Two fallacies ensure meaningless public discussion about climate “crisis” or “emergency.” H/T to Terry Oldberg for comments and writings prompting me to post on this topic.

One corruption is the numerous times climate claims include fallacies of Equivocation. For instance, “climate change” can mean all observed events in nature, but as defined by IPCC all are 100% caused by human activities.  Similarly, forecasts from climate models are proclaimed to be “predictions” of future disasters, but renamed “projections” in disclaimers against legal liability.  And so on.

A second error in the argument is the Fallacy of Misplaced Concreteness, AKA Reification. This involves mistaking an abstraction for something tangible and real in time and space. We often see this in both spoken and written communications. It can take several forms:

♦ Confusing a word with the thing to which it refers

♦ Confusing an image with the reality it represents

♦ Confusing an idea with something observed to be happening

Examples of Equivocation and Reification from the World of Climate Alarm

“Seeing the wildfires, floods and storms, Mother Nature is not happy with us failing to recognize the challenges facing us.” – Nancy Pelosi

Mother Nature’ is a philosophical construct and has no feelings about people.

“This was the moment when the rise of the oceans began to slow and our planet began to heal …”
– Barack Obama

The ocean and the planet do not respond to someone winning a political party nomination. Nor does a planet experience human sickness and healing.

“If something has never happened before, we are generally safe in assuming it is not going to happen in the future, but the exceptions can kill you, and climate change is one of those exceptions.” – Al Gore

The future is not knowable, and can only be a matter of speculation and opinion.

“The planet is warming because of the growing level of greenhouse gas emissions from human activity. If this trend continues, truly catastrophic consequences are likely to ensue. “– Malcolm Turnbull

Temperature is an intrinsic property of an object, so temperature of “the planet” cannot be measured. The likelihood of catastrophic consequences is unknowable. Humans are blamed as guilty by association.

“Anybody who doesn’t see the impact of climate change is really, and I would say, myopic. They don’t see the reality. It’s so evident that we are destroying Mother Earth. “– Juan Manuel Santos

“Climate change” is an abstraction anyone can fill with subjective content. Efforts to safeguard the environment are real, successful and ignored in the rush to alarm.

“Climate change, if unchecked, is an urgent threat to health, food supplies, biodiversity, and livelihoods across the globe.” – John F. Kerry

To the abstraction “Climate Change” is added abstract “threats” and abstract means of “checking Climate Change.”

“Climate change is the most severe problem that we are facing today, more serious even than the threat of terrorism.” -David King

Instances of people killed and injured by terrorists are reported daily and are a matter of record, while problems from Climate Change are hypothetical

 

Climate Models Are Built on Fallacies

 

A previous post Chameleon Climate Models described the general issue of whether a model belongs on the bookshelf (theoretically useful) or whether it passes real world filters of relevance, thus qualifying as useful for policy considerations.

Following an interesting discussion on her blog, Dr. Judith Curry has written an important essay on the usefulness and limitations of climate models.

The paper was developed to respond to a request from a group of lawyers wondering how to regard claims based upon climate model outputs. The document is entitled Climate Models and is a great informative read for anyone. Some excerpts that struck me in italics with my bolds and added images.

Climate model development has followed a pathway mostly driven by scientific curiosity and computational limitations. GCMs were originally designed as a tool to help understand how the climate system works. GCMs are used by researchers to represent aspects of climate that are extremely difficult to observe, experiment with theories in a new way by enabling hitherto infeasible calculations, understand a complex system of equations that would otherwise be impenetrable, and explore the climate system to identify unexpected outcomes. As such, GCMs are an important element of climate research.

Climate models are useful tools for conducting scientific research to understand the climate system. However, the above points support the conclusion that current GCM climate models are not fit for the purpose of attributing the causes of 20th century warming or for predicting global or regional climate change on timescales of decades to centuries, with any high level of confidence. By extension, GCMs are not fit for the purpose of justifying political policies to fundamentally alter world social, economic and energy systems.

It is this application of climate model results that fuels the vociferousness
of the debate surrounding climate models.

The actual equations used in the GCM computer codes are only approximations of
the physical processes that occur in the climate system.

While some of these approximations are highly accurate, others are unavoidably crude. This is because the real processes they represent are either poorly understood or too complex to include in the model given the constraints of the computer system. Of the processes that are most important for climate change, parameterizations related to clouds and precipitation remain the most challenging, and are the greatest source of disagreement among different GCMs.

There are literally thousands of different choices made in the construction of a climate model (e.g. resolution, complexity of the submodels, parameterizations). Each different set of choices produces a different model having different sensitivities. Further, different modeling groups have different focal interests, e.g. long paleoclimate simulations, details of ocean circulations, nuances of the interactions between aerosol particles and clouds, the carbon cycle. These different interests focus their limited computational resources on a particular aspect of simulating the climate system, at the expense of others.

Overview of the structure of a state-of-the-art climate model. See Climate Models Explained by R.G. Brown

Human-caused warming depends not only on how much CO2 is added to the atmosphere, but also on how ‘sensitive’ the climate is to the increased CO2. Climate sensitivity is defined as the global surface warming that occurs when the concentration of carbon dioxide in the atmosphere doubles. If climate sensitivity is high, then we can expect substantial warming in the coming century as emissions continue to increase. If climate sensitivity is low, then future warming will be substantially lower.

In GCMs, the equilibrium climate sensitivity is an ‘emergent property’
that is not directly calibrated or tuned.

While there has been some narrowing of the range of modeled climate sensitivities over time, models still can be made to yield a wide range of sensitivities by altering model parameterizations. Model versions can be rejected or not, subject to the modelers’ own preconceptions, expectations and biases of the outcome of equilibrium climate sensitivity calculation.

