Climate sensitivity reconciled?

There’s a really interesting paper by Mark Richardson, Kevin Cowtan, Ed Hawkins, and Martin Stolpe called Reconciled climate response estimates from climate models and the energy budget of Earth. Ed already has post that explains it all very clearly, so I won’t need to go into too much detail.

The basic issue is that energy balance estimates for climate sensitivity use observed changes in temperature, and estimated changes in forcing, to determine the climate sensitivity. The results tend to suggest values lower than suggested by climate models. However, the temperature datasets used have some coverage bias (not all areas have coverage) and the temperature measurements vary from being air temperatures over land to sea surface temperaures over the oceans. For climate models, on the other hand, there is no coverage bias and the temperatures are typically air temperatures everywhere. Therefore, it’s not a like-for-like comparison.

Credit: Richardson et al. (2016)

Credit: Richardson et al. (2016)

Richardson et al. (2016) illustrate that this can make quite a difference. The figure on the left shows the HadCRUT4 temperatures (grey line), together with the models temperatures with full coverage and using air temperatures everywhere (red line), model temperatures determined using air and sea surface temperatures (red dotted), and model temperatures using air and sea surface temperatures and also masked to account for coverage bias in the HadCRUT4 data (blue triangles). The effect is not trivial, with as much as 0.2oC warming missing due to coverage bias and the use of sea surface, rather than air, temperatures.

Credit: Richardson et al. (2016)

Credit: Richardson et al. (2016)

The figure on the right shows the impact on climate sensitivity etimates (Transient Climate Response (TCR) only). The bottom blue bar shows the original Otto et al. result. The next is the Otto et al. method, but updated using Lewis & Curry forcings. The one above that, however, is the CMIP5 result, but using blended temperatures (air over land, and sea surface over ocean) and masked to compensate for coverage bias in the observational dataset. Essentially, they all produce best estimates of around 1.4oC.

The top two bars in the figure on the right, however, show the CMIP5 estimate using air temperatures only (top) and an observationally-based estimate that uses inferred air temperatures. They produce best estimates of 1.8oC (CMIP5 tas) and 1.7o (observationally-based, inferred tas). The key point is that if you correct the models by using blended temperatures and account for coverage bias, you get a result that is consistent with the observationally-based estimates. Similarly, if you correct the observationally-based estimates to account for blended temperatures and coverage bias, you get a result consistent with the model result. Essentially, this is an argument that there isn’t necessarily a discrepancy between climate model estimates and energy balance estimates; you just need to do a like-for-like comparison.

So, this suggests that there might not really be a discrepency between observationally-based and model-based estimates for climate sensitivity, but also suggests we might have to be careful as to what we mean when we discuss global temperatures. As Ed’s post concludes

Finally, if the reported air-ocean warming and masking differences are robust, then which global mean temperature is relevant for informing policy? As observed? Or what those observations imply for ‘true’ global near-surface air temperature change? If it is decided that climate targets refer to the latter, then the warming is actually 24% (9-40%) larger than reported by HadCRUT4.

Advertisements
This entry was posted in Climate change, Climate sensitivity, Global warming, Science and tagged , , , , , , . Bookmark the permalink.

106 Responses to Climate sensitivity reconciled?

  1. John Hartz says:

    Also see:
    Scientists find common ground over climate sensitivity by Roz Pidcock, The Carbon Brief, June 28, 2016

  2. Michael Hauber says:

    Does this mean anything for paleo estimates of sensitivity and temperature history? Naively I would expect paleoclimatology to be tuned to the observational record. If the observational record needs to be inflated by 24% to be on the same coverage/blending basis as the models, would that mean paleoclimatology needs to be inflated by 24% as well? Paleo and models are usually described as agreeing (although I haven’t looked into paleo stuff in much detail). Does this mean paleo estimates get pushed above models? Or is it all still within uncertainty levels?

  3. Michael, as far as I know the paleo estimates come from a local comparison of the proxy and the observations. Furthermore, locally most of the variability is in the year to year variability and not in the trend. Thus I would not expect an influence. If I am wrong, paleo people please correct me.

  4. Todd Friesen says:

    It would be interesting to see this analysis combined with the analysis produced by Kate Marvel and Gavin Schmidt with regard to varying TCR sensitivities to GHGs, land use, aerosols, ozone, etc.

  5. JCH says:

    From Climate Etc. …

    Here is why I think it might be better to use HADCRUT4 than a truly global data set for doing climate sensitivity analysis. The main thing missing in the HADCRUT4 analysis is the Arctic region. And the Arctic is hugely sensitive to the multi-decadal ocean oscillations, which was noted in the AR5:

    “Arctic temperature anomalies in the 1930s were apparently as large as those in the 1990s and 2000s. There is still considerable discussion of the ultimate causes of the warm anomalies in the 1920s and 1930s.” (IPCC AR5)

    “A recent multi-proxy 2000-year Arctic temperature reconstruction shows that temperatures during the first centuries were comparable or even higher than during the 20th century.”

    Assuming that these high amplitude variations in the Arctic have a substantial component from natural internal variability, this variation should not be included when you are trying to infer the externally forced variability from CO2. How much of the current warming and sea ice loss in the Arctic is natural versus forced by CO2 remains highly uncertain, the AR5 makes a very conservative statement:

    “Anthropogenic influences have very likely contributed to Arctic sea ice loss since 1979.”

    ‘Contributed’ – apparently the AR5 did not have sufficient confidence to say anything like ‘more than half’.

    So how was natural variability reconciled in this paper?

  6. Ethan Allen says:

    Well, I’m glad to see that there was absolutely NO coverage bias in 1860.

  7. Steven Mosher says:

    Here is what I dont get.

    using the same sparse data… with three different methods… we can demonstrate hadcruts issues.

    http://static.berkeleyearth.org/memos/robert-rohde-memo.pdf

    but still folks wont move on to better methods…

    heck Id be happy if folks adopted C&W

  8. Steven Mosher says:

    “Well, I’m glad to see that there was absolutely NO coverage bias in 1860.

    coverage bias is a function of where the warming happens. not the actual percentage of coverage

  9. Ethan Allen says:

    “coverage bias is a function of where the warming happens. not the actual percentage of coverage”

    Well excuse me.

    How about …

    Absolutely NO difference in the HadCRUT temperature and the blended AND masked AOGCM temperature in 1860 (see Supplementary Figure 6).

    1860-1880 is a pi** poor place to start for empirical/observational modeling BECAUSE of the extremely pi** poor surface temperature coverage.

    Now go back to Figure 1 and look at 1860-1920, all three curves are virtually identical (yeah 1860-1880 baseline). Heck, I don’t trust any reconstructions of Earth’s temperature prior to say 1950. It’s pretty much a flatline between 1860 and 1920.

    Flip, for alI know the AOGCM’s don’t account for the surface boundary layers at all (air and water) and certainly never will account for wind driven surface mixing via water waves.

  10. Harry Twinotter says:

    Uncertainty is no one’s friend. And underestimate is just as likely as an overestimate, despite what the climate change deniers wanted.

  11. Steven Mosher says:

    “1860-1880 is a pi** poor place to start for empirical/observational modeling BECAUSE of the extremely pi** poor surface temperature coverage.”

    err not really.

    http://berkeleyearth.lbl.gov/auto/Global/Animations/OceanPlusLand.mpg

    That time period is about 50-75% coverage.

    Its so funny.. when skeptics want to prove the MWP was warmer they love talking about a couple places as if they were the globe.

    But when the foot is on the other hand

  12. Steven Mosher says:

    “Heck, I don’t trust any reconstructions of Earth’s temperature prior to say 1950.”

    ahh you just vanished the dust bowl.. it wasnt hot in the 30s?

    Why then do skeptics point out that the warm 1910 to 1940 is not captured by the models?

    that warming wasnt real?

    was any warming warming before 1950 real? was there an LIA… I dunno you dont trust any reconstructions… it could have been warmer during those frost fairs.. ya…

  13. MarkR says:

    Michael Hauber: We looked at transient response which you can’t get yet from palaeoclimate because of the time resolution of proxies. Our TCR largely agrees with models and models largely seem to agree with palaeoclimate estimates, but our results hint that we should maybe check which temperatures exactly we’re referring to with the palaeoclimate evidence. I don’t know enough about them to say more.

    Tood Friesen: Kyle Armour wrote a commentary on the paper in the same journal:
    http://www.nature.com/nclimate/journal/vaop/ncurrent/full/nclimate3079.html
    He included the Marvel approach plus some knowledge about time-sensitivity of feedbacks. I’d say that Marvel et al. needs to be replicated in more models before we know how big it’s likely to be IMO. The full model range of the effect we identified is 9-40 % for example, so model choice might matter. The Schmidt updated forcing seems to be worth 10-15 % as well but it would need more rigorous work to check how this works with the observational forcing we use, since that is already different from the CMIP5 forcing Gavin was updating.

    It’s also possible that weird temperature change patterns and issues in changing SST measurements around WWII could shift the results one way or the other. And there are suggestions that aerosol cooling is overestimated (Stevens and the recent CERN work) which would lead to a lower best estimate of TCR. Our uncertainties are large enough to cover the range of values these hint at though.

  14. MarkR says:

    JCH: Natural variability was included in our estimate of +-0.2 C uncertainty in the temperature change for subtracting a 20-year average from a 10-year average. Supplementary Tables 12 and 13 show the effect of increasing or decreasing this uncertainty. Reducing it by 75% gives TCR of 1.7 C (range 1.2-3.2 C), increasing it by 100% gives TCR of 1.7 C (range 0.7-3.5 C).

    The extra variability of the high latitude regions is included in the original estimate. Curry is right that if you wait to see which regions warm fastest and then throw them out then you get less warming.

