Home / Weather / Remark by way of Cowtan & Jacobs on Lewis & Curry 2018 and Answer: Section 1

Remark by way of Cowtan & Jacobs on Lewis & Curry 2018 and Answer: Section 1

From Dr. Judith Curry’s Local weather And many others.

Posted on December 16, 2019 by way of niclewis | 52 Feedback

Through Nic Lewis

A touch upon LC18 (contemporary paper by way of Lewis and Curry on local weather sensitivity)  by way of Cowtan and Jacobs has been printed, together with our reaction.


In an previous article right here I mentioned the Lewis and Curry (2018) paper “The affect of new forcing and ocean warmth uptake knowledge on estimates of local weather sensitivity” (LC18) and set out its effects.

The LC18 research used an international calories finances mannequin to estimate the planetary equilibrium local weather sensitivity (ECS) and brief local weather reaction (TCR). ECS and TCR are estimated from adjustments (Δ) in international imply floor temperature [T], efficient radiative forcing (ERF) [F] and the planetary radiative imbalance[1] [N] between a base and a last duration, as:

ECS = F2×CO2  × ΔT /(ΔF – ΔN)  and  TCR = F2×CO2 × ΔTF

the place F2×CO2 is the ERF for a doubling of atmospheric CO2 focus.

The primary LC18 estimates for ECS and TCR have been as in keeping with Desk 1. The details of observe are that they lie close to the ground finish of the IPCC AR5 ‘most probably’ levels for ECS and TCR, and that they’re each much less unsure and moderately less than the ones given within the predecessor learn about, Lewis & Curry (2015) when the usage of HadCRUT4 international floor temperature knowledge. The LC18 very best estimates in accordance with the speedier warming infilled Cowtan & Approach Had4_krig_v2 temperature dataset are similar to the HadCRUT4-based ends up in Lewis & Curry (2015).

Desk 1 (in accordance with Desk three in LC18) Very best estimates (medians) and uncertainty levels for ECS and TCR the usage of the bottom and ultimate sessions indicated. Values in roman sort compute the temperature alternate concerned (ΔT) the usage of the HadCRUT4v5 dataset; values in italics compute the usage of the infilled, globally-complete Had4_krig_v2 (Cowtan & Approach) dataset. The most well liked estimates are proven in daring. Levels are mentioned to the closest zero.05 Ok. Additionally proven are the similar effects (the usage of the HadCRUT4v2 dataset) from LC15 for the primary two duration mixtures given in that paper. ECS estimates suppose that efficient local weather sensitivity does no longer alternate with time elapsed since imposition of forcing.

Abstract of the Remark and Answer

A Remark on LC18 by way of Kevin Cowtan and Peter Jacobs, and a Answer from myself and Judith Curry, have simply been printed by way of Magazine of Local weather. A replica of the Answer is to be had right here.

The Remark (known as CJ20, because it  seems within the 1 January 2020 factor) is arguably extra a critique of observational sea floor temperature (SST) datasets than of the strategies and result of LC18. Its summary reads as follows:

Lewis and Curry (2018) (hereafter LC18) provide one way for the estimating the brief local weather reaction (TCR) of the local weather gadget from the temperature alternate between two time home windows – an early baseline duration within the 19th century, and a contemporary duration basically within the 21st century. The effects counsel a decrease price of TCR than estimates from local weather mannequin simulations. Earlier research have known uncertainty within the ancient forcings, the affect of the time evolution of the forcing on temperature reaction, and observational problems as contributory components to this war of words. We examine an additional issue: uncertainty within the bias corrections carried out to ancient sea floor temperature knowledge. This uncertainty can specifically affect the estimation of variables on decadal timescales, and subsequently affect the estimation of TCR the usage of the window way in addition to estimates of inside variability. We show that use of the entire ancient file can mitigate the affects of running with couple of minutes home windows to some degree, specifically with admire to the early a part of the file.