Further, the discrepancy between observational and climate model-based estimates of climate sensitivity is substantial and of significant importance to policymakers. Equilibrium climate sensitivity, and the level of uncertainty in its value, is a key input into the economic models that drive cost-benefit analyses and estimates of the social cost of carbon.

Variations in climate can be caused by external forcing, such as solar variations, volcanic eruptions or changes in atmospheric composition such as an increase in CO2. Climate can also change owing to internal processes within the climate system (internal variability). The best known example of internal climate variability is El Nino/La Nina. Modes of decadal to centennial to millennial internal variability arise from the slow circulations in the oceans. As such, the ocean serves as a ‘fly wheel’ on the climate system, storing and releasing heat on long timescales and acting to stabilize the climate. As a result of the time lags and storage of heat in the ocean, the climate system is never in equilibrium.

The combination of uncertainty in the transient climate response (sensitivity) and the uncertainties in the magnitude and phasing of the major modes in natural internal variability preclude an unambiguous separation of externally forced climate variations from natural internal climate variability. If the climate sensitivity is on the low end of the range of estimates, and natural internal variability is on the strong side of the distribution of climate models, different conclusions are drawn about the relative importance of human causes to the 20th century warming.

Anthropogenic (human-caused) climate change is a theory in which the basic mechanism is well understood, but whose potential magnitude is highly uncertain.

What does the preceding analysis imply for IPCC’s ‘extremely likely’ attribution of anthropogenically caused warming since 1950? Climate models infer that all of the warming since 1950 can be attributed to humans. However, there have been large magnitude variations in global/hemispheric climate on timescales of 30 years, which are the same duration as the late 20th century warming. The IPCC does not have convincing explanations for previous 30 year periods in the 20th century, notably the warming 1910-1945 and the grand hiatus 1945-1975. Further, there is a secular warming trend at least since 1800 (and possibly as long as 400 years) that cannot be explained by CO2, and is only partly explained by volcanic eruptions.

Summary

There is growing evidence that climate models are running too hot and that climate sensitivity to CO2 is on the lower end of the range provided by the IPCC. Nevertheless, these lower values of climate sensitivity are not accounted for in IPCC climate model projections of temperature at the end of the 21st century or in estimates of the impact on temperatures of reducing CO2 emissions.

The climate modeling community has been focused on the response of the climate to increased human caused emissions, and the policy community accepts (either explicitly or implicitly) the results of the 21st century GCM simulations as actual predictions. Hence we don’t have a good understanding of the relative climate impacts of the above (natural factors) or their potential impacts on the evolution of the 21st century climate.

Footnote:

There are a series of posts here which apply reality filters to attest climate models.  The first was Temperatures According to Climate Models where both hindcasting and forecasting were seen to be flawed.

Others in the Series are:

Sea Level Rise: Just the Facts

Data vs. Models #1: Arctic Warming

Data vs. Models #2: Droughts and Floods

Data vs. Models #3: Disasters

Data vs. Models #4: Climates Changing

Climate Medicine

Climates Don’t Start Wars, People Do

Patrick J. Michaels reports at Real Clear Policy Nobel Prize Awarded for the Worst Climate Model. Excerpts in italics with my bolds and added images.

Given the persistent headlines about climate change over the years, it’s surprising how long it took the Nobel Committee to award the Physics prize to a climate modeler, which finally occurred earlier this month.

Indeed, Syukuro Manabe has been a pioneer in the development of so-called general circulation climate models (GCMs) and more comprehensive Earth System Models (ESMs). According to the Committee, Manabe was awarded the prize “For the physical modelling of the earth’s climate, quantifying variability, and reliably predicting global warming.”

What Manabe did was to modify early global weather forecasting models, adapting them to long-term increases in human emissions of carbon dioxide that alter the atmosphere’s internal energy balance, resulting in a general warming of surface temperatures, along with a much larger warming of temperatures above the surface over the earth’s vast tropics.

Unlike some climate modelers, like NASA’s James Hansen — who lit the bonfire of the greenhouse vanities in 1988, Manabe is hardly a publicity hound. And while politics clearly influences it (see Al Gore’s 2007 Prize), the Nobel Committee also respects primacy, as Manabe’s model was the first comprehensive GCM. He produced it at the National Oceanic and Atmospheric Administration’s Geophysical Fluid Dynamics Laboratory (GFDL) in Princeton NJ. The seminal papers were published in 1975 and 1980.

And, after many modifications and renditions, it is also the most incorrect of all the world’s GCMs at altitude over the vast tropics of the planet.

Getting the tropical temperatures right is critical. The vast majority of life-giving moisture that falls over the worlds productive midlatitude agrosystems originates as evaporation from the tropical oceans.

The major determinant of how much moisture is wafted into our region is the vertical distribution of tropical temperature. When the contrast is great, with cold temperatures aloft compared to the normally hot surface, that surface air is buoyant and ascends, ultimately transferring moisture to the temperate zones. When the contrast is less, the opposite occurs, and less moisture enters the atmosphere.

Every GCM or ESM predicts that several miles above the tropical surface should be a “hot spot,” where there is much more warming caused by carbon dioxide emissions than at the surface. If this is improperly forecast, then subsequent forecasts of rainfall over the world’s major agricultural regions will be unreliable.

That in turn will affect forecasts of surface temperature. Everyone knows a wet surface heats up (and cools down) slower than a dry one (see: deserts), so getting the moisture input right is critical.

Following Manabe, vast numbers of modelling centers popped up, mushrooms fertilized by public — and only public — money.

Every six years or so, the U.S. Department of Energy collects all of these models, aggregating them into what they call Coupled Model Intercomparison Projects (CMIPs). These serve as the bases for the various “scientific assessments” of climate change produced by the U.N.’s Intergovernmental Panel on Climate Change (IPCC) or the U.S. “National Assessments” of climate.