    Arctic sea ice extent in summer appears to be far lower than at any other time in at least 1,450 years.
    http://www.nature.com/nature/journal/v479/n7374/full/nature10581.html

  15. MarkR says:

    Ethan Allen:
    RE: “for alI know the AOGCM’s don’t account for the surface boundary layers at all (air and water) and certainly never will account for wind driven surface mixing via water waves.”

    >> Some models include the “effects of surface waves, Langmuir circulations, submesoscale
    eddies, as well as additional wind mixing” (Huang et al., doi: 10.1002/2013JC009535)

    The periods we used are based on Otto et al. (but one period was shifted by a year to include simulations beginning in 1861), and since the calculation uses the differences in temperature the choice of zero point makes no difference.

    Supplementary Table 6 shows that shifting the beginning period forward to 1900-1919 instead leads to observations looking hotter compared with models than the results we reported (at the 50-74th percentile of models, instead of the 33-51st percentile for our baseline).

  16. Dikran Marsupial says:

    This sort of careful attention to detail and checking out possible sources of bias is an important way in which science moves forward. I suspect in many fields (certainly true in mine) these kinds of bias are assumed to be benign and not investigated because there are more glamorous/interesting things to research instead, which is a shame, because there is often a really good paper to be written showing that they are not as benign as previously thought (even if adoption can indeed be slow even after this has been demonstrated).

  17. MarkR says:

    Further points on what Curry brings up:

    1) we use forcing in the same way as Lewis & Curry (2015). Yes, partial coverage could affect this and contribute to uncertainty, but we use the same approach so the LC15 results actually largely agree with models. I don’t remember seeing this issue raised by Curry related to that paper.

    2) Yes, some regions have more variability but this is no justification for throwing out the fastest-warming regions. This variability is in our reported uncertainties.

    3) Sure, Curry does not like infilling, but those methods have been validated and shown to work with known uncertainties. I don’t think we should throw out work that’s been done following standard scientific methods, even if the numbers disagree with Curry’s opinion. That’s not relevant for our main reconciliation result though.

    4) I saw little discussion of the blending effect which accounts for a difference of about 9%.

  18. Dikran Marsupial says:

    “Sure, Curry does not like infilling.” I would hope that she doesn’t like known biases either (and presumably provides caveats regarding them in her published work). At the end of the day you have to do one or the other – pretending the biases don’t exist is the unacceptable option.

  19. Mark,
    Thanks for all the comments and clarifications.

  20. Ethan Allen says:

    MarkR,

    Thanks for the link to …
    Evaluating CMIP5 simulations of mixed layer depth during summer
    http://onlinelibrary.wiley.com/doi/10.1002/2013JC009535/pdf
    (parameterizations, but can’t expect much more, ever even)

    I do have a question in reference to Figure 1 (kind of hoping all data used in that plot are in some of the ascii files), that I am unable to find in the SI_richardson16tcr.zip file …

    1) 1pctCO2
    3) hist_rcp85

    Both state …

    “Contains folders Txax and Had4, definding blending method. Each file contains 4 columns: time, air temperature anomaly, blended anomaly, air temperature minus blended anomaly.”

    Looks kind of boilerplate cut-and-paste to me.

    1) 1pctCO2 has four columns, 1st column starts at x.x or xx.x and ends at xxx.x, so zero = ?
    3) hist_rcp85 84 CMIP5 files in two directories, year then only two columns of data?

    SM,

    NOAA/GISS start in 1880, I wonder why? Also see Supplementary Figure 6 of the current paper …
    “Supplementary Figure 6: Change in near-surface air temperature from 1861-1880 to 2000-2009 seen globally (left), seen by typical HadCRUT4 data coverage over 2000-2009 (centre) and by typical HadCRUT4 data coverage over 1861-1880 (right). Typical coverage refers to cases where more than 25 % of months within that period report data.”

    I’m now thinking the authors should have run daily masks exactly matching AOGCM’s with HadCRUT data areas. Ha Ha 🙂

  21. Ethan Allen says:

    Hmm, interesting ECMWF reanalysis report submitted for publication in the Quarterly Journal of the Royal Meteorological Society …

    Estimates of variations and trends of global surface temperature
    http://www.ecmwf.int/sites/default/files/elibrary/2016/16437-estimates-variations-and-trends-global-surface-temperature.pdf

    From p. 8 (4 Global temperature trends) …

    “The full-period trends differ little among the datasets. ERA-Interim, JRA-55 and GISTEMP give warming rates of 0.17°C/decade, HadCRUT4 gives 0.18°C/decade while NOAAGlobalTemp gives 0.16°C/decade. Trends range from 0.16 to 0.19°C/decade for the HadCRUT4 ensemble, and Had4_UAH_v2 and Had4_krig_v2 both give 0.18°C/decade. The trends from both reanalyses are slightly smaller if sea-surface temperature rather than marine air temperature is used, rounding to 0.17°C/decade again for ERA-Interim but to 0.16°C/decade for JRA-55. Trends are slightly smaller still if background values of air temperature are used over land rather than the values from analysing screen level observations: in this case the trends from both reanalyses round to 0.16°C/decade. Although the full period concludes at a time of extreme values, the largest 30-year trend within the period runs from1982 to 2011 in five of the six datasets, and from 1981 to 2010 in the other.”

    See also pp. 21-2 Figure 10 …

    “Figure 10 Surface temperature anomalies (°C) relative to 1981-2010 from (a, b, c) HadCRUT4 (d, e, f) Had4_krig_v2, (g, h, i) NOAAGlobalTemp and (j, k, l) GISTEMP for 1881-1910 (left), 1911-1940 (centre) and 1941-1970 (right), and from the ERA-20CM ensemble-mean for (m) 1911-1940 and (n) 1941-1970. Grid boxes where values are missing are coloured grey. Lighter grey colouring indicates boxes that would have had values had maps been presented as anomalies relative to the datasets’ standard reference periods. Two small regions around Antarctica where cold anomalies exceed 4°C have been shaded at the 4°C level in panels (m) and (n).”

    “There are other sources of uncertainty, however, which lead to a quite considerable spread among the HadCRUT4 ensemble for the 19th century and early decades of the 20th century. A single member stands out for almost thirty years from around 1890 in having a much warmer average temperature over Europe.”

  22. CCHolley says:

    JCH, here are papers discussing the arctic ice variability as related to internal variability versus the anthropogenic influences:

    J J Day, J C Hargreaves, J D Annan and A Abe-Ouchi, “Sources of multi-decadal variability in Arctic sea ice extent” IOP Publishing Ltd. (2012)

    Per the author’s findings between 5% and 30% of the Arctic sea ice decline from 1979 to 2010 could be attributed to the natural cycles of the AMO and AO, and even less can be attributed to natural cycles since 1953, since these natural cycles tend to average out over longer time frame. “despite increased observational uncertainty in the pre-satellite era, the trend in [Arctic sea ice extent] over this longer period [1953–2010] is more likely to be representative of the anthropogenically forced component.”

    Julienne C. Stroeve, James Maslanik, Mark C. Serreze, Ignatius Rigor, Walter Meier, “Sea ice response to an extreme negative phase of the Arctic Oscillation during winter 2009/2010” Geophysical Research Letters (2011)

    The authors found for the period of 2009-2010, the AO was in a state which should have resulted in a large sea ice extent; the fact that 2010 was a year of relatively low sea ice extent is indicative long-term human-caused sea ice decline. “Based on relationships established in previous studies, the extreme negative phase of the Arctic Oscillation (AO) that characterized winter of 2009/2010 should have favored retention of Arctic sea ice through the 2010 summer melt season. The September 2010 sea ice extent nevertheless ended up as third lowest in the satellite record, behind 2007 and barely above 2008, reinforcing the long-term downward trend.”

    Dirk Notz, Jochem Marotzke, “Observations reveal external driver for Arctic sea-ice retreat”
    Geophysical Research Letters (2012)

    Authors found very poor correlation between the AMO and PDO with Arctic sea ice extent. They concluded: “the available observations are sufficient to virtually exclude internal variability and self-acceleration as an explanation for the observed long-term trend, clustering, and magnitude of recent sea-ice minima. Instead, the recent retreat is well described by the superposition of an externally forced linear trend and internal variability. For the externally forced trend, we find a physically plausible strong correlation only with increasing atmospheric CO2 concentration. Our results hence show that the observed evolution of Arctic sea-ice extent is consistent with the claim that virtually certainly the impact of an anthropogenic climate change is observable in Arctic sea ice already today.”

  23. Steven Mosher says:

    “NOAA/GISS start in 1880, I wonder why?

    1. Because their sources of choice start there
    2. because their method is incapable of using all the data

    People need to understand the insanity of their method.

    Suppose you have 10 trillion perfect measurements for 1850-1880.
    but those stations only last 30 years.

    Suppose you have 10 trillion measurements from 1980 to 2016.

    That would give you great accuracy in calculating delta T.. BUT

    GISS would have to dump all that data on the floor as would CRU.

    The reason is simple. They use anomalies and base periods.

    Who taught us that this approach is sub optimal? skeptics at climate audit.

  24. For what it is worth: I feel that the Berkeley Earth series is too long. In the beginning it is basically the temperature in Europe (and a little coastal USA) and at the end it is the global temperature. That is comparing apples and oranges. This is only partially communicated by the uncertainty in the mean; also the autocorrelations in the uncertainties are a lot stronger in the beginning.

    GHCNv3 may be shorter than CRUTEM because GHCN homogenize the data themselves and thus need a sufficient network density to have enough stations to compare with each other to see data problems. CRUTEM uses data that is homogenized by the national weather services if available and they have more data and metadata (station history information) to guide their homogenization. The approach of GHCN is more transparent. If there are data problems in Berkeley Earth in the beginning, they are likely only able to remove a small part of it.