Firstly, CJ20 asserted that the bottom and ultimate sessions – what they name early and overdue home windows – selected in LC18 – that have been matched as regards volcanic forcing and affect from multidecadal inside variability – resulted in decrease values of TCR (CJ20 didn’t deal with the LC18 ECS estimates). They therefore got rid of that declare, which the research in our submitted Answer disproved. The general model of CJ20 specializes in the conceivable affect of the usage of home windows reasonably than all of the ancient knowledge, particularly the affect – in accordance with evaluating warming in CMIP5 (present technology) local weather fashions and in observations – of the collection of various dates for the home windows, and on uncertainty in bias corrections to ancient SST knowledge. CJ20 center of attention on use of the HadCRUT4 temperature file, however – as LC18 made transparent –  it’s suitable to make use of a globally whole file for comparability with local weather mannequin effects. We accordingly used handiest Kevin Cowtans’s infilled model of HadCRUT4, Had4_krig_v2, in our Answer.

The summary for my and Judith Curry’s Respond to CJ20 reads as follows:

Cowtan and Jacobs assert that the process utilized by Lewis and Curry  in 2018 (LC18) to estimate the local weather gadget’s brief local weather reaction (TCR) from adjustments between two time home windows is much less powerful – particularly in opposition to sea floor temperature bias correction uncertainty – than one way that makes use of all the ancient file. We show that TCR estimated the usage of all knowledge from the temperature file is intently in step with that estimated the usage of the LC18 home windows, as is the median TCR estimate the usage of all pairs of person years. We additionally display that the median TCR estimate from all pairs of decade-plus duration home windows is intently in step with that estimated the usage of the LC18 home windows, and that incorporating window variety uncertainty would make little distinction to overall uncertainty in TCR estimation. We discover that after variations within the evolution of forcing are accounted for, the connection through the years between warming in CMIP5 fashions and observations is in line with the connection between CMIP5 TCR and LC18’s TCR estimate, however fluctuates because of multidecadal inside variability and volcanism. We additionally display that quite a lot of different issues raised by way of Cowtan and Jacobs have negligible implications for TCR estimation in LC18.

In a nutshell, we refuted all issues of substance made in CJ20. I plan to care for the diversities between noticed and CMIP5 model-simulated ancient warming, which shaped the root of CJ20’s numerical research, in a next article. On this article, I will be able to elaborate on our refutation of issues in the rest of CJ20.

Window variety comparable uncertainty

In regards to the declare by way of CJ20 regarding uncertainty brought about by way of window selection, that is what we needed to say within the Answer, having examined the results of random collection of home windows from a decade upwards in duration,[2] all of which resulted in median TCR estimates very as regards to LC18’s 1.33 °C [= 1.33 K]:

For estimates with the perfect (2.zero Wm−2) minimal forcing building up, which might be maximum related to LC18’s TCR estimate, the Five–95% TCR uncertainty vary bobbing up from random window variety is 1.08–1.54 Ok, or 1.20–1.59 Ok the usage of zero.55-scaled volcanic forcing. The width of those levels – zero.103 and zero.073, respectively, in fractional same old deviation phrases[3] –  displays the truth that most of the window mixtures contain mismatched influences from inside variability and/or volcanism. Those window variety uncertainty levels don’t indicate that LC18 underestimated uncertainty in international temperature alternate: the 1σ fractional uncertainty in LC18’s most popular TCR estimate as a result of temperature alternate uncertainty (together with that from inside variability) on my own was once zero.103.[4] Additionally, despite the fact that no allowance is made for double counting of temperature alternate uncertainty, estimated total TCR uncertainty would building up little if window variety uncertainty have been added. Including (in quadrature) the zero.103 or zero.073 1σ fractional uncertainty in TCR from window variety to the 1σ fractional uncertainty of the most well liked LC18 TCR estimate, would handiest building up it to at least one.13⤬ its authentic stage, or to at least one.07⤬ that stage if the usage of zero.55-scaled volcanic forcing.[5]

This presentations that uncertainty in TCR estimation bobbing up from window variety is minor despite the fact that no allowance is made for double counting of temperature uncertainty, and negligible if allowance is made for such doubling counting.

