Reposted from Dr. Roy Spencer’s International Warming Weblog.
by Roy W. Spencer, Ph. D.
In response to reviewers’ feedback on a paper John Christy and I submitted relating to the impression of El Nino and La Nina on local weather sensitivity estimates, I made a decision to alter the main focus sufficient to require a complete re-write of the paper.
The paper now addresses the query: If we take all the numerous floor and sub-surface temperature datasets and their differing estimates of warming over the past 50 years, what does it indicate for local weather sensitivity?
The difficulty with estimating local weather sensitivity from observational information is that, even when the temperature observations have been globally full and error-free, you continue to should know fairly precisely what the “forcing” was that triggered the temperature change.
(Sure, I do know a few of you don’t just like the forcing-feedback paradigm of local weather change. Be happy to disregard this put up if it bothers you.)
As a reminder, all temperature change in an object or system is because of an imbalance between charges of power gained and power misplaced, and the worldwide warming speculation begins with the belief that the local weather system is of course in a state of power steadiness. Sure, I do know (and agree) that this assumption can’t be demonstrated to be strictly true, as occasions just like the Medieval Heat Interval and Little Ice Age can attest.
However for the aim of demonstration, let’s assume it’s true in at present’s local weather system, and that the one factor inflicting current warming is anthropogenic greenhouse fuel emission (primarily CO2). Does the present price of warming recommend (as we’re informed) {that a} world warming catastrophe is upon us? I feel this is a vital query to handle, separate from the query of whether or not among the current warming is pure (which might make AGW even much less of an issue).
Lewis and Curry (most lately in 2018) addressed the ECS query in an identical method by evaluating temperatures and radiative forcing estimates between the late 1800s and early 2000s, and received solutions someplace within the vary of 1.5 to 1.8 deg. C of eventual warming from a doubling of the pre-industrial CO2 focus (2XCO2). These estimates are significantly decrease than what the IPCC claims from (principally) local weather mannequin projections.
Our method is considerably completely different from Lewis & Curry. First, we use solely information from the latest 50 years (1970-2021), which is the interval of most fast development in CO2-caused forcing, the interval of most fast temperature rise, and about way back to one can go and discuss with any confidence about ocean warmth content material (an important variable in local weather sensitivity estimates).
Secondly, our mannequin is time-dependent, with month-to-month time decision, permitting us to look at (as an example) the current acceleration in deep ocean temperature (ocean warmth content material) rise.
In distinction to Lewis & Curry and differencing two time durations’ averages separated by 100+ years, our method is to make use of a time-dependent mannequin of vertical power flows, which I’ve blogged on earlier than. It’s run at month-to-month time decision, so permits examination of such points because the current acceleration of the rise in oceanic warmth content material (OHC).
In response to reviewers feedback, I prolonged the area from non-ice lined (60N-60S) oceans to world protection (together with land), in addition to borehole-based estimates of deep-land warming developments (I consider a primary for this sort of work). The mannequin stays a 1D mannequin of temperature departures from assumed power equilibrium, inside three layers, proven schematically in Fig. 1.
One factor I discovered alongside the best way is that, though borehole temperatures recommend warming extending to virtually 200 m depth (the reason for which appears to extent again a number of centuries), fashionable Earth System Fashions (ESMs) have embedded land fashions that reach to solely 10 m depth or so.
One other factor I discovered (in the midst of responding to reviewers feedback) is that the assumed historical past of radiative forcing has a pretty big impact on recognized local weather sensitivity. I’ve been utilizing the RCP6 radiative forcing state of affairs from the earlier (AR5) IPCC report, however in response to reviewers’ solutions I’m now emphasizing the SSP245 state of affairs from the latest (AR6) report.
I run all the mannequin simulations with both one or the opposite radiative forcing dataset, initialized in 1765 (a standard start line for ESMs). All outcomes under are from the latest (SSP245) efficient radiative forcing state of affairs most well-liked by the IPCC (which, it seems, truly produces decrease ECS estimates).