    I am sure the people at climate audit are very happy with the Berkeley Earth dataset and have stopped their campaign against the instrumental record.

  25. cce says:

    I’ve always thought it was strange that model output is constantly being compared the temperature indexes, but model output isn’t comparable to the temperature indexes. The first I heard of this was in a Hansen paper (2007, I think) when he said that actual warming was somewhere between the “classic” Met stations only index, and modern SST+SAT GISTEMP. Made sense. But here’s the thing. Models output SST, right? I mean, they have to. Why did it take until 2016 to write this paper? Why hasn’t this been done since day one? What has been stopping modeling groups from creating hybrid output (SST+SAT) that is actually comparable to the indexes? Do they like being punished and ridiculed by “skeptics?” I don’t get it.

    In any case, produce the hybrid output. Mask so that the output matches real world observations. Then run the hybrid output through the various index algorithms. THEN do the comparison between observations and models.

  26. cce,
    I don’t know the answer to that. My suspicion is that people thought it wouldn’t make much difference, and if it weren’t for “skeptics” continually claiming models have failed, it probably been seen as an issue. It’s probably a combination of trying to counter the latter and that it does actually make some difference, that has lead to this.

  27. izen says:

    @- cce
    “Why did it take until 2016 to write this paper? Why hasn’t this been done since day one? ”

    I expect ATTP is right in thinking it was because nobody thought it made much difference until otters made an issue of it. After all those in the field knew that models and observations are reporting different metrics and therefore did not expect close agreement.

    I suspect another reason is precisely because it wasn’t done since day one.
    The early development of climate models that has evolved to full GCMs did not start with results that could be easily converted to be comparable to the various observational temperature data. Like qwertyuiop, the form of the output from the next generation of climate models was kept consistent with previous models to enable inter-comparison between model types and generations. That may have been seen as more important than changing the reported output to a different format.

  28. cce says:

    This is the quote from Hansen (2006, not 2007):

    Temperature change from climate models, including that re-
    ported in 1988 (12), usually refers to temperature of surface air over
    both land and ocean. Surface air temperature change in a warming
    climate is slightly larger than the SST change (4), especially in
    regions of sea ice. Therefore, the best temperature observation for
    comparison with climate models probably falls between the mete-
    orological station surface air analysis and the land–ocean temper-
    ature index.
    http://www.pnas.org/content/103/39/14288.full.pdf

    The difference between the two indexes is somewhat large — “large” in the sense that it represents a significant portion of expected warming.

    I don’t think it should have taken so long to produce an apples-to-apples comparison like this. But, then again, I didn’t have to do the work.

  29. Steven Mosher says:

    “For what it is worth: I feel that the Berkeley Earth series is too long. ”

    that’s like saying CET is too long

    “If there are data problems in Berkeley Earth in the beginning, they are likely only able to remove a small part of it.:”

    I look at it this way. Given the data what can we say and with how much confidence.

    it would be cool to get an update on this

    http://iedro.org/articles/south-american-data-rated-as-second-highest-priority-climate-data/

    “Meteorological observations such as temperature, pressure, and precipitation in South America date back nearly 500 years to 1534 AD, when Spaniards began keeping daily weather logs in present-day Ecuador. Today, as we try to understand how our climate will change in the near future, we look toward records like these to see how our climate has changed in the past. While a great deal of data from South America has already been collected and digitized by IEDRO and other entities, each country still has vast stores containing observations and historic data on old parchments that have not been touched.”

    It would be interesting to here your thoughts Victor on data rescue and the priority it gets

  30. Steven Mosher says:

    Victor..

    do you know

    Fernando Domínguez Castro?

    he has a fascinating list of publications

    http://link.springer.com/article/10.1007/s11629-015-3795-0

  31. Ron Graf says:

    Steven Mosher: “Its so funny.. when skeptics want to prove the MWP was warmer they love talking about a couple places as if they were the globe.
    But when the foot is on the other hand….
    [V Venema] ‘For what it is worth: I feel that the Berkeley Earth series is too long. ‘
    …that’s like saying CET is too long”

    The issues of mixing qualities is subjective. The issue brought up in Richardson et al of apples and apples, and the other issues here in this string, are addressable by mature bench-marking. I think that is what Victor is getting at. For example, splicing proxy data with thermometer data should have had been proscribed by protocol. It would have saved a lot of distracted time, resources and feelings.

    Anders: “…My suspicion is that people thought it [SST versus air temp] wouldn’t make much difference, and if it weren’t for “skeptics” continually claiming models have failed, it probably been seen as an issue… “

    Harry Twinotter: “Uncertainty is nobody’s friend. And underestimate is just as likely as an overestimate, despite what the climate change deniers wanted.”

    Actually, in this case everyone could assume that the SST, and every inch of depth, would dampen the signal. I would be floored if this had not been already addressed. Are there corrections in the surface historical record that Karl, Hansen, Easterling, etc..have missed? I kind of doubt it.

    Back to benchmarking, it should be clear that all the analysis in this post assume.
    1) Zero centennial and millennial variability (forced or unforced). Don’t be confused by those who mentions of the LIA and MWP in passing.

    2) There are no gradual land use chances like road building, land clearing, swamp draining or rural city growth that can are not documented and adjusted for or diagnosed by trend breaks from neighboring stations.

    3) Urban Heat Island only affects big cities and is adequately adjusted for.

    4) Micro-site issues like HVAC installation or airport growth are accounted for.

  32. Windchaser says:

    Back to benchmarking, it should be clear that all the analysis in this post assume.
    1) Zero centennial and millennial variability (forced or unforced).

    Hmm. No, that’s not a clear assumption. What’s your logic?

  33. Ron Graf says:

    “What’s your logic?”
    From Anders Post: The basic issue is that energy balance estimates for climate sensitivity use observed changes in temperature, and estimated changes in forcing, to determine the climate sensitivity.

    1) If any of the observed warming is from heating by a change in overturning or polar gradient (unforced heating), that would confound the equation since all the warming is assumed forced.

    2) If any of the observed warming is from heating by a change in volcanic frequency, solar variation or cosmic ray intensity, (which could affect cloud formation according to CERN), then any appropriation such warming to GHG would be erroneous. Climate sensitivity, by definition in this post is confined to GHG forcing,

    Does anyone know why in the world NOAA did not include an air temperature sample ability in Argo’s engineering specification? Was it just funding?

  34. Steven Mosher says:

    “2) If any of the observed warming is from heating by a change in volcanic frequency, solar variation or cosmic ray intensity, (which could affect cloud formation according to CERN), then any appropriation such warming to GHG would be erroneous. Climate sensitivity, by definition in this post is confined to GHG forcing,”

    wrong

    The equation is pretty simple

    ECS = 3.71 * Delta T / Delta F = Delta Q

    DeltaT/ Delta F+Delta Q is the climate sensitivity.. ECS is the sensitivity * the forcing due to
    doubling c02 or 3.71 watts

    Delta T is just The temperature in the final period ( say 1995-2015) – the temperature in the base period ( 1860-1880)

    Delta Q is just the change in OHC

    Delta F. includes the changes in all forcing. So for example Lewis picked the base period because volcanic forcing was close to zero then. Same for the current period
    0-0=0
    Solar forcing. Go check the delta in solar forcing. Since there is VIRTUALLY NO secular trend in solar forcing delta F for solar is Zero
    GCRs…. Unicorns… For GCR to me important there has to be
    1) an observed effect in ACTUAL DATA.
    2) That effect would have to be non zero between the base period and the final period
    Since you cant even FIND a change in clouds as GCRs change you are just talking
    about the possibiliyy of unicorns

  35. Steven Mosher says:

    1) If any of the observed warming is from heating by a change in overturning or polar gradient (unforced heating), that would confound the equation since all the warming is assumed forced.

    Or if Unicorns caused the warming…
    U are basically just making stuff up.. “heating by change in overturning? by polar gradiant?

    here is the thing.

    Delta T is what it is. IF you want to argue that some internal variation “caused” the delta T
    then you have to describe the mechanism.. in physical terms.

    For example, I’ll turn to Lewis and Curry again. They picked a base period that they tough would minimize the impact of things like AMO. So, you left with arguing that there is some unknown LTP
    that we havent named yet that was at a low point during 1860-1880 and at a high point now.

    What is that process? when does it start and stop? how does it express itself? is it like some super slow El nino? There is basically no evidence for a such a process. Of course we can hypothesize that it might exists.. and yes unicorns might exist as well. But If I told you unicorns existed you expect some kind of evidence for that…

    Hypothisizing new Entities ( like planet vulcan or dark matter ) can be done but you have to specify what to look for and how you know when you will find it.

    As it stands GHGs explain the difference in temperature. That means if you want to promote unicorns as an alternative explanation, you’ve got a lot of proving to do. and then you have to explain why it cant be GHGs..

  36. VV: “For what it is worth: I feel that the Berkeley Earth series is too long. ”

    Mosher: that’s like saying CET is too long

    That’s like saying Trump’s hands are too small.

    You can do better.

    VV: “If there are data problems in Berkeley Earth in the beginning, they are likely only able to remove a small part of it.:”

    Mosher: I look at it this way. Given the data what can we say and with how much confidence.

    No, your confidence intervals are for the sampling, not for the remaining trend bias after homogenization.

    Mosher: It would be interesting to here your thoughts Victor on data rescue and the priority it gets

    I do not work on data rescue myself, but know a little because meetings and sessions are often combined with homogenization and quality control.

    My impression is that data rescue is mainly (but certainly not only) a problem for daily data, at least for rich countries. For developing countries it is hard to know what is there and thus whether data rescue is needed. The number of monthly averages we have is often used to estimate how much daily data is still missing. It has much too little priority. At the current rate we will need many decades until most historical paper data is available for research. It is expensive, the records are typically hand written and machine learning is still not sufficiently accurate to digitise hand written documents.