The usage of knowledge from all the ancient file

CJ20 suggest use of information from all the ancient file. If truth be told, LC18 examined doing so, by way of the standard regression way, however discovered mismatching volcanic affect made estimation delicate to the scaling issue used for volcanic forcing. With out cutting down volcanic forcing the TCR estimate from regression over the entire ancient duration is a long way less than that from the usage of the home windows way. That is what we stated within the Answer:

When AR5 volcanic forcing is scaled by way of zero.55, regression of median annual-mean temperature on forcing over 1850–2016 offers a 1.27 Ok Had4_krig_v2-based TCR estimate, marginally less than LC18’s 1.33 Ok two-window founded most popular estimate. Regressing pentadal way (over 1852–2016) considerably improves the have compatibility (to an R2 of zero.92) and offers a TCR estimate of one.33 Ok. The usage of such pentadal-mean regression on every of the 500,000 pairs of samples of temperature and forcing time collection offers a Five–95% TCR vary of zero.91–1.84 Ok, marginally decrease and narrower than the LC18 most popular estimate vary.

So, the result of TCR estimation the usage of knowledge from all the ancient file is intently in step with the ones the usage of LC18’s window way and selected home windows, supplied the volcanic forcing is scaled down as in keeping with LC18’s advice. Alternatively, the uncertainty brought about by way of having to estimate the precise volcanic forcing scaling issue arguably makes the usage of knowledge from the whole ancient file a much less ample means than the usage of the home windows way.

Problems with ancient sea floor temperature knowledge

There may be certainly vital uncertainty as to the accuracy of the worldwide SST file. Alternatively, CJ20 didn’t display that the LC18 TCR estimates have been materially suffering from any known mistakes in SST bias corrections. Nor did they display that uncertainty within the SST file was once more than that estimated by way of the suppliers of the datasets utilized in LC18.

CJ20 make the purpose that protection of the ‘water hemisphere’ was once virtually non-existent within the 1860s. Alternatively, the 1869–82 number one early window utilized in LC18 avoids the 1860s (save for 1869, when protection was once higher), and offers moderately upper protection within the (land-sparse) southern hemisphere than within the northern hemisphere.

CJ20 additionally state that 19th century temperatures are depending on massive ‘bucket corrections’ to sea floor temperature (SST) observations, on the other hand CJ20 themselves counsel that the alternate from wood buckets to poorly insulated canvas buckets requiring a big bias correction happened basically throughout 1890–1910. Bucket corrections have been quite small throughout 1869–82, the LC18 early window.

Imaginable misestimation of forcings

That is what we wrote within the Answer regarding two forcing estimation problems raised in CJ20:

CJ20 declare that earlier research have known variations in inferred forcings and within the temperature affect of ancient as opposed to brief forcing adjustments as doable explanatory components for contemporary observational energy-budget TCR estimates being less than reasonable local weather mannequin TCR values. Not one of the 3 supporting research that they cite helps both rivalry.


CJ20 declare that comparability of modeled and noticed temperatures for overdue home windows beginning after 2005 is suffering from overestimation of forcings in fashions. Since LC18 didn’t make any comparisons of modeled and noticed temperatures over the ancient duration, the one factor of relevance to LC18 is whether or not it misestimated contemporary forcing. Not one of the 3 supporting research that CJ20 cite point out that LC18 misestimated contemporary forcing.

If truth be told, a extra complete learn about[6] discovered, of their CMIP5-specification ancient simulations, that for the reason that mid-2000s underestimation of adjustments in different forcing brokers greater than counteracted overestimation of adjustments in sun and volcanic forcing. Additionally, not one of the research cited in CJ20 addressed the actual drawback, of bias in CMIP5 mannequin forcing that already existed a number of a long time in the past (because of mainly to over the top aerosol forcing); none in their analyses began prior to 1980.

Ocean and air floor temperature in fashions and observations

In CMIP5 fashions near-surface marine air temperature warms greater than the sea floor temperature box (‘tos‘). CJ20 state that “Lewis and Curry argue that this box [tos] isn’t the highest layer of the majority ocean floor temperature” (to which measured SST widely corresponds). Alternatively, this straw guy argument, which CJ20 disprove, was once by no means made in LC18. As the answer states:

CJ20’s declare that LC18 “argue that this box [tos] isn’t the highest layer of the majority ocean floor temperature” is improper. Reasonably, LC18 argued that the tas/tos warming distinction displays the model-simulated warming distinction between tas and ocean pores and skin temperature, which is able to heat in a different way from SST.