In 2017, University of Alabama’s John Christy, along with Richard McNider, published a paper that, among other things, examined the 25 applicable families of CMIP-5 models, comparing their performance to what’s been observed in the three-dimensional global tropics. Take a close look at Figure 3 from the paper, in the Asia-Pacific Journal of Atmospheric Sciences, and you’ll see that the model GFDL-CM3 is so bad that it is literally off the scale of the graph. [See Climate Models: Good, Bad and Ugly]

At its worst, the GFDL model is predicting approximately five times as much warming as has been observed since the upper-atmospheric data became comprehensive in 1979. This is the most evolved version of the model that won Manabe the Nobel.

In the CMIP-5 model suite, there is one, and only one, that works. It is the model INM-CM4 from the Russian Institute for Numerical Modelling, and the lead author is Evgeny Volodin. It seems that Volodin would be much more deserving of the Nobel for, in the words of the committee “reliably predicting global warming.”

Might this have something to do with the fact that INM-CM4 and its successor models have less predicted warming than all of the other models?

Patrick J. Michaels is a senior fellow working on energy and environment issues at the Competitive Enterprise Institute and author of “Scientocracy: The Tangled Web of Public Science and Public Policy.”

 

Roger Pielke Jr. has a new paper at Science Direct Distorting the view of our climate future: The misuse and abuse of climate pathways and scenarios.  Excerpt in italics with my bolds.

Abstract

Climate science research and assessments under the umbrella of the Intergovernmental Panel on Climate Change (IPCC) have misused scenarios for more than a decade. Symptoms of misuse have included the treatment of an unrealistic, extreme scenario as the world’s most likely future in the absence of climate policy and the illogical comparison of climate projections across inconsistent global development trajectories.

Reasons why such misuse arose include (a) competing demands for scenarios from users in diverse academic disciplines that ultimately conflated exploratory and policy relevant pathways, (b) the evolving role of the IPCC – which extended its mandate in a way that creates an inter-relationship between literature assessment and literature coordination, (c) unforeseen consequences of employing a temporary approach to scenario development, (d) maintaining research practices that normalize careless use of scenarios, and (e) the inherent complexity and technicality of scenarios in model-based research and in support of policy.

Consequently, much of the climate research community is presently off-track from scientific coherence and policy-relevance.

Attempts to address scenario misuse within the community have thus far not worked. The result has been the widespread production of myopic or misleading perspectives on future climate change and climate policy. Until reform is implemented, we can expect the production of such perspectives to continue, threatening the overall credibility of the IPCC and associated climate research. However, because many aspects of climate change discourse are contingent on scenarios, there is considerable momentum that will make such a course correction difficult and contested – even as efforts to improve scenarios have informed research that will be included in the IPCC 6th Assessment.

Discussion of How Imaginary Scenarios Spoil Attempts to Envision Climate Futures

The article above is paywalled, but a previous post reprinted below goes into the background of the role of scenarios in climate modelling, and demonstrates the effects referring to results from the most realistic model, INMCM5

Roger Pielke Jr. explains that climate models projections are unreliable because they are based on scenarios no longer bounded by reality.  His article is The Unstoppable Momentum of Outdated Science.  Excerpts in italics with my bolds.

Much of climate research is focused on implausible scenarios of the future, but implementing a course correction will be difficult.

In 2020, climate research finds itself in a similar situation to that of breast cancer research in 2007. Evidence indicates the scenarios of the future to 2100 that are at the focus of much of climate research have already diverged from the real world and thus offer a poor basis for projecting policy-relevant variables like economic growth and carbon dioxide emissions. A course-correction is needed.

In a new paper of ours just out in Environmental Research Letters we perform the most rigorous evaluation to date of how key variables in climate scenarios compare with data from the real world (specifically, we look at population, economic growth, energy intensity of economic growth and carbon intensity of energy consumption). We also look at how these variables might evolve in the near-term to 2040.

We find that the most commonly-used scenarios in climate research have already diverged significantly from the real world, and that divergence is going to only get larger in coming decades. You can see this visualized in the graph above, which shows carbon dioxide emissions from fossil fuels from 2005, when many scenarios begin, to 2045. The graph shows emissions trajectories projected by the most commonly used climate scenarios (called SSP5-8.5 and RCP8.5, with labels on the right vertical axis), along with other scenario trajectories. Actual emissions to date (dark purple curve) and those of near-term energy outlooks (labeled as EIA, BP and ExxonMobil) all can be found at the very low end of the scenario range, and far below the most commonly used scenarios.

Our paper goes into the technical details, but in short, an important reason for the lower-than-projected carbon dioxide emissions is that economic growth has been slower than expected across the scenarios, and rather than seeing coal use expand dramatically around the world, it has actually declined in many regions.

It is even conceivable, if not likely, that in 2019 the world has passed “peak carbon dioxide emissions.” Crucially, the projections in the figure above are pre-Covid19, which means that actual emissions 2020 to 2045 will be even less than was projected in 2019.

While it is excellent news that the broader community is beginning to realize that scenarios are increasingly outdated, voluminous amounts of research have been and continue to be produced based on the outdated scenarios. For instance, O’Neill and colleagues find that “many studies” use scenarios that are “unlikely.” In fact, in their literature review such “unlikely” scenarios comprise more than 20% of all scenario applications from 2014 to 2019. They also call for “re-examining the assumptions underlying” the high-end emissions scenarios that are favored in physical climate research, impact studies and economic and policy analyses.

Make no mistake. The momentum of outdated science is powerful. Recognizing that a considerable amount of climate science to be outdated is, in the words of the late Steve Rayer, “uncomfortable knowledge” — that knowledge which challenges widely-held preconceptions. According to Rayner, in such a context we should expect to see reactions to uncomfortable knowledge that include:

• denial (that scenarios are off track),
• dismissal (the scenarios are off track, but it doesn’t matter),
• diversion (the scenarios are off track, but saying so advances the agenda of those opposed to action) and,
• displacement (the scenarios are off track but there are perhaps compensating errors elsewhere within scenario assumptions).