    The main problem for monthly averages, which are in the focus of the climate “debate”, (and more so for daily data) is that the data is often not shared internationally and that also information on inhomogeneities, which is often in the local language, is not shared. Climate data has strategic and commercial value and most weather services are understaffed and do not have much time for climate work. I would hope that we can get the international sharing of climate data into the climate treaty. That would likely have to include compensations for the loses in sales of data.

    Going back in the past of the instrumental records before 1850 likely does not help the study of climate change much. Measurement methods were very different then, thus the measurements are hard to compare with current ones. That is also why you often find articles on one or a few such old series (like Fernando Domínguez-Castro ones) because you need a lot of background to help interpreting the observations. If you do not have multiple series so that you can homogenize the data, it will be hard to assess trends from such very old data.

    Such very old meteorological data are still a gold mine for historians and frequently help to understand historical events. They also often use proxy data or other documentary evidence (e.g., grape and grain harvests). There is a lot of beautiful work being done in climate history or environmental history.

  37. This bit of Judith Curry’s reflections jumped out at me:

    The key issue that this paper raises is about the quality of the global surface temperature data sets, particularly related to sampling. The paper infers that the trend of surface temperature anomalies in HadCRUT4 are 9-40% lower than the true global averages. Is this believable? Well the uncertainty in surface temperature trends (9-40%) doesn’t seem implausible, but the inference that the uncertainty is only on the side of insufficient warming doesn’t seem plausible. There are many uncertainties in these global surface temperature data sets, although sampling is probably the largest uncertainty. Where does the (9-40%) numbers come from? Climate models.

    Right, Judy. So you can’t be certain that the actual trend is NOT 9-40% higher than the central estimate. Oh, but *climate* models.

    Doesn’t Richardson et al. Figure 3 contain central estimates with two-sided error bars? Don’t those intervals *include* Lewis & Curry’s (and Otto’s) original TCR central estimates, not to mention more than a good portion of the total range of same? I must be imagining things again.

    Next paragraph:

    If you want to sort out the sampling uncertainty of a dataset like HADCRUT4, the best way to approach this is compare with the global reanalysis datasets like ECMWF and the NCEP reanalysis (both of which disagree significantly with each other). By subsampling the ECMWF and NCEP reanalyses, and then comparing with the global reanalyses, you could get a handle on the biases in HadCRUT4 from under sampling (and the uncertainty in this bias). This is a much better approach than trying to infer the bias in a trend using global climate models that are biased warm.

    Not that we’d ever beg the question. Nope. No siree. I really like how doing it the “best way” is someone else’s job.

  38. Ron Graf says:

    SM: “Or if Unicorns caused the warming…
    U are basically just making stuff up.. “heating by change in overturning? by polar gradient?”

    I’m not making anything up Steven. You are wrong.

    The ocean rate of overturning produces a non-forced steady-state surface temperature. Are you disputing this? Are you disputing that the rate of overturning is variable on lower frequency scales than ENSO? Are you disputing that GCMs are without skill at these frequencies?

    The polar gradient is also a non-forced steady state. It’s just like overturning except it works longitudinally rather and vertically, In addition to the direct surface temperature effect, the polar gradient has a small forcing (plank effect) if it decreases the overall temperature gradient since by plank effects alone this raises GMST.
    Temp K ———- Black body Radiant Emittance: W/m-2
    295 ——————— 429.5
    300 ——————– 459.3
    305 ———————490.7
    ((490.7+429)*.5-(459.3) = 0.80 W/m-2 for 10K (0.33%) overall mixing or global temp

    Steven, where did I dispute that aerosols and all other observed factors are not used in the observed EBM method for gaining CS? What I said is that all the non-GHG forcings, feedbacks and non-forced variability is assumed to be constant on the centennial and millennial scale. Thus, the only variable is AGW from GHG.

    Actually, according to Hansen 2001, the GISS temperature record at least contains intentionally unadjusted land use and land cover change LULCC, which is not AGW and not from GHG.

    If you do not dispute heart of my comment points then your comment was extremely misleading to readers. BTW, climate sensitivity per se does not necessarily include the OHC; as in transient climate response TCR, for example. You went out of your way to complicate and obfuscate it seems.

  39. Steven Mosher says:

    “Mosher: I look at it this way. Given the data what can we say and with how much confidence.

    No, your confidence intervals are for the sampling, not for the remaining trend bias after homogenization.”

    yes are confidence intervals are for sampling and measurement error, ie the data. If folks want to estimate trends I would tell them to think twice about what models they want to apply to data. and what time periods they want to use.. So given the data we say what we can say about the the sampling and measurement error.. since there is not much one can say about the trend bias from say 1750 to 1850.. we dont say what we cant say. maybe I am missing your point.. Does CET publish a remaining trend bias after their homogenization?

  40. Marco says:

    “land use and land cover change LULCC, which is not AGW”

    A lot of it *is* AGW. If humans convert forest to something else (agricultural land, cities, whatever), this is an *anthropogenic* forcing.

  41. Dikran Marsupial says:

    Victor wrote “For what it is worth: I feel that the Berkeley Earth series is too long. In the beginning it is basically the temperature in Europe (and a little coastal USA) and at the end it is the global temperature. That is comparing apples and oranges. [emphasis mine]

    Steven Mosher wrote
    “For what it is worth: I feel that the Berkeley Earth series is too long. ”

    that’s like saying CET is too long

    If the spatial coverage of the CET changed over time, Steven might have a point there, but as it is Steven seems to have missed Victor’s point (which seems reasonable to me).

  42. Dikran Marsupial says:

    Steven Mosher wrote “Who taught us that this approach is sub optimal? skeptics at climate audit.”

    I don’t know how true this is, but if so it provides an example of how not to communicate science in a way that will gain traction, i.e. publishing sober, measured analyses that demonstrate the case in journal papers is probably better than antagonism. Scientists should judge arguments on their merits, yes, but they are also human beings, so if you think someone is wrong, some methods are better than others in getting them to change their minds (if that is actually what you want to achieve). Sadly blog science currently doesn’t seem to encourage this sort of cooperation, but that doesn’t mean it can’t be done.

    Of course CA doesn’t always get it right, e.g. the Douglass et al paper on tropical trends.

  43. Dikran Marsupial says:

    I should add that (of course), not that I am perfect in communicating my ideas 100% of the time… ;o)

  44. Ron,
    The ocean heat content is increasing. How, therefore, can surface warming be internal?

  45. BBD says:

    The ocean heat content is increasing. How, therefore, can surface warming be internal?

  46. izen says:

    @-Ron Graf
    “In addition to the direct surface temperature effect, the polar gradient has a small forcing (plank effect) if it decreases the overall temperature gradient since by plank effects alone this raises GMST.
    Temp K ———- Black body Radiant Emittance: W/m-2
    295 ——————— 429.5
    300 ——————– 459.3
    305 ———————490.7
    ((490.7+429)*.5-(459.3) = 0.80 W/m-2 for 10K (0.33%) overall mixing or global temp”

    As your figures show if you flatten the polar gradient by heating the poles you need lots of energy to maintain that reduced temperature difference. The change in polar gradient does not magically generate that energy.

    You can reduce the polar gradient with a fixed energy input by cooling the equator in balance with warming the poles. A more efficient heat engine transporting that energy pole-ward would have this effect. However the colder tropics would be the result.

    The only way you can warm the poles more than the equator, with BOTH warming is to retain more energy in the surface climate system. Variations in the water-phase heat engine that moves energy from tropics to poles, and day to night sides, can only alter the distribution of that energy.

    I have tried to think of a natural, unforced process that can increase the amount of energy at the surface, rather than just move what there is around. Perhaps you could give a clear example?

  47. Steven Mosher says:

    “If the spatial coverage of the CET changed over time, Steven might have a point there, but as it is Steven seems to have missed Victor’s point (which seems reasonable to me).”

    The problem is that people are unclear when they refer to the “series”

    The data is the data. If you want a long series that spatially doesnt change from 1753 to 2016
    YOU CAN GO GET THAT.. its called masking. just do it.

    So I really dont see what victor means by “the series” begin too long.

  48. Steven Mosher says:

    or its like saying CRU is too long because in 1850 it only covers Europe and the coastal US.

    Anyone ever say that?

    The comment “it only covers europe and the eastern US” ( with respect to berkeley ) was made long ago.. by a reviewer.

  49. Steve Mosher: “ So given the data we say what we can say about the the sampling and measurement error.. since there is not much one can say about the trend bias from say 1750 to 1850..

    Then we agree. And maybe one should tell users that this period should thus not be used to study climate change. For that you need to have an idea about the errors in the observed changes. It could still be a nice dataset for meteorological questions in this period.

    Steve Mosher: “Does CET publish a remaining trend bias after their homogenization?

    Read the paper. I work on relative homogenization and they did not use that. They used a lot of information on how the measurements were performed where. Sometimes used parallel measurements. No idea whether they assessed the trend error.

    Steve Mosher: “or its like saying CRU is too long because in 1850 it only covers Europe and the coastal US. Anyone ever say that? The comment “it only covers europe and the eastern US” ( with respect to berkeley ) was made long ago.. by a reviewer.

    It wasn’t me. 🙂 Could be an older person, not too long ago this was still true.

    I also said something subtly different: “In the beginning it is basically the temperature in Europe (and a little coastal USA) and at the end it is the global temperature. That is comparing apples and oranges.”

    I did not specify a period for this statement.

    I did specify a period for the unreliable part of Berkeley Earth. The data before 1850 is unreliable not because you do not have the occasional station outside of Europe, but because you do not have sufficient station density outside of Europe to remove trend biases. You are basically looking at raw data for this period. Biases which are enormous in this period, where many different measurement methods and protocols were used.

