wrote last month

  12. 12
    mike says:

    @ al at 8: So hottest ever non-El nino March. I can’t process your numeric stuff, so will just ask questions. Can you list 4 or 5 of the other hot non-en march years and the temp delta for each? I would like to have a sense about how March 2020 looks in a set of similar Marches.

    Other question: do you have any suspicion that we are seeing a temp bump from decreasing aerosols related to slowed global economy?

    An aside about CO2: I think we might be seeing a bump in atmospheric CO2 related to the warming that has eclipsed the slight drop that we might otherwise have seen due to slowed global economy. I can’t crunch the numbers to show that, but the atmospheric CO2 number look a bit higher than I had hoped for with slowed global economy.



  13. 13
    Guest (O.) says:

    Nice graph for paleoclimatology …

    … found on the german Wikipedia in an article about climate change.

    The article:

    The graph:

    It’s the old graph pimped up with another graph plus source information…

  14. 14
    MA Rodger says:

    mike @12,

    The RSS March anomaly list @8 with non-El-Niño-boosted Marches enboldened runs – 2016 (+1.10ºC), 2020 (+0.84ºC), 2010 (+0.76ºC), 2019 (+0.72ºC), 2017 (+0.62ºC), 1998 (+0.60ºC), 2018 (+0.57ºC), 2004 (+0.56ºC) & 2015 (+0.51ºC). (Looking at MEI, I think 2004 just about rates as being El-Niño-boosted.) Perhaps this RSS year-on-year graphic (usually 2 clicks to ‘download your attachment’) may assist in visualising the context of recent RSS TLT global temperature anomalies.
    (Mind, as an indicator of AGW, I don’t put a great deal of store in RSSv4 managing to give a good account of itself when plotted out against global surface records. If you look at a hemispherical level, that ‘good account’ looks a lot less ‘good’. So my attempts to demonstrate graphically what may be potentially accelerating AGW rest on GISS & other SAT [eg here – 2 clicks) and not on TLT.)

    Concerning aerosols & Covid-19 – Okay the TLT record is atmosphere so a sudden change in atmosperic absorbing properties would have quick impacts but how much of the aerosol effects is that and how much albedo? Taking it as a simple climate forcing & its effect on surface temperatures, a back-of-fag-packet calculation on the likely scale of any aerosol effect caused by their diminution, consider a reduction of 0.74Wm^-2 which is likely a big chunk of such forcing and also equal to 20% of 2xCO2.
    Thus with ECS=3 we’d expect an eventual impact of -0.6ºC but only 40% of that would have appeared in a decade so 4% per year or 0.4% per month. It might begin to become significant/noticeable within global temperatures if it lasts a couple of years or so.

    And in my view, the daily/weekly CO2 increases on last year are presently high (at time of writing) not because of a CO2 uppy-blip this year but a CO2 downy-bulp last year.

  15. 15
    mike says:

    Ugly weekly number on CO2:

    Last Week

    March 22 – 28, 2020 415.52 ppm
    March 22 – 28, 2019 411.24 ppm

    Consider the context: A significant portion of the global CO2-emitting economy is idle because of Covid 19.

    Fossil fuel aerosols are likely down along with the CO2 emissions, so some particulate reduction of global warming is likely reduced.

    Basic global warming with a boost now from some loss of particulate reduction is expected to impact the natural carbon cycle in a manner that is generally expected to see a bump in CO2 accumulation in atmosphere and oceans.

    It’s too soon to make too much of a weekly CO2 number that is inherently noisy, but that does not mean we should ignore these data points.

    Stay well, don’t feed the trolls.


  16. 16

    This is an interesting recent paper

    Sutherland, B. R. & Jefferson, R. Triad resonant instability of horizontally periodic internal modes. Phys. Rev. Fluids 5, 034801 (2020).

    Alberto said:

    “The models without this problem have very poor predictive performance.”

    My model matches the data and has a significant asymmetry a la a single ITCZ.

    What I don’t understand is your implication that the state-of-the-art models can have good predictive performance yet also have the double-ITCZ problem — by definition they would be poor if they don’t match all the empirical data.

  17. 17
    nigelj says:

    Could the Atlantic Overturning Circulation ‘shut down’? 11 February 2020

    This guest post is by:Dr Richard Wood, who leads the climate, cryosphere and oceans group at the UK’s Met Office Hadley Centre. Dr Laura Jackson, a scientist in the same group.