There are theoretical causes for anticipating air simply above the sea floor to heat moderately sooner than the sea pores and skin temperature. Alternatively, the level of the adaptation is determined by many components and is unsure, as is the adaptation between the warming charges of SST and of ocean pores and skin temperature. LC18 subsequently curious about observational reasonably than CMIP5 mannequin proof on this house. We are saying within the Answer:

LC18 (segment 7e) concluded from observational and reanalysis proof that during the actual local weather gadget, tas warmed at maximum a couple of in keeping with cent greater than a mix of tas and tos (mannequin best ocean layer temperature), a considerably smaller distinction than that claimed by way of CJ20. Certainly, the 1979-onwards ERA-interim reanalysis globally-complete floor air temperature file, adjusted for inhomogeneities of their SST supply (Simmons et al. 2017), presentations moderately decrease warming over 1979–2016 than does Had4_krig_v2.

It is usually value noting that during CMIP5 fashions tas, in contrast to tos, is a diagnostic reasonably than a prognostic variable – this can be a parameterised extraneous variable, no longer a variable that includes within the elementary mannequin physics.


Not one of the criticisms of LC18 within the Answer rise up to exam. I go away exam of variations between noticed and CMIP5 model-simulated ancient warming, which shaped the root of CJ20’s numerical research, to a next article. Suffice to mention right here that such variations, when correctly analysed within the gentle of variations in forcing evolution, are totally in line with the LC18 TCR estimate.

Nicholas Lewis    December 2019

[1] N is estimated from its counterpart, the velocity of local weather gadget warmth uptake, which is basically by way of the sea.

[2] Since small inter-window forcing will increase supply deficient TCR estimation, minimal required inter-window forcing will increase, starting from 1.zero to two.zero Wm−2, have been imposed. (The better the forcing building up the decrease the relative uncertainty, as regards each forcing and the alternate in temperature that it reasons. The home windows used for LC18’s primary ECS and TCR estimates gave a forcing building up of two.52 Wm−2.) There have been over 11,000 decade plus lengthy window mixtures giving a forcing building up of two.zero Wm−2 or extra. For computational tractability, early and overdue home windows have been specified to be of equivalent duration. When the usage of LC18’s instructed zero.55 scaling of volcanic forcing the median TCR estimates have been even nearer to at least one.33 Ok in any respect ranges of required forcing building up, and had decrease uncertainty levels, than when the usage of unscaled volcanic forcing.

[3] As a way to be ready readily to mix uncertainties, we paintings with 1 same old deviation fractional uncertainties, right here derived by way of scaling from 17-83% levels and medians in Desk 1

[4] Scaling from the Five-95% vary and median for Had4_krig_v2 ΔT in Desk 2 of LC18. If temperature uncertainty on my own is integrated, the fractional uncertainty in TCR equals that during ΔT.

[5] Scaling from the 17-83% vary in Desk three of LC18, giving a fractional same old deviation of zero.193 for the most well liked LC18 TCR estimate. Uncertainties are taken to be usually dispensed and unbiased for the needs of deriving their same old deviations and mixing them. Including in quadrature a fractional same old deviation of zero.103 (zero.073) to the unique stage of zero.193 will increase it to zero.219 (zero.207).

[6] Outten, S., Thorne, P., Bethke, I. and Seland, Ø., 2015. Investigating the new obvious hiatus in floor temperature will increase: 1. Development of 2 30‐member Earth Gadget Style ensembles. Magazine of Geophysical Analysis: Atmospheres, 120(17), pp.8575-8596.

About admin

Check Also


Blockbuster information from China about COVID-19

Reposted from the Fabius Maximus Weblog Through Larry Kummer, Editor / 25 February 2020 Abstract: …

Leave a Reply

Your email address will not be published. Required fields are marked *