Such responses reinforce the momentum of outdated science and make it more difficult to implement a much needed course correction.

Responding to climate change is critically important. So too is upholding the integrity of the science which helps to inform those responses. Identification of a growing divergence between scenarios and the real-world should be seen as an opportunity — to improve both science and policy related to climate — but also to develop new ways for science to be more nimble in getting back on track when research is found to be outdated.

[A previous post is reprinted below since it demonstrates how the scenarios drive forecasting by CMIP6 models, including the example of the best performant model: INMCM5]

Background from Previous Post : Best Climate Model: Mild Warming Forecasted

Links are provided at the end to previous posts describing climate models 4 and 5 from the Institute of Numerical Mathematics in Moscow, Russia.  Now we have forecasts for the 21st Century published for INM-CM5 at Izvestiya, Atmospheric and Oceanic Physics volume 56, pages218–228(July 7, 2020). The article is Simulation of Possible Future Climate Changes in the 21st Century in the INM-CM5 Climate Model by E. M. Volodin & A. S. Gritsun.  Excerpts are in italics with my bolds, along with a contextual comment.

Abstract

Climate changes in 2015–2100 have been simulated with the use of the INM-CM5 climate model following four scenarios: SSP1-2.6, SSP2-4.5, and SSP5-8.5 (single model runs) and SSP3-7.0 (an ensemble of five model runs). Changes in the global mean temperature and spatial distribution of temperature and precipitation are analyzed. The global warming predicted by the INM-CM5 model in the scenarios considered is smaller than that in other CMIP6 models. It is shown that the temperature in the hottest summer month can rise more quickly than the seasonal mean temperature in Russia. An analysis of a change in Arctic sea ice shows no complete Arctic summer ice melting in the 21st century under any model scenario. Changes in the meridional stream function in atmosphere and ocean are studied.

Overview

The climate is understood as the totality of statistical characteristics of the instantaneous states of the atmosphere, ocean, and other climate system components averaged over a long time period.

Therefore, we restrict ourselves to an analysis of some of the most important climate parameters, such as average temperature and precipitation. A more detailed analysis of individual aspects of climate change, such as changes in extreme weather and climate situations, will be the subject of another work. This study is not aimed at a full comparison with the results of other climate models, where calculations follow the same scenarios, since the results of other models have not yet been published in peer reviewed journals by the time of this writing.

The INM-CM5 climate model [1, 2] is used for the numerical experiments. It differs from the previous version, INMCM4, which was also used for experiments on reproducing climate change in the 21st century [3], in the following:

• an aerosol block has been added to the model, which allows inputting anthropogenic emissions of aerosols and their precursors;
• the concentrations and optical properties of aerosols are calculated, but not specified, like in the previous version;
• the parametrizations of cloud formation and condensation are changed in the atmospheric block;
• the upper boundary in the atmospheric block is raised from 30 to 60 km;
• the horizontal resolution in the ocean block is doubled along each coordinate; and,
• the software related to adaptation to massively parallel computers is improved, which allows the effective use a larger number of compute cores.

The model resolution in the atmospheric and aerosol blocks is 2° × 1.5° in longitude and latitude and 73 levels and, in the ocean, 0.5° × 0.25° and 40 levels. The calculations were performed at supercomputers of the Joint Supercomputer Center, Russian Academy of Sciences, and Moscow State University, with the use of 360 to 720 cores. The model calculated 6–10 years per 24 h in the above configuration.

Four scenarios were used to model the future climate: SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5. The scenarios are described in [4]. The figure after the abbreviation SSP (Shared Socioeconomic Pathway) is the number of the mankind development path (see the values in [4]). The number after the dash means the radiation forcing (W m–2) in 2100 compared to the preindustrial level. Thus, the SSP1-2.6 scenario is the most moderate and assumes rapid actions which sharply limit and then almost completely stop anthropogenic emissions. Within this scenario, greenhouse gas concentrations are maximal in the middle of the 21st century and then slightly decrease by the end of the century. The SSP5-8.5 scenario is the warmest and implies the fastest climate change. The scenarios are recommended for use in the project on comparing CMIP6 (Coupled Model Intercomparison Project, Phase 6, [5]) climate models.  Each scenario includes the time series of:

• carbon dioxide, methane, nitrous oxide, and ozone concentrations;
• emissions of anthropogenic aerosols and their precursors;
• the concentration of volcanic sulfate aerosol; and
• the solar constant. 

One model experiment was carried out for each of the above scenarios. It began at the beginning of 2015 and ended at the end of 2100. The initial state was taken from the so-called historical experiment with the same model, where climate changes were simulated for 1850–2014, and all impacts on the climate system were set according to observations. The results of the ensemble of historical experiments with the model under consideration are given in [6, 7]. For the SSP3-7.0 scenario, five model runs was performed differing in the initial data taken from different historical experiments. The ensemble of numerical experiments is required to increase the statistical confidence of conclusions about climate changes.

[My Contextual Comment inserted Prior to Consideration of Results]

Firstly, the INM-CM5 historical experiment can be read in detail by following a linked post (see Resources at the end), but this graphic summarizes the model hindcasting of past temperatures (GMT) compared to HadCrutv4.

Secondly, the scenarios are important to understand since they stipulate data inputs the model must accept as conditions for producing forecasts according to a particular scenario (set of assumptions).  The document with complete details referenced as [4] is The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6.

All the details are written there but one diagram suggests the implications for the results described below.