  18. 18
    Karsten V. Johansen says:

    Mike #16: possible explanation for the steeper rise in ppmv CO2:

    Maybe we are now very close to important tipping points. This could manifest itself in springtime over the northern hemisphere. It seems to me that it is often then we register the most extremely high temperature values. Fx in 2017 the temperatures in western Canada were twenty degrees C above normal for around a month. Here in Scandinavia and lots of other regions we are seing this, spring by now often arrive in an explosive manner now. May 2018 had mean temperatures like a warm scandinavian july!

    These extremes could be partly because it is in the spring we have the yearly culmination of the CO2-level, just before the greening of lots of vegetation and the blooming of photosynthezising microorganisms at sea, which then creates a drawdown in tropospheric CO2 until the Nhemisphere autumn. So maybe we are now in springtime just tipping above some temperature/heat content tropospheric border values in the system, beyond which we tip over into quite another climatic regime, fx. regarding the global wind system. Something like this has happened before. Around 12 ka BP it has been shown that the global wind regime changed very suddenly:

    Of course some of these extreme spring temperature readings in recent decades are also due to the fact that northern skies have less cloud-cover in springtime, because the spring heating is “drying out” the airmasses, it takes some time before the oceanic surface heats up enough to let the evaporation catch up.

  19. 19
    MA Rodger says:

    Copernicus ERA5 Reanalysis has been posted with a March anomaly of +0.68ºC, a drop down below both the Jan & Feb anomalies (+0.77ºC & +0.80ºC respectively). March 2020 is the 4th warmest March on the ERA5 record (March 2020 2nd in RSS TLT, 3rd in UAH TLT).
    The warmest Marchs in ERA5 now run 2016 (+0.82ºC), 2019 (+0.70ºC), 2017 (+0.69ºC), 2020 (+0.68ºC), then a bit of a gap down to 2018 (+0.47ºC), 2010 (+0.46ºC), 2015 (+0.42ºC), 2002 (+0.39ºC) & 2014 (+0.28ºC).

    March 2020 sits 13th in the ERA5 all-month anomaly record (11th in RSS TLT, =21st in the UAH TLT).

    Now a quarter-way through the year, the start to 2020 averages +0.75ºC, 2nd warmest on record after El-Niño-boosted 2016 (+0.82ºC) with 3rd spot 2017 (+0.66ºC), 4th 2019 (+0.56ºC) and 5th 2018 (+0.45ºC), followed by El-Niño-boosted 2010 (+0.42ºC), then 2015 (+0.40ºC), El-Niño-boosted 2007 (+0.35ºC), 2002 (+0.29ºC) & 2005 (+0.28ºC).
    To assist in assessing the level of ‘scorchio!!!’ we are experiencing, a year-on-year graph of ERA5 monthly anomalies is posted here (usually 2 clicks to ‘downlaod your attahcment’).

  20. 20


    “This is an interesting recent paper

    Sutherland, B. R. & Jefferson, R. Triad resonant instability of horizontally periodic internal modes. Phys. Rev. Fluids 5, 034801 (2020).

    Application of the analysis here =>

    Hope this can help with the double-ITCZ issue. The best way to gain an understanding is to start with the highest-symmetry and lowest DOF region and work off of this foundation and outward from this region.

    So let’s build from the foundation:

  21. 21
    nigelj says:

    The irreversible emissions of a permafrost ‘tipping point’. Dr Christina Schädel, assistant research professor in the Center for Ecosystem Science and Society at Northern Arizona University. 12.02.2020.

  22. 22
    MA Rodger says:

    With the ESRL MLO CO2 for March posted, an update is due on the modelling exercise I kicked off back in August. There is still a final number in the table to fill in. (The number graphed out [usually two clicks to ‘download your attachment’] has all its dots now as it doesn’t plot the smoothed MLO data.) And the blank spot on in the table is not all that’s needed to complete the exercise. I have been using a predicted Global CO2 values post-May 2019 simply equal to +2.5ppm/year so there is the actual Global CO2 numbers to put through the mix. Given all the complications, it’s probably best to leave a proper assessment of the exercise until all the numbers are in. And with Global numbers reporting two-months behind & then subject to substantial revision for two-to-three following months, that will not be fore some time yet.