    GHCN starting in 1880 is already very ambitious. I would love to see an error analysis focussing on the station density we had around that time.

  50. Christian John says:

    On the Topic,

    I came on my own analyis very well to Richardson et al. but i used another approach and it is only a short contribution. (without Uncertainy Analysis)

    So just look here: (it would be usefull if you understand german) : http://www.wzforum.de/forum2/read.php?6,3193835

    Well what i did? Simple to say, i use Hadcrut4 as Cowtan and Way, eliminate El-Nino-Effects, choose carefully the basline and compare it then with an Adjust RCP4.5 Output. So how do i adjust RCP4.5, well so KMNI-Explorer gives your 2 usefully variables, TAS and TOS (near Surface Temperature and SST only), there i masked RCP4.5-mean(TAS) in Land and Ocean and look for the weigthing between Land and Ocean.

    So i found:
    Ozean: 0.741
    Land: 0.259

    That allowed me the rebuild on Land*weigth + Ocean * weigth the orginal RCP4.5-Output in TAS with an Error of 2.75*10-6K. The next Step to get the SST by TOS for RCP4.5 and then used them with the orginal weigth + the near Surface Temperatur on Land with its weigth and build a RCP4.5-Run in a comparable format to the observations.

    As you can see in the link before, its works out, that RCP4.5 is well with the observations.

  51. MarkR says:

    Hi Ethan Allen,
    The code produces all figures and comments explain which bit produces which figure. There’s an error in README_data that you’ve found, sorry about that. The last column has been deleted from the final version of the hist_rcp85 data since anyone can do that calculation. The description should read:

    “Contains folders Txax and Had4, where Txax is global and Had4 follows HadCRUT4 historical masking. Each file contains 3 columns: time (years), air temperature anomaly (K), blended anomaly (K).”

    The time in the 1pctCO2 scenarios is years and the starting year is irrelevant. The README_data issue does not affect the results as the correct columns are loaded in each case,

    I’ll wait a bit for any other issues to be raised and then do a version update, thanks for finding this.

  52. MarkR says:

    Hi cce,
    “But here’s the thing. Models output SST, right? I mean, they have to. Why did it take until 2016 to write this paper? Why hasn’t this been done since day one?”

    I have some speculations.

    Imagine you’re calculating TCR in models. You check both air and surface temperatures from CMIP5 and it only makes ~3 % difference (roughly). Sensibly you move on, there’s lots of work to do. Kevin Cowtan didn’t though, he was stubborn and found that the way HadCRUT4 handles sea ice hides warming when sea ice is retreating, his paper explains why (doi: 10.1002/2015GL064888 ).

    Secondly, the models behave differently over land and ocean. Most models seem to have faster land surface warming compared with the air above (at least in abrupt4xCO2 simulations), so this dilutes some of the extra SST warming effect when you do your global average air vs surface temperature comparison.

    Combined, turns a ~9% blending effect into a 2-3% effect in the most sensible first test that people would do. And I’ve spoken to modellers about this who did this test and said the effect was small.

    Plus the masking effect on top, which people knew about and sometimes include in other work but didn’t necessarily realise how important it is for this particular calculation.

    Finally, Ben Santer and others did the blending and masking in a 2000 paper (doi: 10.1126/science.287.5456.1227 ) but didn’t make a big deal out of it and it didn’t affect their results very much.

  53. MarkR says:

    brandonrgates: “Right, Judy. So you can’t be certain that the actual trend is NOT 9-40% higher than the central estimate. Oh, but *climate* models.”

    Ignoring the masking effect, we have solid evidence for the 9% blending bias being positive.

    1) The sea-ice component is identified by Kevin Cowtan, makes logical sense, and agrees with BEST’s comparison of using air vs SST data over sea ice.

    2) The faster-warming-air comes from robust changes in the surface energy balance. Textbook physics gives the same answer as climate models on this.

    We’ve got good reasons to believe both of them and they are both biases that mean reported temperature is less than reported by HadCRUT4. I don’t think anyone should give much credit to Curry’s opinion on this until she comes up with evidence that the BEST data and textbook physics are wrong.

    The masking effect might be opposite in the real world to that seen in models, but the Arctic dominates so I’d like to see some evidence that the Arctic, where all evidence points to faster warming than the global average, or other missing regions are not warming as expected. Although from what Curry says, I think she’s implying we should ignore the fastest warming area on Earth because internal variability is big there?

    I see that Curry also said:
    “Well, the biggest outstanding issue is that of aerosol forcing; if you use Bjorn Stevens’ aerosol forcing values, Nic Lewis (from table above) finds the (0-95%) range to be 0.93 – 1.67 C, whose upper limit is below the mean of Richardson’s subsampled CMIP5 sensitivities.”

    The CMIP5 ensemble mean for blended-masked simulations is about 1.5 C (1.49 C iirc), and median is smaller, around 1.4 C.

  54. MarkR says:

    brandonrgates: that post wasn’t necessarily aimed at you, I was using your comment as a springboard.

  55. Ron Graf says:

    Anders: “Ron,
    The ocean heat content is increasing. How, therefore, can surface warming be internal?”

    Yes. The OHC should be able to flatten out the ENSO surface trends into a trend that accurately reflects the direction of decadal trend in radiative forcing. The difficulty is that we have only been able to observe OHC accurately (via Argo) for a single decade. So, although the OHC trend is the best gauge for measuring mean decadal radiative imbalance, this reveals little without being able to compare past decade’s or century’s imbalance and mix of forcings.

  56. Ron Graf says:

    Izen: “As your figures show if you flatten the polar gradient by heating the poles you need lots of energy to maintain that reduced temperature difference. The change in polar gradient does not magically generate that energy.”

    I am not implying that the gradient generates energy merely that it radiates more efficiently the less mixed the TOA temp is, notwithstanding entropic energy, which I have not calculated.

    “You can reduce the polar gradient with a fixed energy input by cooling the equator in balance with warming the poles. A more efficient heat engine transporting that energy pole-ward would have this effect. However the colder tropics would be the result.”

    1) A more efficient longitudinal heat flux would come at the expense of energy at the tropics but not necessarily reducing the surface temp. The tropics seems to have a surface feedback of surface evaporation cooling whereby the latent WV heat is transported to the upper troposphere, condenses and warms the TOA. Thus there is little “tropical hotspot” during global warming and no tropical cool spot in global cooling.

    2) Increasing the efficiency of the flux through the polar gradient decreases outgoing longwave radiation (OLR) thus warming the globe generally, and in the tropics making up for some of the loss.

    “The only way you can warm the poles more than the equator, with BOTH warming is to retain more energy in the surface climate system. Variations in the water-phase heat engine that moves energy from tropics to poles, and day to night sides, can only alter the distribution of that energy.I have tried to think of a natural, unforced process that can increase the amount of energy at the surface, rather than just move what there is around. Perhaps you could give a clear example?”

    The best example may be the onset of the Quaternary Ice Age correlating with the merging of the North American and South American continents and thus creating the Atlantic and Pacific as separately circulating bodies, perhaps less efficiently overall. Also, perhaps the cooling effect of a slowdown of the AMOC is enhanced by the change in polar gradient efficiency.

  57. Ron Graf says:

    From Anders’ Post: “The results tend to suggest values lower than suggested by climate models. However, the temperature datasets used have some coverage bias (not all areas have coverage) and the temperature measurements vary from being air temperatures over land to sea surface temperatures over the oceans.”

    I believe the divergence over time between the mean SST and mean air temp over the sea was deemed at some point to be insignificant compared the cost of adding yet another adjustment. After all, the difference in trends of SST v SAT anomaly are likely small. Also, apparently SAT data exists as Karl et al (2015) used it.

    …one of the
    improvements to ERSST version 4 is extending the ship-bias
    correction to the present, based on information derived
    from comparisons with night marine air temperatures…

  58. Windchaser says:

    The best example may be the onset of the Quaternary Ice Age correlating with the merging of the North American and South American continents and thus creating the Atlantic and Pacific as separately circulating bodies,.

    Yeah, fundamental changes to things like the topography of continents, or the composition of the atmosphere would constitute what we call a “forced” change.

    There’s a problem here in that the skeptics don’t have much more than speculation. They don’t have solid data behind large unforced natural variability, they don’t have mechanisms. They just have a little bit of curve-fitting and speculation.

    It’s a crazy idea to hang your hat on. I mean, by all means speculate, but keep a critical mind and realize that, without evidence, it’s still just speculation. Don’t get too attached to your ideas, either, until there’s data, or hell, at least a mechanism to back it up.

  59. Ron Graf says:

    “It’s a crazy idea to hang your hat on. I mean, by all means speculate, but keep a critical mind and realize that, without evidence, it’s still just speculation. Don’t get too attached to your ideas, either, until there’s data, or hell, at least a mechanism to back it up.”

    “Uncertainty is nobody’s friend.”

    The starting point is critical thinking, then analysis, then investigation. Rinse and repeat.

  60. Marco says:

    “The difficulty is that we have only been able to observe OHC accurately (via Argo) for a single decade. So, although the OHC trend is the best gauge for measuring mean decadal radiative imbalance, this reveals little without being able to compare past decade’s or century’s imbalance and mix of forcings.”

    And then there’s the sea level rise…which has some other components also, but it would be really hard to come up with an explanation for that sea level rise that does not include a significant steric component.

    Ron, be careful not to mistake “critical thinking” with “confirmation bias”. I often tell my undergrads to do a “sanity check” – if your calculations show that the temperature for a reaction has to be -10 K, you know you did something wrong – and they can expect to get things that largely fit with theory. I tell my PhD students to not only do a “sanity check” but also a “sanity check” on that “sanity check” (and/or that in the literature). If you are critical of a result because it does not fit your bias, you can easily do an analysis that is set up to confirm your bias. Critical thinking includes challenging your own ideas.