As shown, the SSP1-26 is virtually the same scenario as the former RCP2.6, while SSP5-85 is virtually the same as RCP8.5, the wildly improbable scenario (impossible according to some analysts).  Note that FF CO2 emissions are assumed to quadruple in the next 80 years, with atmospheric CO2 rising from 400 to 1000 ppm ( +150%).  Bear these suppositions in mind when considering the INMCM5 forecasts below.

Results [Continuing From Volodin and Gritsun]

Let us describe some simulation results of climate change in the 21st century. Figure 1 shows the change in the globally averaged surface air temperature with respect to the data of the corresponding historical experiment for 1850–1899. In the warmest SSP5-8.5 scenario (circles), the temperature rises by more than 4° by the end of the 21st century. In the SSP3-7.0 scenario (crosses), different members of the ensemble show warming by 3.4°–3.6°. In the SSP2-4.5 scenario (squares), the temperature increases by about 2.4°. According to the SSP1-2.6 scenario (triangles) , the maximal warming by ~1.7° occurs in the middle of the 21st century, and the temperature exceeds the preindustrial temperature by 1.4° by the end of the century.

[My comment: Note that the vertical scale starts with +1.0C as was seen in the historical experiment. Thus an anomaly of 1.4C by 2100 is an increase of only 0.4C, while the SSP2-4.5 result adds 1.4C to the present]. 

The results for other CMIP6 models have not yet been published in peer-reviewed journals. However, according to the preliminary analysis (see, e.g.  https://cmip6workshop19.sciencesconf.org/ data/Session1_PosterSlides.pdf, p.29), the INM-CM5 model shows the lowest temperature increase among the CMIP6 models considered for all the scenarios due to the minimal equilibrium sensitivity to the CO2 concentration doubling, which is ~2.1° for the current model version, like for the previous version, despite new condensation and cloud formation blocks. [For more on CMIP6 comparisons see post Climate Models: Good, Bad and Ugly]

The changes in the surface air temperature are similar for all scenarios; therefore, we analyze the difference between temperatures in 2071–2100 and 1981–2010 under the SSP5-8.5 and SSP1-2.6 scenarios (Fig. 2). The warming is maximal in the Arctic; it reaches 10° and 3°, respectively. Other features mainly correspond to CMIP5 data [8], including the INMCM4 model, which participates in the comparison. The warming on the continents of the Northern Hemisphere is about 2 times higher than the mean, and the warming in the Southern Hemisphere is noticeably less than in the Northern Hemisphere. The land surface is getting warmer than the ocean surface in all the scenarios except SSP1-2.6, because the greenhouse effect is expected to weaken in the second half of the 21st century in this scenario, and the higher heat capacity of the ocean prevents it from cooling as quickly as the land.

The changes in precipitation in December–February and June–August for the SSP3-7.0 scenario averaged over five members of the ensemble are shown in Fig. 4. All members of the ensemble show an increase in precipitation in the winter in a significant part of middle and high latitudes. In summer, the border between the increase and decrease in precipitation in Eurasia passes mainly around or to the north of 60°. In southern and central Europe, all members of the ensemble show a decrease in precipitation. Precipitation also increases in the region of the summer Asian monsoon, over the equatorial Pacific, due to a decrease in the upwelling and an increase in ocean surface temperature (OST). The distribution of changes in precipitation mainly corresponds to that given in [6, Fig. 12.22] for all CMIP5 models.

The change in the Arctic sea ice area in September, when the ocean ice cover is minimal over the year, is of interest. Figure 5 shows the sea ice area in September 2015–2019 to be 4–6 million km2 in all experiments, which corresponds to the estimate from observations in [11]. The Arctic sea ice does not completely melt in any of the experiments and under any scenario. However, according to [8, Figs. 12.28 and 12.31], many models participating in CMIP6, where the Arctic ice area is similar to that observed at the beginning of the 21st century, show the complete absence of ice by the end of the 21st century, especially under the RCP8.5 scenario, which is similar to SSP5-8.5.

The reason for these differences is the lower equilibrium sensitivity of the INM-CM5 model.

Note that the scatter of data between experiments under different scenarios in the first half of the 21st century is approximately the same as between different members of the ensemble under the SSP3-7.0 scenario and becomes larger only after 2070. The sea ice area values are sorted in accordance with the radiative forcing of the scenarios only after 2090. This indicates the large contribution of natural climate variability into the Arctic ice area. In the SSP1-2.6 experiment, the Arctic ice area at the end of the 21st century approximately corresponds to its area at the beginning of the experiment.

Climate changes can be also traced in the ocean circulation. Figure 6 shows the change in the 5-year averaged intensity of the Atlantic meridional circulation, defined as the maximum of the meridional streamfunction at 32° N. All experiments show a decrease in the intensity of meridional circulation in the 21st century and natural fluctuations against this decrease. The decrease is about 4.5–5 Sv for the SSP5-8.5 scenario, which is close to values obtained in the CMIP5 models [8, Fig. 12.35] under the RCP8.5 scenario. Under milder scenarios, the weakening of the meridional circulation is less pronounced. The reason for this weakening of the meridional circulation in the Atlantic, as far as we know, is not yet fully understood.

Conclusion

Numerical experiments have been carried out to reproduce climate changes in the 21st century according to four scenarios of the CMIP6 program [4, 5], including an ensemble of five experiments under the SSP3-7.0 scenario. The changes in the global mean surface temperature are analyzed. It is shown that the global warming predicted by the INM-CM5 model is the lowest among the currently published CMIP6 model data. The geographical distribution of changes in the temperature and precipitation is considered. According to the model, the temperature in the warmest summer month will increase faster than the summer average temperature in Russia.

None of the experiments show the complete melting of the Arctic ice cover by the end of the 21st century. Some changes in the ocean dynamics, including the flow velocity and the meridional stream function, are analyzed. The changes in the Hadley and Ferrel circulation in the atmosphere are considered.