    12-Month MLO CO2 increase (ppm/yr) Dec numbers below next table
    … … … … … Met Office… … … .Modelled… … … … ..Actual… … … .. ..Actual
    … … … . forecast [Smothd]. .[Original,Smoothed]… .[Unsmoothed]… .[Smoothed]

    Jan19 … … … 2.64 … … … … … 2.74 … … … … … … 2.87 … … … … 2.85
    Feb19 … … … 2.64 … … … … … 2.92 … … … … … … 3.43 … … … … 2.95
    Mar19 … … … 3.04 … … … … … 3.13 … … … … … … 2.56 … … … … 3.02
    Apr19 … … … 3.24 … … … … … 3.10 … … … … … … 3.08 … … … … 3.02
    May19 … … … 3.38 … … … … … 3.16 … … … … … … 3.42 … … … … 3.21
    Jun19 … … … 3.22 … … … … … 3.24 … … … … … … 3.13 … … … … 3.20
    Jul19 … … … 3.00 … … … … … 3.07 … … … … … … 3.06 … … … … 3.05
    Aug19 … … … 2.86 … … … … … 2.94 … … … … … … 2.96 … … … … 3.02
    Sep19 … … … 2.63 … … … … … 2.78 … … … … … … 3.03 … … … … 2.84
    Oct19 … … … 2.42 … … … … … 2.66 … … … … … … 2.53 … … … … 2.60
    Nov19 … … … 2.44 … … … … … 2.44 … … … … … … 2.25 … … … … 2.49
    Dec19 … … … 2.46 … … … … … 2.13 … … … … … … 2.69 … … … … 2.50
    Jan20 … … … 2.48 … … … … … 2.07 … … … … … … 2.57 … … … … 2.54
    Feb20 … … … 2.58 … … … … … 2.06 … … … … … … 2.36 … … … … 2.49
    Mar20 … … … 2.89 … … … … … 1.87 … … … … … … 2.53

  23. 23
    Adam Lea says:

    Here is some pessimistic viewing about the climate impact of the pandemic from a scientist, theorising that the global pandemic will only cut CO2 emissions by a small amount this year, and the increase will start again next year. It is based on the experience of previous major global recessions/depressions where country leaders have pumped huge monetary stimulus packages to get their economies kick started again, particularly in heavy industry and construction.

  24. 24
    GBR says:

    21st century “Battle of the Coral Sea” already lost?

    Great Barrier Reef’s third mass bleaching in five years the most widespread yet

    Government’s chief marine scientist says he fears people will lose hope for the future of the reef but it is a clear signal for action

    apparently 60% of the whole reef has been bleached but they will not know for another 6 months how much of that will die or survive. Guesstimates are half will die.

    Last bleaching event I heard say reefs needed a 10 year break between bleaching heat stresses to recover properly.

    In other news from years ago projections of reef scientists (iirc) was that unless temps held below 1.5C GBR was likely to mostly dead and gone circa 2050.

    From 2016 interview with Dr John (Charlie) Veron, the ‘Godfather of Coral’

  25. 25
    GBR says:

    … 2.13 … … … … … … 2.69 … … … … 2.50
    … 2.07 … … … … … … 2.57 … … … … 2.54
    … 2.06 … … … … … … 2.36 … … … … 2.49
    . 2.89 … … … … … 1.87 … … … … … … 2.53

    Time for a change?

  26. 26
    GBR says:


    … 1.87 … … … … … … 2.53

  27. 27
    mike says:

    At Karsten: yes, thanks for mentioning and linking to the bit about ocean capacity to moderate CO2 accumulation in the atmosphere. That relationship is pretty important. Here’s a prediction: we are about to see the hottest summer on record. I believe that will be the case because aerosols that reflect solar radiation have been reduced by the economic downturn driven by Covid 19.

    Probably also worth mentioning that this will likely occur without a heat boost from EN warm cycle. Al would know for sure, but I think that means two the hottest years on record will be in place at 1 and 3 with 2019 in #3, an EN year in #2 and 2020 in #1 (if that is the way it works out for 2020).

    I am following your CO2 numbers, Al. I think the covid downturn is a confounding variable that should be expected to reduce CO2 ppm in atmosphere, and yet the number coming in appear to be higher than you plotted without any knowledge that a covid downturn was coming. Is that correct?