The Mannequin Experiments
Along with the belief that the radiative forcing eventualities are a comparatively correct illustration of what has been inflicting local weather change since 1765, there may be additionally the belief that our temperature datasets are sufficiently correct to compute ECS values.
So, taking these on religion, let’s forge forward…
I ran the mannequin with hundreds of combos of warmth switch coefficients between mannequin layers and the online suggestions parameter (which determines ECS) to get 1970-2021 temperature developments inside sure ranges.
For land floor temperature developments I used 5 “completely different” land datasets: CRUTem5 (+0.277 C/decade), GISS 250 km (+0.306 C/decade), NCDC v3.2.1 (+0.298 C/decade), GHCN/CAMS (+0.348 C/decade), and Berkeley 1 deg. (+0.280 C/decade).
For world common sea floor temperature I used HadCRUT5 (+0.153 C/decade), Cowtan & Method (HadCRUT4, +0.148 C/decade), and Berkeley 1 deg. (+0.162 C/decade).
For the deep ocean, I used Cheng et al. 0-2000m world common ocean temperature (+0.0269 C/decade), and Cheng’s estimate of the 2000-3688m deep-deep-ocean warming, which quantities to a (very unsure) +0.01 whole warming over the past 40 years. The mannequin should produce the floor developments throughout the vary represented by these datasets, and produce 0-2000 m developments inside +/-20% of the Cheng deep-ocean dataset developments.
Since deep-ocean warmth storage is such an necessary constraint on ECS, in Fig. 3 I present the 1D mannequin run that most closely fits the 0-2000m temperature development of +0.0269 C/decade over the interval 1970-2021.
Lastly, the storage of warmth within the land floor is often ignored in such efforts. As talked about above, local weather fashions have embedded land floor fashions that reach to solely 10 m depth. But, borehole temperature profiles have been analyzed that recommend warming as much as 200 m in depth (Fig. 4).
This nice depth, in flip, means that there was a multi-century warming development occurring, even within the early twentieth Century, which the IPCC ignores and which suggests a pure supply for long-term local weather change. Any pure supply of warming, if ignored, results in inflated estimates of ECS and of the significance of accelerating CO2 in local weather change projections.
I used the black curve (backside panel of Fig. 4) to estimate that the near-surface layer is warming 2.5 occasions quicker than the 0-100 m layer, and 25 occasions quicker than the 100-200 m layer. In my 1D mannequin simulations, I required this quantity of deep-land warmth storage (analogous to the deep-ocean warmth storage computations, however requiring weaker warmth switch coefficients for land and completely different volumetric warmth capacities).
The distributions of recognized ECS values I recover from land and ocean are proven in Fig. 5.
The ultimate, world common ECS from the central estimates in Fig. 5 is 2.09 deg. C. Once more, that is considerably increased than the 1.5 to 1.8 deg. C obtained by Lewis & Curry, however a part of this is because of bigger estimates of ocean and land warmth storage used right here, and I’d suspect that our use of solely the latest 50 years of knowledge has some impression as nicely.
Conclusions
I’ve used a 1D time-dependent mannequin of temperature departures from assumed power equilibrium to handle the query: Given the varied estimates of floor and sub-surface warming over the past 50 years, what do they recommend for the sensitivity of the local weather system to a doubling of atmospheric CO2?
Utilizing the latest estimates of efficient radiative forcing from Annex III within the newest IPCC report (AR6), the observational information recommend decrease local weather sensitivities (ECS) than promoted by the IPCC with a central estimate of +2.09 deg C. for the worldwide common. That is on the backside finish of the newest IPCC (AR6) probably vary of two.0 to 4.5 deg. C.
I consider that is nonetheless probably an higher sure for ECS, for the next causes.
- Borehole temperatures recommend there was a long-term warming development, a minimum of up into the early twentieth Century. Ignoring this (no matter its trigger) will result in inflated estimates of ECS.
- I nonetheless consider that some portion of the land temperature datasets has been contaminated by long-term will increase in City Warmth Island results, that are indistinguishable from climatic warming in homogenization schemes.