  61. Dikran Marsupial says:

    “The data is the data. If you want a long series that spatially doesn’t change from 1753 to 2016
    YOU CAN GO GET THAT.. its called masking. just do it.”

    I don’t think many of the people discussing climate on blogs are actually capable of doing that. I think Victors comment was reasonable, personally rather than masking or truncating the data I would give a caveat about the spatial change (rather in the same way caveats ought to be added in talking about the medieval warm period). Basically in public communication of science (which is hopefully what climate blogs are doing) we can’t take it for granted that the reader will know about these things (indeed I didn’t when I started reading climate blogs), so from that perspective it would be reasonable to suggest that the series may be “too long”.

    “or its like saying CRU is too long because in 1850 it only covers Europe and the coastal US.”

    That would be a better comparison.

  62. BBD says:

    Ron G

    Yes. The OHC should be able to flatten out the ENSO surface trends into a trend that accurately reflects the direction of decadal trend in radiative forcing. The difficulty is that we have only been able to observe OHC accurately (via Argo) for a single decade. So, although the OHC trend is the best gauge for measuring mean decadal radiative imbalance, this reveals little without being able to compare past decade’s or century’s imbalance and mix of forcings.

    While ARGO has improved the accuracy of OHC measurement, the pre-ARGO data are more than adequate to demonstrate a multidecadal *increase* in OHC in all major basins.

    How can OHC be increasing in all major basins if surface warming is internally forced? OHC should be *falling* if there were a net transfer of energy from the oceans (the only possible reservoir) to the atmosphere.

    Please actually answer the question this time.

  63. MarkR says:

    BBD:

    Maybe some natural processes could both increase OHC and cause warming. Maybe some internal ocean circulation change that brings warmer water to the surface at the edges of the marine stratocumulus regions, reducing the stratocumulus area. Or changes in circulation that affects cirrus cast off from deep convection near the equator.

    I wouldn’t be surprised that _if these things happened_ you could eventually see a net increase in both OHC and warming without any forced components. Then you need to explain why this “natural cycle” is happening over an ever-increasing period of time, why no-one has found it, where the measured anthropogenic heating has disappeared to and explain why climate in the past seems to have changed as expected but is no longer doing so. To paraphrase John Cook from a while ago: “it walks like a duck, swims like a duck, flies like a duck and quacks like a duck, but you expect us to believe that there is no duck and the walking, swimming, flying and quacking is being done by a unicorn.”

    It’s getting closer and closer to one of Russell’s teapots.

  64. BBD says:

    MarkR

    Just for fun, I’m not even sure if the hypothetical unicorn you sketched out could take on substance and quack:

    Maybe some natural processes could both increase OHC and cause warming. Maybe some internal ocean circulation change that brings warmer water to the surface at the edges of the marine stratocumulus regions, reducing the stratocumulus area. Or changes in circulation that affects cirrus cast off from deep convection near the equator.

    Wouldn’t the reduction in cloud area tend to balance the energy fluxes to and from the surface? Reduced cloud = increased TSI but also increased IR flux from the surface to TOA. Plug in the diurnal cycle and what you gain during the day you lose during the night. I think.

  65. MarkR says:

    BBD, I think both of those should have a net positive effect. Mauritsen and Stevens (doi: 10.1038/ngeo2414) for the cirrus “iris” and Brient and Schneider (doi: 10.1175/JCLI-D-15-0897.1) for low-cloud. But like you say, sunlight and infrared effects oppose each other which weakens the effect.

    So in our little game, let’s say you insist that sensitivity is low and you call on something like the iris to explain it, then why is there an iris hiding so much human-caused heating (we know CO2’s direct heating is almost exactly as predicted, Feldman et al., doi: 10.1038/nature14240), that is super-powerful but disagrees with our radiative physics and hasn’t been found by measurements, but there’s also an “anti iris” at the same time that only works for natural variability, just happens to be correlated excellently with our radiative forcing over about half a century and hasn’t been found by measurements?

    Another step closer to Russell’s teapot.

  66. John Hartz says:

    Speaking of ocean warming…

    When it comes to fundamental drivers of climate and weather across the Earth, it is hard to think of a region more important than the Indo-Pacific Warm Pool, an enormous area stretching across the Pacific and Indian oceans on both sides of the equator.

    This is, basically, the biggest body of warm water there is. Indeed, the warm pool, which is fueled by the intense sunlight striking the equator and tropics, is defined as the area where the average surface ocean temperature is greater than about 82 degrees Fahrenheit all year round (a temperature, incidentally, that is well above the threshold level needed for tropical cyclone or hurricane formation).

    The warm pool drives monsoons, tropical cyclones and much more. Its warm ocean surface is the home to deep atmospheric “convection,” or the rising of warm, moist air, which leads to atmospheric circulation and rainfall patterns that influence the entire planet.

    And the warm pool is growing.

    The biggest body of warm water on Earth is getting even bigger by Chris Mooney, Energy & Environment, Washington Post, July 1, 2016

  67. BBD says:

    I concede 🙂 I just can’t do it. Contrarianism requires bigger brains than mine.

  68. Ron Graf says:

    BBD, MarkR, it is humorous to watch you both having a John Cook type debate with yourselves.

    BDD says: How can OHC be increasing in all major basins if surface warming is internally forced? OHC should be *falling* if there were a net transfer of energy from the oceans (the only possible reservoir) to the atmosphere.

    1) I never said there was zero AGW. There should be warming; thank goodness there’s warming. Mankind’s inadvertent tripping over CO2 before the interglacial ran into full stop was just one in a long line of cosmic lucky breaks. It certainly saved us from all flocking to the tropics right now in a global re-glaciation. I am hoping we can switch to safe nuclear quickly and save some of the CO2 for 100 generations from now — they will need to keep the atmosphere >350ppm for at least 15k more years to stave of the freezemeister.

    2) The ocean layers from 0-2000m are warming but the deep ocean is not warming*, and NASA does not have a clue why. I guess physics isn’t so simple as some might guess. Also, we have no idea if the OHC has been rising since 1900 or 1700. The later would seem more likely since all the proxy reconstructions show the ocean surface cooled a very significant 0.7C from 1250 to 1700.

    3) If we really want to link OHC to radiative influence we should be able to correlate La Nina with OHC rise and El Nino with OHC fall since a warmer surface is radiating the OLR more effectively thus expelling the heat outward rather than inward. Does anyone know if that has been done?

    *Llovel (2014) news release says:

    The deep parts of the ocean are harder to measure,” said JPL’s William Llovel, lead author of the study published Sunday in the journal Nature Climate Change. “The combination of satellite and direct temperature data gives us a glimpse of how much sea level rise is due to deep warming. The answer is — not much.

  69. BBD,

    Contrarianism requires bigger brains than mine.

    I think of it as artistic license.

  70. MarkR,

    that post wasn’t necessarily aimed at you, I was using your comment as a springboard.

    Happy to have given an assist.

  71. JCH says:

    3) If we really want to link OHC to radiative influence we should be able to correlate La Nina with OHC rise and El Nino with OHC fall since a warmer surface is radiating the OLR more effectively thus expelling the heat outward rather than inward. Does anyone know if that has been done?

    I see you have applied critical thinking.

    The guys over at moyhu checked about a month ago, with El Nino quickly winding down, and there was little sign of a drop in OHC.

    How can significant energy leave the oceans if there is a persistent TOA imbalance where more energy is going into the earth system, some 93% of it into the oceans, than is leaving?

  72. MarkR says:

    John Hartz: we recently published a paper looking at a related topic in models (Stephens, Kahn & Richardson, doi: 10.1175/JCLI-D-15-0234.1 ).

    A warming surface means more evaporation. In the warm pool, convection carries the moisture high into the air where it makes the greenhouse effect much stronger, reducing heat escape to space. Warming alone increases heat loss, but in some areas the moisture effect is stronger. We have a “super-greenhouse effect” and these regions in models have a “runaway greenhouse effect”. As the surface warms in those regions, the heat escaping to space decreases, which encourages them to warm further.

    This is possibly what happened globally on Venus, causing it to warm up so much that nowadays its surface is hot enough to melt lead, although the water since escaped into space (molecular mass 18, easier for it to escape than chubby CO2 with a molecular mass of 44).

    On Earth it’s not global though, so heat can flow to higher latitudes which act as “cooling fins”, e.g. the cloudless but hot Sahara desert is like a big radiator pointed into space.

    We found that in models, this super-greenhouse region expands as warming goes on. This is a testable prediction, and it could have important effects for global weather patterns.

  73. BBD says:

    Ron G

    There should be warming; thank goodness there’s warming. Mankind’s inadvertent tripping over CO2 before the interglacial

    One factory producing CFCs could have averted a glaciation. There is no need for rapid, large and sustained CO2 forcing with all its attendant risks and hazards.

    The ocean layers from 0-2000m are warming but the deep ocean is not warming*, and NASA does not have a clue why. I guess physics isn’t so simple as some might guess.

    The abyssal ocean takes millennia to warm and no, this is *not* a surprise to climatologists and oceanographers nor an indication that the physics is poorly understood. The 0 – 2000m layer is the relevant slice when it comes to AGW, at least on centennial scales.

    This is now the third time you have dodged the question:

    How can OHC be increasing in all major basins if surface warming is internally forced? OHC should be *falling* if there were a net transfer of energy from the oceans to the atmosphere.

    Please answer this time.

  74. JCH says:

    I believe what Wunsch actually said is it cannot currently be determined whether or not the abyssal ocean is net warming or net cooling… some areas clearly warming and some clearly cooling.

  75. JCH says:

    Trenberth says OHC went down as a result of the 97-98 El Nino, and he says it went down after the 09-10 El Nino, though I’ve never seen a graph where 09-10 dip is evident.