    March CO2

    Mar. 2020: 414.50 ppm
    Mar. 2019: 411.97 ppm


  28. 28
  29. 29
    MA Rodger says:

    mike @28,
    The difficult question with any Covid-19 impact on CO2 emissions is the estimated drop over the year. There is a lot of big numbers being mentioned for specific and short-term reductions (eg a 25% drop in China’s emissions at the start of the year reported here) but the global annual figures being toted seem to be 4% or 5% with some a lot lower (eg 0.2% to 1.2% argued here).
    If we run with a 5% drop in global CO2 emissions, that would equate to just -0.26ppm, not a large impact within the natural wobbles. For comparison, the Global Carbon Project put the last five years emissions increase (2014-18) averaging 1%-per-year and you may recall the fun trying to spot any resulting increase in atmspheric CO2 levels.

    On the subject of the year-on-year CO2 increase over the last 3 weeks being particularly high (averaging 3.8ppm in the MLO numbers) after a 6 week period of small increases (averaging 2.0ppm), I should point out that the Met Office 2020 CO2 forecast numbers do give a spike in the MLO CO2 increase for March (+2.93ppm) and April(+3.48ppm) with the rest of the year after April averaging +2.8ppm. The spike, indeed the whole forecast has to be based on something, presumably something that includes ENSO but not Covid-19. My own efforts at a CO2 projection are simply based on the difference between Global CO2 averages and MLO CO2 from 12-months ago (with an assumed value for Global CO2 this year) and such considerations suggest (with the Global numbers as at present) a low increase through the first half of the year (averaging +1.9ppm) and higher though the second half (+2.6ppm).

    And I should also point out that referring to me as ‘Al’ on the basis that my Facebook persona uses that name will surely present some difficulties for the many folk who do not appreciate my use elsewhere of that Facebook persona.

  30. 30
    Ray Ladbury says:

    Steven Emmerson: “S. Fred Singer has died.”

    To paraphrase Dorothy Parker, “Really? How could they tell?”

    I won’t believe it ’til they drive the stake through his heart.

  31. 31
    MA Rodger says:

    Both GISS & NOAA have posted their March global temperature anomalies, NOAA with the anomaly unchanged since February (still at +1.16ºC) and GISS with a small drop (down from +1.25ºC to +1.19ºC). March 2020 is the 2nd warmest March in both the GISS record and the NOAA record (March 2020 positioned 2nd in RSS TLT, 3rd in UAH TLT, 4th in the ERA5 reanalysis).

    Hottest March League Tables
    2016 … … +1.36ºC — — | — — 2016 … … +1.31ºC
    2020 … … +1.19ºC — — | — — 2020 … … +1.16ºC
    2019 … … +1.18ºC — — | — — 2019 … … +1.09ºC
    2017 … … +1.16ºC — — | — — 2017 … … +1.08ºC
    2015 … … +0.96ºC — — | — — 2015 … … +0.92ºC
    2010 … … +0.91ºC — — | — — 2018 … … +0.89ºC
    2018 … … +0.90ºC — — | — — 2010 … … +0.88ºC
    2002 … … +0.88ºC — — | — — 2002 … … +0.82ºC
    1990 … … +0.80ºC — — | — — 1990 … … +0.77ºC
    2014 … … +0.79ºC — — | — — 2014 … … +0.77ºC
    2008 … … +0.74ºC — — | — — 2008 … … +0.75ºC
    2005 … … +0.74ºC — — | — — 2005 … … +0.74ºC

    March 2020 sits 4th in the GISS all-month anomaly record and =3rd in NOAA (11th in RSS TLT, =21st in the UAH TLT, 13th in ERA5 reanalysis).

    Hottest Jan-March years (identical sequence for GISS & NOAA over top eleven places.)
    2016 … … +1.30ºC … … … +1.23ºC
    2020 … … +1.20ºC … … … +1.15ºC
    2017 … … +1.11ºC … … … +1.03ºC
    2019 … … +1.02ºC … … … +0.96ºC
    2015 … … +0.90ºC … … … +0.88ºC
    2018 … … +0.86ºC … … … +0.80ºC
    2010 … … +0.83ºC … … … +0.80ºC
    2002 … … +0.82ºC … … … +0.77ºC
    2007 … … +0.81ºC … … … +0.76ºC
    1998 … … +0.70ºC … … … +0.72ºC
    2014 … … +0.70ºC … … … +0.68ºC

    The last decade of monthly global anomalies for GISS, NOAA, HadCRUT, RSS & UAH graphed out here(usually 2 clicks to ‘download your anomaly’).