Resources:

Climate Models: Good, Bad and Ugly

2018 Update: Best Climate Model INMCM5

Temperatures According to Climate Models

Roger Pielke Jr. explains that climate models projections are unreliable because they are based on scenarios no longer bounded by reality.  His article is The Unstoppable Momentum of Outdated Science.  Excerpts in italics with my bolds.

Much of climate research is focused on implausible scenarios of the future, but implementing a course correction will be difficult.

In 2020, climate research finds itself in a similar situation to that of breast cancer research in 2007. Evidence indicates the scenarios of the future to 2100 that are at the focus of much of climate research have already diverged from the real world and thus offer a poor basis for projecting policy-relevant variables like economic growth and carbon dioxide emissions. A course-correction is needed.

In a new paper of ours just out in Environmental Research Letters we perform the most rigorous evaluation to date of how key variables in climate scenarios compare with data from the real world (specifically, we look at population, economic growth, energy intensity of economic growth and carbon intensity of energy consumption). We also look at how these variables might evolve in the near-term to 2040.

We find that the most commonly-used scenarios in climate research have already diverged significantly from the real world, and that divergence is going to only get larger in coming decades. You can see this visualized in the graph above, which shows carbon dioxide emissions from fossil fuels from 2005, when many scenarios begin, to 2045. The graph shows emissions trajectories projected by the most commonly used climate scenarios (called SSP5-8.5 and RCP8.5, with labels on the right vertical axis), along with other scenario trajectories. Actual emissions to date (dark purple curve) and those of near-term energy outlooks (labeled as EIA, BP and ExxonMobil) all can be found at the very low end of the scenario range, and far below the most commonly used scenarios.

Our paper goes into the technical details, but in short, an important reason for the lower-than-projected carbon dioxide emissions is that economic growth has been slower than expected across the scenarios, and rather than seeing coal use expand dramatically around the world, it has actually declined in many regions.

It is even conceivable, if not likely, that in 2019 the world has passed “peak carbon dioxide emissions.” Crucially, the projections in the figure above are pre-Covid19, which means that actual emissions 2020 to 2045 will be even less than was projected in 2019.

While it is excellent news that the broader community is beginning to realize that scenarios are increasingly outdated, voluminous amounts of research have been and continue to be produced based on the outdated scenarios. For instance, O’Neill and colleagues find that “many studies” use scenarios that are “unlikely.” In fact, in their literature review such “unlikely” scenarios comprise more than 20% of all scenario applications from 2014 to 2019. They also call for “re-examining the assumptions underlying” the high-end emissions scenarios that are favored in physical climate research, impact studies and economic and policy analyses.

Make no mistake. The momentum of outdated science is powerful. Recognizing that a considerable amount of climate science to be outdated is, in the words of the late Steve Rayer, “uncomfortable knowledge” — that knowledge which challenges widely-held preconceptions. According to Rayner, in such a context we should expect to see reactions to uncomfortable knowledge that include:

• denial (that scenarios are off track),
• dismissal (the scenarios are off track, but it doesn’t matter),
• diversion (the scenarios are off track, but saying so advances the agenda of those opposed to action) and,
• displacement (the scenarios are off track but there are perhaps compensating errors elsewhere within scenario assumptions).

Such responses reinforce the momentum of outdated science and make it more difficult to implement a much needed course correction.

Responding to climate change is critically important. So too is upholding the integrity of the science which helps to inform those responses. Identification of a growing divergence between scenarios and the real-world should be seen as an opportunity — to improve both science and policy related to climate — but also to develop new ways for science to be more nimble in getting back on track when research is found to be outdated.

[A previous post is reprinted below since it demonstrates how the scenarios drive forecasting by CMIP6 models, including the example of the best performant model: INMCM5]

Background from Previous Post : Best Climate Model: Mild Warming Forecasted

Links are provided at the end to previous posts describing climate models 4 and 5 from the Institute of Numerical Mathematics in Moscow, Russia.  Now we have forecasts for the 21st Century published for INM-CM5 at Izvestiya, Atmospheric and Oceanic Physics volume 56, pages218–228(July 7, 2020). The article is Simulation of Possible Future Climate Changes in the 21st Century in the INM-CM5 Climate Model by E. M. Volodin & A. S. Gritsun.  Excerpts are in italics with my bolds, along with a contextual comment.

Abstract

Climate changes in 2015–2100 have been simulated with the use of the INM-CM5 climate model following four scenarios: SSP1-2.6, SSP2-4.5, and SSP5-8.5 (single model runs) and SSP3-7.0 (an ensemble of five model runs). Changes in the global mean temperature and spatial distribution of temperature and precipitation are analyzed. The global warming predicted by the INM-CM5 model in the scenarios considered is smaller than that in other CMIP6 models. It is shown that the temperature in the hottest summer month can rise more quickly than the seasonal mean temperature in Russia. An analysis of a change in Arctic sea ice shows no complete Arctic summer ice melting in the 21st century under any model scenario. Changes in the meridional stream function in atmosphere and ocean are studied.

Overview

The climate is understood as the totality of statistical characteristics of the instantaneous states of the atmosphere, ocean, and other climate system components averaged over a long time period.

Therefore, we restrict ourselves to an analysis of some of the most important climate parameters, such as average temperature and precipitation. A more detailed analysis of individual aspects of climate change, such as changes in extreme weather and climate situations, will be the subject of another work. This study is not aimed at a full comparison with the results of other climate models, where calculations follow the same scenarios, since the results of other models have not yet been published in peer reviewed journals by the time of this writing.