    But in general, it looks to me like OHC goes up, on net, during all phases of ENSO, which I believe is consistent with Minnett and other descriptions of the greenhouse theory.

    But on the internet you can read dozens of descriptions of El Nino disgorging massive amounts of stored energy from the last La Nina into the atmosphere.

  76. Ron Graf says:

    JCH, I found the string on Moyhu thread regarding OHC: https://moyhu.blogspot.com/2016/05/surface-templs-global-temperature-down.html

    It seems that the El Nino warming of the 0-100m layer has been related to cooling of the 100-700m layer. That makes sense. What is hard to see, within error, is the ENSO effect on the entire OHC. in order to get a controlled analysis of the entire system (the hydrosphere) there needs to be accurate measurements of all the atmospheric forcings, including radiative feedbacks like cloud cover, accurate surface, 0-2000m ocean and deep ocean. Only in this way can true energy balance closure be obtained. It seems that surface warming should increase OLR but even that should be confirmed since as some point out the increased water vapor may just trap the increased OLR, resulting in little radiative adjustment. I believe this is a big fear by some and should be verified or invalidated.

    BBD: The abyssal ocean takes millennia to warm and no, this is *not* a surprise to climatologists and oceanographers nor an indication that the physics is poorly understood.

    According to Llovel they were surprised at the zero change in the OHC in the deep ocean (>2000m). I believe that your assumption: the abyss is immune from overturning at a less than 1000-yr rate, is one without much evidence. Since the deep ocean contains on the order of the same heat content as the 0-2000m a chnage in overtuning rate could have a major impact on the 0-2000m OHC. The reason we should be using the ENSO – 0-2000m OHC analysis is to verify our measurements are correct and verify the radiative flux in OLR, which can help verify the degree of imbalance at given surface conditions. A major volcanic eruption would grant a superb opportunity to test predictions on all of the system values for a given transient drop in radiative forcing. All of these observational analyses could be used for model validation or falsification. Of course this still would not answer if BBD’s assumption about a stable steady-state of the abyss, immune from millennial disruption in either the vertical and polar gradient dynamics, which could could very well explain the LIA a MWP, etc…

  77. BBD says:

    Of course this still would not answer if BBD’s assumption about a stable steady-state of the abyss,

    It’s not *my* assumption and it is supported by Llovel et al. which is the source you introduced into the discussion.

    You still have not answered the question:

    How can OHC be increasing in all major basins if surface warming is internally forced? OHC should be *falling* if there were a net transfer of energy from the oceans to the atmosphere.

    Come on.

  78. BBD says:

    Ron G

    According to Llovel they were surprised at the zero change in the OHC in the deep ocean (>2000m).

    Really? Where does it say that then? In the paper? In the press release? Can’t find it anywhere. I thing you just made it up, along with your other stuff about the physics etc.

    Let’s look at the press release for Llovel et al. which you did not link to and have misrepresented. It says:

    Scientists at NASA’s Jet Propulsion Laboratory (JPL) in Pasadena, California, analyzed satellite and direct ocean temperature data from 2005 to 2013 and found the ocean abyss below 1.24 miles (1,995 meters) has not warmed measurably. Study coauthor Josh Willis of JPL said these findings do not throw suspicion on climate change itself.

    And:

    The remainder was essentially zero. Deep ocean warming contributed virtually nothing to sea level rise during this period.

    Coauthor Felix Landerer of JPL noted that during the same period warming in the top half of the ocean continued unabated, an unequivocal sign that our planet is heating up. Some recent studies reporting deep-ocean warming were, in fact, referring to the warming in the upper half of the ocean but below the topmost layer, which ends about 0.4 mile (700 meters) down.

    Landerer also is a coauthor of another paper in the same journal issue on 1970-2005 ocean warming in the Southern Hemisphere. Before Argo floats were deployed, temperature measurements in the Southern Ocean were spotty, at best. Using satellite measurements and climate simulations of sea level changes around the world, the new study found the global ocean absorbed far more heat in those 35 years than previously thought — a whopping 24 to 58 percent more than early estimates.

    Not once does it say that anyone was “surprised”. Stop the bullshitting, please.

  79. JCH says:

    Sea -haha – any OHC exiting the numerous El Nino events here? What would it look like?

  80. Christian says:

    JHC,

    You overlook that the upper ocean is more sensitiv in heat as the lower ocean (the upper is expanding more by equal energy increase as the lower do) and also, that El-Nino change the Rain-Pattern (often there is more rain above the ocean itself on el-Nino as in normal state or la nina)

    As i posted on myhu, you have to watch the differences betwwen the upper or mixed layer and the lower ocean. So therefor it could be, that you loos a little bit energy integrated over all depts, but also get a increase in SSH.

  81. JCH says:

    Hi Christian – I would say a little bit of energy is a lot different than the predominant internet belief that an El Nino warms the atmosphere, 98-style, by jettisoning a vast amount of OHC into the atmosphere that was stored in the oceans by a previous La Nina. It is very hard to get the extra, sequestered energy out of the oceans once it gets in there… because the enhanced greenhouse effect is progressively make its escape from its ocean prison ever incrementally more difficult. An El Nino atmosphere is primarily warmed by what takes place in the skin layer, and the vast majority of the energy in the skin layer is likely to be current sunlight and long wave radiation.

  82. Christian says:

    JHC,

    Now i see the point, yes the “internet belief” has nothing to do with reality. On El-Nino you dosent loos very much energy from the ocean, this is because, the normal state in eqatorial pacific is upwelling of cooler water because of trade wind, also because of trade winds + coriolis force, there is some transport from the eqatorial pacific away.

    So if there is El-Nino, you have less trade winds and less upwelling and this alone is cause warming of SST, you need not anythink else to warm up the SST. Another Part is below the surface, the OHC increase due Kelvin-Waves, which is more pushing down the thercline then a real transport together. Also this all taken together, you get a interesting Feedback (i just only want to explain the feedback on El-Nino creation, there is also a feedback in El-Nino destruction from the Rosby-Waves but not the point here) that makes things following up:

    There is a strong West-Wind-Burst ———————–> cause a Kelvinwave with Downwelling ———–> Downwelling propagates form W to E in tropical Pacific ——————–> on the propagation it pushes down the thermocline ———————–> since then Upwelled Water by Trade-Winds get warmer(but always cooler as its on Surface) ———————> less cool Upwelling water weakens the Trade-Winds —————-> weaker Tradewinds cause less Upwelling Volume ————–> generell SST increase ——————-> more frequent of West-Wind-Bursts ——————– (See the Beginning)

    So you can see well, this isnt really need any energy which was stored before. The only loose is of transport of some energy due middle latitudes where of it goes out to space.

    So whats really lower a little bit the heat content of the ocean? Mhm, you have to look at the climate system as full, since tropical basin is more are heat sink (because of Uwelling) and arctic/antarctic are more heat realease and of water formation brings cool water in the lower ocean in direction to the equator there should be a small drop in OHC above all depts, because if you weaken your heat sink, but not your water formation on equal mass, its simply getting colder in the lower ocean.

    So if you now goes to the radiative imbalance cause by GHGs, i would say, it cause simply, that El-Nino cant cancelt out by La-Nina, so therefor, on La-Nina-Years we should see the a constant warming (every La-Nina warmer then La-Nina before) with the exception of vulcanic erruptions on the same time.

    Greets

  83. JCH,

    Sea -haha – any OHC exiting the numerous El Nino events here? What would it look like?

    Mebbe like this?

    OHC are the 700m pentdatal data from NODC. Annual rate of change was calculated by simply subtracting the former year’s value from the current year. ENSO is represented by the annual mean of the NINO3.4 index derived from ERRST v4, obtained from KNMI Climate Exploder. Sign is flipped in the plot since we’re looking for OHC loss during El Nino conditions. Both series plotted with cubic spline interpolation to make pretty curves. (Curvy models = good ….)

    Not a great fit … doing an actual regression reduces the amplitude of the “prediction” by a factor of five. I reckon the annual OHC deltas are somewhat spurious, especially pre-ARGO. I’m sure there’s some relationship, but teasing it out of the noise and controlling for confounding factors is likely less trivial than the bollocks I’ve done here. But it was fun doing.

    Oh … I tried the same technique (such as it is) down to 2 km, and the results were worse.

  84. JCH says:

    Well, one thing is… what does this effect, exaggerated, have, if any, on those wiggles:

  85. Harry Twinotter says:

    JCH.

    “It is very hard to get the extra, sequestered energy out of the oceans once it gets in there… because the enhanced greenhouse effect is progressively make its escape from its ocean prison ever incrementally more difficult.”

    Can you explain this statement to me? And what is an “ocean prison”?

    Yes, El Ninos do warm the atmosphere significantly. They are a redistribution of heat that was subducted into the ocean during neutral and La Nina years. They do not change the global OHC much relatively-speaking because it is large. So it is not hard to get the heat out, actually my guess is the heat gets out easier than it gets in.

  86. JCH says:

    Because the energy that is accumulating in the oceans due to the enhanced greenhouse effect is essentially trapped there for a very long time… centuries. Locked up. Assuming conditions remain essentially as they are, the accumulation of energy in the oceans will continue until equilibrium. If conditions change, a negative imbalance at the TOA, Ibelieve the oceans would rapidly vent energy to the atmosphere.

    In the current situation, if it were easy for the energy to get out, it would go.

    The textbooks say all energy that transfers from the oceans to the atmosphere does so within the skin layer. When there is an El Nino, the area of the ocean skin layer that is not being cooled by upwelling of extremely cold water from the depths appears to be increased. This would increase the transfer of energy from the oceans to the atmosphere: sensible heat, latent heat, and net radiation.

    The ocean water that is in the skin layer is primarily warmed by recent days’ sunlight, and that day’s down welling long wave radiation. I don’t see how that can be controversial.