  32. 32
    Al Bundy says:


    Hmmm. By definition permafrost is dirt that accumulated when it was NOT frozen. So thawing it can represent an opportunity for further dirt to accumulate. And the total carbon contained in said dirt is not terribly relevant; what matters is whether the biome involved will grow or shrink the amount of dirt (and methane effects, of course).

    Not claiming anything much here. Just opening up a can of worms and noting that we have the ability to influence biomes. I don’t see why we should assume that modern thawed biomes as guided by humans will by definition not just under-perform but inversely perform as compared to the thawed biomes of the past.

    Your handle is an acronym for Great Barrier Reef. That a happenstance? And yes, the GBR is toast. I’d bet there is no way to save it. The only hope might be to replace it with a reef made up of totally different breeds of coral, perhaps from the Red Sea, perhaps with GMO coral developed in labs.

    Ray L: I won’t believe it ’til they drive the stake through his heart.

    AB: Did Singer ever realize how much damage he did? Or does the old saw about advancement not being achieved through winning over but plowing under apply?

  33. 33

    #32, MAR–

    Thanks once again for updating us in your usual thoroughgoing manner.

    But when, oh when, will the long-promised cooling start? ;-)

  34. 34
    Mr. Know It All says:

    Oh my goodness, look what the pandemic models tell us about climate change models – excellent article and comments:

  35. 35
    Ray Ladbury says:

    Al Bundy,
    Max Planck famously said, “Physics advances one funeral at a time.”

    Or, as Tommy Lee Jones says in Men in Black: “A person is smart. People are dumb, panicky, dangerous animals, and you know it!”

    And I’m not sure about whether a singular person is smart.

  36. 36
    Ray Ladbury says:

    Oh, look. Mr. KIA provides us with evidence that Breitbart correspondents are utterly clueless about the purposes and process of scientific–and particularly epidemiological–modeling.

    A Prime example of a man who is too stupid to know he should be embarrassed.

  37. 37
    nigelj says:

    Mr. Know It All @35 posts some opinion from ‘Breitbart’ , claiming the covid 19 models over estimated numbers of hospital admissions , so this is allegedly proof that climate models don’t work.

    This truly is intellectual nonsense on a gargantuan scale.

    You cannot compare the modelling of covid 19 and climate models, because we know very little about covid 19 because its only been around a couple of months and is a new virus, while we have at least 100 years of published research on climate change and associated climate data, so climate models are much more firmly grounded than covid 19 models.

    In addition, Breitbart choose the wrong metric with hospital admission rates, because what really counts is the mortality rate. This is the pointy end of data. Not sure what numbers models predicted, but they certainly predicted it would be shocking and the numbers of fatalities in New York are clearly shocking. Without very strong mitigation this will spread elsewhere.

    For general interest, the raw data clearly shows a covid 19 mortality rate of about 5% for New York (looking at 2000 covid 19 pandemic for the USA on wikipedia) , compared to 0.1% for seasonal flu, so covid 19 is over 50 times higher than seasonal flu. It’s likely that not all cases of covid 19 have been identified, but experts say this only reduces the mortality rate to about 2% at best, still 20 times worse than seasonal flu.

    Breitbart also cherry pick the states with the lowest hospitalisations, and ignore New York. Although New York is admittedly the worst case, its a red flag for what could happen elsewhere as the virus spreads, which cannot be ruled out. So the models might be wrong on the timing, but not the eventual outcome.

    Breitbart are just useless, and also have it backwards. Covid 19 is a warning of what will happen to the climate if we do nothing. Covid 19 is a lesson in the problem of accelerating growth curves and we have this with both covid 19 and various climate trends.

  38. 38
    nigelj says:

    Mr. Know It All @35, addendum to my previous comment. Breitbart claim models over predicted actual hospital admissions for covid 19, and this is based on models that took account of mitigation measures. It looks more like the models did not allow for all the actual lock down measures that have been implemented from what Ive read, so of course you would expect hospital admissions to be less than models predicted.

  39. 39

    #35, KIA–


    OOOOOO-kay, then.