The INM-CM5 climate model [1, 2] is used for the numerical experiments. It differs from the previous version, INMCM4, which was also used for experiments on reproducing climate change in the 21st century [3], in the following:

• an aerosol block has been added to the model, which allows inputting anthropogenic emissions of aerosols and their precursors;
• the concentrations and optical properties of aerosols are calculated, but not specified, like in the previous version;
• the parametrizations of cloud formation and condensation are changed in the atmospheric block;
• the upper boundary in the atmospheric block is raised from 30 to 60 km;
• the horizontal resolution in the ocean block is doubled along each coordinate; and,
• the software related to adaptation to massively parallel computers is improved, which allows the effective use a larger number of compute cores.

The model resolution in the atmospheric and aerosol blocks is 2° × 1.5° in longitude and latitude and 73 levels and, in the ocean, 0.5° × 0.25° and 40 levels. The calculations were performed at supercomputers of the Joint Supercomputer Center, Russian Academy of Sciences, and Moscow State University, with the use of 360 to 720 cores. The model calculated 6–10 years per 24 h in the above configuration.

Four scenarios were used to model the future climate: SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5. The scenarios are described in [4]. The figure after the abbreviation SSP (Shared Socioeconomic Pathway) is the number of the mankind development path (see the values in [4]). The number after the dash means the radiation forcing (W m–2) in 2100 compared to the preindustrial level. Thus, the SSP1-2.6 scenario is the most moderate and assumes rapid actions which sharply limit and then almost completely stop anthropogenic emissions. Within this scenario, greenhouse gas concentrations are maximal in the middle of the 21st century and then slightly decrease by the end of the century. The SSP5-8.5 scenario is the warmest and implies the fastest climate change. The scenarios are recommended for use in the project on comparing CMIP6 (Coupled Model Intercomparison Project, Phase 6, [5]) climate models.  Each scenario includes the time series of:

• carbon dioxide, methane, nitrous oxide, and ozone concentrations;
• emissions of anthropogenic aerosols and their precursors;
• the concentration of volcanic sulfate aerosol; and
• the solar constant. 

One model experiment was carried out for each of the above scenarios. It began at the beginning of 2015 and ended at the end of 2100. The initial state was taken from the so-called historical experiment with the same model, where climate changes were simulated for 1850–2014, and all impacts on the climate system were set according to observations. The results of the ensemble of historical experiments with the model under consideration are given in [6, 7]. For the SSP3-7.0 scenario, five model runs was performed differing in the initial data taken from different historical experiments. The ensemble of numerical experiments is required to increase the statistical confidence of conclusions about climate changes.

[My Contextual Comment inserted Prior to Consideration of Results]

Firstly, the INM-CM5 historical experiment can be read in detail by following a linked post (see Resources at the end), but this graphic summarizes the model hindcasting of past temperatures (GMT) compared to HadCrutv4.

Secondly, the scenarios are important to understand since they stipulate data inputs the model must accept as conditions for producing forecasts according to a particular scenario (set of assumptions).  The document with complete details referenced as [4] is The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6.

All the details are written there but one diagram suggests the implications for the results described below.

As shown, the SSP1-26 is virtually the same scenario as the former RCP2.6, while SSP5-85 is virtually the same as RCP8.5, the wildly improbable scenario (impossible according to some analysts).  Note that FF CO2 emissions are assumed to quadruple in the next 80 years, with atmospheric CO2 rising from 400 to 1000 ppm ( +150%).  Bear these suppositions in mind when considering the INMCM5 forecasts below.

Results [Continuing From Volodin and Gritsun]

Let us describe some simulation results of climate change in the 21st century. Figure 1 shows the change in the globally averaged surface air temperature with respect to the data of the corresponding historical experiment for 1850–1899. In the warmest SSP5-8.5 scenario (circles), the temperature rises by more than 4° by the end of the 21st century. In the SSP3-7.0 scenario (crosses), different members of the ensemble show warming by 3.4°–3.6°. In the SSP2-4.5 scenario (squares), the temperature increases by about 2.4°. According to the SSP1-2.6 scenario (triangles) , the maximal warming by ~1.7° occurs in the middle of the 21st century, and the temperature exceeds the preindustrial temperature by 1.4° by the end of the century.

[My comment: Note that the vertical scale starts with +1.0C as was seen in the historical experiment. Thus an anomaly of 1.4C by 2100 is an increase of only 0.4C, while the SSP2-4.5 result adds 1.4C to the present]. 

The results for other CMIP6 models have not yet been published in peer-reviewed journals. However, according to the preliminary analysis (see, e.g.  https://cmip6workshop19.sciencesconf.org/ data/Session1_PosterSlides.pdf, p.29), the INM-CM5 model shows the lowest temperature increase among the CMIP6 models considered for all the scenarios due to the minimal equilibrium sensitivity to the CO2 concentration doubling, which is ~2.1° for the current model version, like for the previous version, despite new condensation and cloud formation blocks. [For more on CMIP6 comparisons see post Climate Models: Good, Bad and Ugly]

The changes in the surface air temperature are similar for all scenarios; therefore, we analyze the difference between temperatures in 2071–2100 and 1981–2010 under the SSP5-8.5 and SSP1-2.6 scenarios (Fig. 2). The warming is maximal in the Arctic; it reaches 10° and 3°, respectively. Other features mainly correspond to CMIP5 data [8], including the INMCM4 model, which participates in the comparison. The warming on the continents of the Northern Hemisphere is about 2 times higher than the mean, and the warming in the Southern Hemisphere is noticeably less than in the Northern Hemisphere. The land surface is getting warmer than the ocean surface in all the scenarios except SSP1-2.6, because the greenhouse effect is expected to weaken in the second half of the 21st century in this scenario, and the higher heat capacity of the ocean prevents it from cooling as quickly as the land.