    So I agree it is true that the oceans readily shed energy… unless something is inhibiting the process. The ever increasing enhanced greenhouse effect is making it more difficult… slowing down the shedding.

    A great deal of ocean surface is evaporated during an El Nino. How much of it came up from 400 to 700 meters below, and what was its starting heat content?

  87. Harry,
    I think it basically goes like this. Below the mixed layer, temperatures decreases with increasing depth. This means that you can’t transfer energy from the deep ocean to the upper ocean because that would mean transferring from a region that is cold to one that is warmer. You can have currents that move water from the deep ocean to the upper ocean, but water is essentially incompressible which means that if a volume of water moves from the deep ocean to the upper ocean, it must displace an essentially equal volume from the upper ocean. Since the deep ocean is colder than the upper ocean, the volume from the deep ocean must carry less energy than the volume it displaced. Therefore it is extremely difficult to move energy out of the ocean.

    What can happen is what is illustrated in the figure above. Winds can force warm water from one side to the other and force some down into a deeper layer. When the winds change, the water can flow back across to the other side and bring some of that warm water back up to the surface.

  88. Steven Mosher says:

    ““or its like saying CRU is too long because in 1850 it only covers Europe and the coastal US.”

    That would be a better comparison.”

    For grins I went and did the comparison of coverage at 1850.

    And then… a light went on.. so I am happy for Victor’s criticism.

    A couple things I noted.

    1. CRU has some sources that we miss. So I went back to look at our source for their data.. it was outdated.. 2011. Given our merge logic we put CRU at the bottom of the heap in terms
    of sources so they will only get used if they are a unique source of data. So.. If I can get an updated version of their station data I can probably fill in a few more blanks.. looking through the sources of their sources I think I found some additional data… so I’ve been busy looking at that.

    2. I have been wondering what I would get if I used CRU stations and the CRU method, BUT
    changed the gridding. They use an equal angle grid.. so at the equator a single station
    can be used as the basis of a 5 degree grid ( roughly 550km on a side ) BUT in the artic
    a 5 degree grid covers less area.. So I may do an equal area version of their data.
    The other thing I noticed is that when when grid lines align with country coasts … interesting issue there as well.

    That said. thank you for your comments they put me down some unexpected paths

  89. Harry Twinotter says:

    JCH.

    “The textbooks say all energy that transfers from the oceans to the atmosphere does so within the skin layer”

    Can you provide the name of one of these textbooks?

    Sure, heat transfer occurs at the interface between air and ocean. Ditto for air and ground. I would guess the heat transfer from ocean to air is reasonably efficient as shown by the moderating effect of the ocean on temperature during the night.

  90. Harry Twinotter says:

    ATTP.

    “Therefore it is extremely difficult to move energy out of the ocean.”

    Yes I know that, although I think you are referring to the deep ocean. What I am querying JCH about is some of his claims, and terms such as “ocean prison”.

    Talking about heat transfers from the deep ocean (which I wasn’t in my first post), sure you need some perturbation to move the deep water to shallower depths. As far as I understand the mechanism for the heat release during an El Nino is the subducted, warmer water in the western pacific is transfered to the eastern pacific via Kelvin waves, where it then upwells and increases the ocean surface temperature. IR radiation and latent heat transfer via evaporation then transfer the ocean surface heat to the atmosphere. Not much of a prison…

  91. Harry,

    Yes I know that, although I think you are referring to the deep ocean. What I am querying JCH about is some of his claims, and terms such as “ocean prison”.

    Ahh, sorry. I assumed that JCH was referring to the deep ocean as the ocean prison.

    I think what you say about the El Nino is about right. You can certainly transfer energy from the ocean to the atmosphere during an El Nino event.

  92. cce says:

    Hey Mosher,

    Some months ago, you implied that there was an all star team working on a grand unifying homogenization project. Is that still the case? Any progress?

    Thanks.

  93. Harry,

    I don’t know if skin-layer dynamics are textbook oceanography or not, but Peter Minnett’s writeup at RealClimate suggest that it’s a widely accepted phenomenon. I’ve looked for, and not found, a refereed publication of this particular experiment in literature, though he does have other papers on similar topics. Anyway, I keep that reference handy if only for this plot:

    … which is the best I know of demonstrating that, yes, incident downward LW radiation does modulate the rate at which the oceans lose retained solar energy.

    Because LW only penetrates sea water a few microns, I like to think of the oceans as sort of the Ultimate Greenhouse of the climate system. Perhaps not quite a prison, but when the only avenue of escape is at the very surface, it does tend to make them highly sensitive to what’s happening at that boundary.e, though he does have other papers on similar topics. Anyway, I keep that reference handy if only for this plot:

    … which is the best I know of demonstrating that, yes, incident downward LW radiation does modulate the rate at which the oceans lose retained solar energy.

    Because LW only penetrates sea water a few microns, I like to think of the oceans as sort of the Ultimate Greenhouse of the climate system. Perhaps not quite a prison, but when the only avenue of escape is at the very surface, it does tend to make them highly sensitive to what’s happening at that boundary.

  94. Sorry for the repeats, text editor user fail on my part ..

  95. JCH,

    Well, one thing is… what does this effect, exaggerated, have, if any, on those wiggles:

    Probably quite a bit. Courtesy of one of Judy’s Denizens, I was recently reading Moreno-Chamarro et al. (2016), though it is specific to the AMOC not ENSO:

    Here, we assess whether a reconstruction of the oceanic circulation can certainly be obtained by using the thermal-wind transport, and hence zonal density gradients, as predictor. This reconstruction technique constitutes the basis of earlier studies by Lund et al. (2006) and Lynch-Stieglitz et al. (2009). It directly quantifies the geostrophic transport of an oceanic current by applying the thermal wind equation to density estimates at its lateral boundaries. The advantage of this approach is that it is obtained from well-known physical relations independent of the models, in contrast to covariances based on model simulations; besides, it has already been proved valid for short, idealized model-based experiments (Hirschi and Lynch-Stieglitz 2006) and is the base for current in-situ AMOC measurements (e.g., Kanzow et al. 2007), though it is still uncertain whether this approach can be extended for estimations of variability on longer (multidecadal) timescales.

    All I need is a supercomputer hefty enough to run a full-blown AOGCM plus knowledge of what the heck I’m doing and I could do better than my toy linear regression models.

    stevenreincarnated provides a number of interesting citations in that thread which I found well worth reading if only as yet another exercise in finding out how much I don’t know about this stuff.

  96. Andy Skuce says:

    On the subject of the “ocean heat prison”, the way I see it is not so much that an El Niño turns an ocean heater up, but rather that it turns a natural air conditioner (upwelling cool water) down for a while.

    Correct me if I’m wrong.

  97. Ron Graf says:

    Andy, I would agree with your analogy. The prison is not so tight because there is a tunnel system that includes the poles where there’s plenty of ice cold (dense) water to replace any up-welling in the tropics. he Earth’s radiator motor runs at variable speed.

  98. Ron Graf says:

    JCH: Sea -haha – any OHC exiting the numerous El Nino events here? What would it look like?

    JCH, Brandon G, thanks for responding to my question regarding OHC response to oscillations in Earth radiance due to ENSO cycle. It appears the ocean shrugs off the atmosphere (mostly), as Brandon’s chart illustrates. There is no ocean energy jail. By the “just physics” part of climate science the radiative imbalance necessarily shifts with the Earth’s GMST.

    BBD, feel free to make an actual point any time you feel ready. But ease off the as hom please.

    JCH, you sea level chart seems to be in conflict with Brandon’s chart as far as expectation that thermal expansion caused SLR is more pronounced for boundary layer warming. Here is a chart from Trenberth(2013) that seems to be in conflict with all the charts. Can anyone reconcile it all to SLR? Trenberth seems to confirm my original idea of OHC decrease for high GMST (but I am not banking on its accuracy).

  99. BBD says:

    Ron G

    BBD, feel free to make an actual point any time you feel ready. But ease off the as hom please.

    I made my point clearly – you are bullshitting about Llovel. Ease of the tone-trolling please.

  100. BBD says:

    And Ron, you still haven’t answered the question…

    How can OHC be increasing in all major basins if surface warming is internally forced? OHC should be *falling* if there were a net transfer of energy from the oceans to the atmosphere.

    Come on.

  101. Climate sensitivity reconciled was the message of the article of Mark Richardson, Kevin Cowtan, Ed Hawkins, and Martin Stolpe on the bias in the temperature change.

    However, if you look at all the three biases we found for the equilibrium climate sensitivity up to now, the equilibrium climate sensitivity of these energy balance models is even (a bit) above the upper bound of the IPCC estimate: a whooping 4.6°C.

    Is it time to freak out about the climate sensitivity estimates from energy budget models?

    (Sneak preview: let’s at least wait a bit longer.)

  102. Steven Mosher: “or its like saying CRU is too long because in 1850 it only covers Europe and the coastal US.

    When I replied to this comment, I thought the CRU was a typo for Berkeley Earth. I now realise that that is unlikely for you as Berkeley member.

    Yes, also the data of CRU is not very trustworthy in that period. Maybe a bit more because they try to gather homogenized data as much as possible. And that nationally homogenized data is likely more reliable, especially in this period using lots of metadata (station history information) is important.

    Steven Mosher: “That said. thank you for your comments they put me down some unexpected paths

    Welcome. We have an article coming up you may like on how good relative homogenization works.

  103. Pingback: Matt Ridley’s lecture | …and Then There's Physics

  104. Pingback: 2016: A year in blogging | …and Then There's Physics

  105. Pingback: Judith Curry confuses laypeople about climate models | …and Then There's Physics

  106. Pingback: Reconciling ECS estimates – again | …and Then There's Physics

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s