  40. 40
    Ric Merritt says:

    Of course, anyone who disses climate models while ignoring their considerable and long-standing successes, well documented on this blog, is just a reality denier blowing smoke, who knows or should know that current energy and climate policies are the best attempt yet known at devastating industrial civilization, and immiserating or prematurely killing our grandchildren.

    This includes whoever wrote that mendacious Breitbart piece, and the real person (if there is one) who posts as Mr KIA.

    Lay off my grandchildren. Become a builder, not a destroyer. Reach for the best humanity can do, instead of bending your best efforts toward increasing misery for billions.

  41. 41
    John Pollack says:

    Mr. Know It All @35: I’ve never dealt epidemiological modeling, but I can see that running one would require a lot of information. The most important would be how many people a person with the virus would infect, and the incubation period. These determine an exponential increase in cases, and any uncertainty would be magnified many times as the numbers balloon.

    If I were the intelligent but skeptical recipient of such model numbers, say a county health commissioner or hospital director, my reaction would not be as Mr. Nolte seems to assume. His choices appear to be “an expert has spoken, so I must believe these numbers” or “it’s from a model, and being touted by an expert, so it’s worthless.”
    Instead, I’d be asking questions: These are really large numbers! What are your critical assumptions in the model? What are the error bars? Have you run models with other assumptions? How did they compare? How has this model performed in the past? I would also be looking around for other models, and inquiring about the professional reputation of the modeler.
    Meanwhile, and this is critical, I would be having an emergency meeting with my staff, saying that we had a preliminary estimate of an epidemic peak that could potentially overwhelm our hospital system, and asking what resources we have available to meet this problem.

    So, the purpose of the model is not to provide one specific number, but to indicate a range of possibilities – which would include being caught short on resources. A planner, once understanding that the error bars on the model numbers were very wide, would begin to look for extra resources, while monitoring updates and other estimates.

    Mr. Nolte merely wishes to use the model estimates merely as a foil for his simplistic approach, not with any intent to understand them.

  42. 42
    John Pollack says:

    Mr. Know It All @35 Pt. 2 The distinction between computer and weather models.

    I am going to get a little personal here. I was a weather forecaster for 30+ years. During my career, I relied on weather models, but not to the exclusion of other information. During that time, I saw the models get a LOT better. For one thing, the computers got a lot better, so that the models could incorporate more physical processes, and in more detail. For another, the observations got a lot better, especially from weather satellites. This was not incidental. The deficiencies in forecasting models were used to help determine which observations were worth spending extra money on. Also, a lot of research went into learning how to best incorporate those observations into the weather models. Epidemiological models do not have this amount of detailed observational history. In addition, the atmosphere obeys well-tested physical laws. People also follow social laws and norms that can’t be specified in the same way.
    By the end of my career, the forecast a week out was about as good as it was 60 hours out when I started, maybe better.

    When I did a forecast, there wasn’t a single day when my goal wasn’t to be 100% accurate. However, that was with the foreknowledge that I would always be wrong about something, and occasionally about something really important. So, the secondary goal was to try to use my time wisely to understand the important stuff, and get it as close to right as possible.
    Weather forecasting models are also always wrong. Some get close, and all are better some days and in some situations than others – but you can’t always tell in advance when a model is having a “good day.” So, an essential part of the forecasting process is comparing models to each other, and also to observations. As you get further out in time, the model solutions will diverge. In some respects, the average of a group of models becomes a better forecast than any individual model run, and it will certainly be more consistent. A disadvantage is that an extreme weather event will deviate from the average, so as time gets shorter, the forecast will get more and more drastic.

    Something else that you learn on the job is that in many harsh situations, it is better to over-forecast than under-forecast, if you can get yourself to believe that things will really get that extreme. The reason is that, just as you may not believe the weather models, other people may not believe you. People are understandably reluctant to believe that their life is at risk, or that the experts have a good handle on this situation, since they’ve sometimes been wrong before. However, when looking at the aftermath of a bad weather event, we always caught more grief when we had a weak forecast than when we had a harsh one, and things weren’t quite as bad as we indicated. A classic example is issuing tornado warnings. The forecast goal is to have a “lead time” so that we identify a potentially tornadic storm on the radar, and get a warning out before it hits. Sometimes, a tornado never develops. Often, it strikes a much smaller area than is included in the warning. If there was no tornado, we would get complaints afterward, of course. However, nobody wanted to be the one whose house got destroyed without warning just so we could be REALLY sure that the tornado had reached the ground.

    Yes, I know that I haven’t said anything about climate models yet. That’s for part 3.

  43. 43
    John Pollack says:

    To Mr. Know It All @135 – Part 2B. This one is even more personal. To me, Nolte comes across much like a wannabe or lousy forecaster. He is shocked, shocked, that a forecast could be wrong. Not that he could come up with a successful one, but it’s sure fun to be a Monday morning quarterback, and criticize the people who are actually making them. Of course, if you are doing that, it is also helpful to pick headline grabbing media depictions rather than original forecasts, and then claim that these are the original products of the “experts” that you are criticizing, rather than media interpretations. As somebody who has had my forecasts sometimes misinterpreted or butchered, I know that it can come out sounding a lot different at the hands of the media.

    “We’re still talking about ‘experts’ that our media and government grovel down to without question.” Really? As an expert forecaster, I expected to be questioned, because I was good, but not always right. I questioned myself, and forecasters questioned each other, so that we could improve. The people who used our forecasts also wanted to know how sure we were. The goal was to be as accurate as we could, and communicate as well as we could, not to have people “grovel.” That sounds like the projection of a weak ego to me, not somebody who takes responsibility for adult decisions, knowing that they will sometimes be wrong, and it will affect other people when you are.

    “We’re still talking about models with the goal of destroying our way of life, our prosperity, our standard of living, and our individual freedoms to live our lives in whatever way we choose.” Well, I guess that’s what he thinks the models are about. In my field, we called that category of stuff “wishcasting” because you start with a wish about the desired outcome, and then work backward to the forecast. Since it ignores or severely distorts reality, it produces a stream of horrible forecasts. It’s also a wannabe mistake, to think that you’ve got it all figured out, and there’s some plot afoot to keep the experts on top – not relating to the fact that they’ve actually put a huge amount of work into getting good at the stuff that they’re expert in.

  44. 44

    #42-4, John Pollack–

    Excellent comments, thank you. It’s illuminating to hear that personal experience.

    I’ve never been a forecaster, but my former father-in-law was–in fact, he ended his career as a very senior official in the then-Atmospheric and Environmental Service (AES) in Canada. So I got to hang out with meteorologists on occasion. They all had ‘war stories’–my father in-law, IIRC, had a ‘personal worst’ forecast record of missing the daily low in Hearst, Ontario, by something like 50 Fahrenheit degrees. *That* generated some complaints!

    He said the maddening thing about it was you could generally see perfectly well, in retrospect, why you had been wrong. Of course the front would speed up, or slow down, or weaken, or whatever! Prospectively, though, it was another story.

    He was a bit skeptical about this whole newfangled numerical prediction thing that was coming in at that time. But not so much so as to stand in its way, thankfully. Even as an outside observer, it’s highly apparent to me how much it has improved operational forecasting.

    FWIW, I can’t trust anything KIA’s Breitbart author wrote. He presents the forecasts without source, or sufficient detail to understand their context, or their underlying assumptions (which I suspect changed over the course of time). Given his obvious ideological bias–as you point out, he himself ascribes the goal of modeling to be “destroying our way of life, our prosperity, our standard of living, and our individual freedoms”–I must believe that even if he wanted to give context he couldn’t do so, because he doesn’t understand the context correctly in the first place. Essentially, he can’t help but cherry-pick.

    And probably doesn’t even try not to.

  45. 45
    John Pollack says:

    Mr. Know It All @135 Part 3 – weather vs. climate models. For the most part, these models are similar, particularly the atmospheric physics, but the goals differ. With the exception of some (mostly research) models, the forecasting models are intended to optimize coverage of weather over a short time in a lot of detail. The climate models are intended to cover a period from years out to millenia. They can be optimized for a huge variety of different things related to understanding the behavior of the climate system, which includes oceans, land surface, cryosphere, vegetation, and other variables. The results are not deterministic forecasts, but expressed typically in averages and ranges of many climate model runs. Models are the only way to do any kind of controlled experiment over large periods of time and amounts of space. However, they are iterative like the forecasting models; discrepancies between models direct attention to critical uncertainties that require better observations, and are used to improve later model versions.

    Weather forecasts can be verified in real time, but climate models require lots of time. Any evidence of past climate conditions that can be dated serves as a valuable set of observations to ground the climate models in reality. A complicating factor is that the quality and availability of these proxy records typically changes over time, so a long record is valuable.

    From my personal perspective as a forecaster, I was reluctant to really put a lot of credence in the climate models through much of the 1980s. It was clear that CO2 and methane were greenhouse gases that could alter the climate, but well dated records of how these had really changed in the past were very sparse. I thought (wrongly, as it turned out) that the climate models could be subject to some kind of subtle but cumulative error that would really throw off the projections. The big breakthroughs started coming in the early 1990s when a 100,000 year ice core from Greenland became available. What first jumped out at me were the huge and sometimes very abrupt changes in temperature shown over time. This clearly established that the climate system was highly nonlinear, and at times chaotic, through direct geophysical evidence. This was enough so that I will never be impressed with an argument that the climate has somehow failed to follow a linear response over a short period, falsifying a model. A good example of an unimpressive argument of this sort is the supposed “hiatus.”

    The results from the first long Antarctic ice core were so powerful that it made the hair stand up on the back of my neck when I first saw the research. It was clear that CO2 was not just a mover of climate, but a major factor. The temperature and CO2 curves were so strongly tied that there was no reason for it to happen otherwise. The modelers were right, and the Earth was speaking in a loud voice that CO2 REALLY MATTERS over time. That doesn’t mean that other stuff doesn’t matter, or that some particular value of climate sensitivity is just right. What it does mean, to me as a forecaster, is that we have to expect the HUGE changes in global climate resulting from the HUGE changes we’ve already made, and are still making, in CO2 concentrations. (Same for methane and various other greenhouse gases, to a lesser degree.) Because we have very strong geophysical evidence that backs up our climate models, the idea that CO2 has little effect on climate is no longer a “null hypothesis” argument. It flies in the face of the evidence we have, and requires strong justification that is lacking in the data.

    Again, as a forecaster, when I see a large change in a major mover, I expect a large response. I may not know exactly what it will be, or the timing, but I do know to expect something big to happen, and to warm people about it – even if they want to call me an “alarmist” before it hits.

  46. 46
    nigelj says:

    J Pollack @44 thanks for the interesting comments. Breitbart talks about “We’re still talking about models with the goal of destroying our way of life, our prosperity, our standard of living, and our individual freedoms to live our lives in whatever way we choose.”

    Nonsensical of course. Much of the denialist dislike of climate and virus modelling appears driven by politics, here is quite a dramatic example regarding covid 19:

  47. 47

    #47, nigel–

    An extreme example of the fallacy that the ‘right’ to do something also implies that doing it must be a good idea.

    Also, of course, that the only freedom that matters is the freedom immediately to gratify every whim, no matter how damaging to others that gratification might prove.

  48. 48

    Can anyone point me to annual time series data for global aerosol optical depth (AOD)? As long as possible.

  49. 49
    Al Bundy says:

    NigelJ: This truly is intellectual nonsense on a gargantuan scale.

    AB: I disagree. There’s nothing intellectual about it. As if Breitbart or MrKIA sees “intellectual” as anything except “the enemy”. Their goal is to sh*t all over anyone who got above a “C” in any STEM-related class.

    John Pollack: Something else that you learn on the job is that in many harsh situations, it is better to over-forecast than under-forecast, if you can get yourself to believe that things will really get that extreme.

    AB: Yeah. Note that MrKIA is whining about how the Covid19 forecasters drastically improved their take in four days, by which time they were seriously accurate. They did what John advocated by taking way sparse data and warning about what it might mean and then, unlike Trump, they did due diligence and Manned-Up by giving their more informed thoughts as soon as they possibly could.

    Obviously MrKia is just being a jerk, b*tching about four days! Once again, he’s lying through his teeth. Nobody could be so brain-dead as to believe the garbage he just spouted. MrKIA, is there ANYTHING that a non-denialist could say or do that you would interpret as “barely acceptable”?

    But, of course, MrKIA wasn’t even slightly interested in truth or rational thought. He got his jollies. So yep, he “won”.

    Kevin McKinney: FWIW, I can’t trust anything KIA’s Breitbart author wrote.

    AB (while pointing at Kevin): HERETIC!!!!

  50. 50
    MA Rodger says:

    Barton Paul Levenson @46,
    Not entirely sure what you ask for, but the KNMI Explorer provides the GISS annual mean strataspheric AOD back to 1850.