The changes in precipitation in December–February and June–August for the SSP3-7.0 scenario averaged over five members of the ensemble are shown in Fig. 4. All members of the ensemble show an increase in precipitation in the winter in a significant part of middle and high latitudes. In summer, the border between the increase and decrease in precipitation in Eurasia passes mainly around or to the north of 60°. In southern and central Europe, all members of the ensemble show a decrease in precipitation. Precipitation also increases in the region of the summer Asian monsoon, over the equatorial Pacific, due to a decrease in the upwelling and an increase in ocean surface temperature (OST). The distribution of changes in precipitation mainly corresponds to that given in [6, Fig. 12.22] for all CMIP5 models.

The change in the Arctic sea ice area in September, when the ocean ice cover is minimal over the year, is of interest. Figure 5 shows the sea ice area in September 2015–2019 to be 4–6 million km2 in all experiments, which corresponds to the estimate from observations in [11]. The Arctic sea ice does not completely melt in any of the experiments and under any scenario. However, according to [8, Figs. 12.28 and 12.31], many models participating in CMIP6, where the Arctic ice area is similar to that observed at the beginning of the 21st century, show the complete absence of ice by the end of the 21st century, especially under the RCP8.5 scenario, which is similar to SSP5-8.5.

The reason for these differences is the lower equilibrium sensitivity of the INM-CM5 model.

Note that the scatter of data between experiments under different scenarios in the first half of the 21st century is approximately the same as between different members of the ensemble under the SSP3-7.0 scenario and becomes larger only after 2070. The sea ice area values are sorted in accordance with the radiative forcing of the scenarios only after 2090. This indicates the large contribution of natural climate variability into the Arctic ice area. In the SSP1-2.6 experiment, the Arctic ice area at the end of the 21st century approximately corresponds to its area at the beginning of the experiment.

Climate changes can be also traced in the ocean circulation. Figure 6 shows the change in the 5-year averaged intensity of the Atlantic meridional circulation, defined as the maximum of the meridional streamfunction at 32° N. All experiments show a decrease in the intensity of meridional circulation in the 21st century and natural fluctuations against this decrease. The decrease is about 4.5–5 Sv for the SSP5-8.5 scenario, which is close to values obtained in the CMIP5 models [8, Fig. 12.35] under the RCP8.5 scenario. Under milder scenarios, the weakening of the meridional circulation is less pronounced. The reason for this weakening of the meridional circulation in the Atlantic, as far as we know, is not yet fully understood.

Conclusion

Numerical experiments have been carried out to reproduce climate changes in the 21st century according to four scenarios of the CMIP6 program [4, 5], including an ensemble of five experiments under the SSP3-7.0 scenario. The changes in the global mean surface temperature are analyzed. It is shown that the global warming predicted by the INM-CM5 model is the lowest among the currently published CMIP6 model data. The geographical distribution of changes in the temperature and precipitation is considered. According to the model, the temperature in the warmest summer month will increase faster than the summer average temperature in Russia.

None of the experiments show the complete melting of the Arctic ice cover by the end of the 21st century. Some changes in the ocean dynamics, including the flow velocity and the meridional stream function, are analyzed. The changes in the Hadley and Ferrel circulation in the atmosphere are considered.

Resources:

Climate Models: Good, Bad and Ugly

2018 Update: Best Climate Model INMCM5

Temperatures According to Climate Models

Links are provided at the end to previous posts describing climate models 4 and 5 from the Institute of Numerical Mathematics in Moscow, Russia.  Now we have forecasts for the 21st Century published for INM-CM5 at Izvestiya, Atmospheric and Oceanic Physics volume 56, pages218–228(July 7, 2020). The article is Simulation of Possible Future Climate Changes in the 21st Century in the INM-CM5 Climate Model by E. M. Volodin & A. S. Gritsun.  Excerpts are in italics with my bolds, along with a contextual comment.

Abstract

Climate changes in 2015–2100 have been simulated with the use of the INM-CM5 climate model following four scenarios: SSP1-2.6, SSP2-4.5, and SSP5-8.5 (single model runs) and SSP3-7.0 (an ensemble of five model runs). Changes in the global mean temperature and spatial distribution of temperature and precipitation are analyzed. The global warming predicted by the INM-CM5 model in the scenarios considered is smaller than that in other CMIP6 models. It is shown that the temperature in the hottest summer month can rise more quickly than the seasonal mean temperature in Russia. An analysis of a change in Arctic sea ice shows no complete Arctic summer ice melting in the 21st century under any model scenario. Changes in the meridional stream function in atmosphere and ocean are studied.

Overview

The climate is understood as the totality of statistical characteristics of the instantaneous states of the atmosphere, ocean, and other climate system components averaged over a long time period.

Therefore, we restrict ourselves to an analysis of some of the most important climate parameters, such as average temperature and precipitation. A more detailed analysis of individual aspects of climate change, such as changes in extreme weather and climate situations, will be the subject of another work. This study is not aimed at a full comparison with the results of other climate models, where calculations follow the same scenarios, since the results of other models have not yet been published in peer reviewed journals by the time of this writing.

The INM-CM5 climate model [1, 2] is used for the numerical experiments. It differs from the previous version, INMCM4, which was also used for experiments on reproducing climate change in the 21st century [3], in the following:

• an aerosol block has been added to the model, which allows inputting anthropogenic emissions of aerosols and their precursors;
• the concentrations and optical properties of aerosols are calculated, but not specified, like in the previous version;
• the parametrizations of cloud formation and condensation are changed in the atmospheric block;
• the upper boundary in the atmospheric block is raised from 30 to 60 km;
• the horizontal resolution in the ocean block is doubled along each coordinate; and,
• the software related to adaptation to massively parallel computers is improved, which allows the effective use a larger number of compute cores.

The model resolution in the atmospheric and aerosol blocks is 2° × 1.5° in longitude and latitude and 73 levels and, in the ocean, 0.5° × 0.25° and 40 levels. The calculations were performed at supercomputers of the Joint Supercomputer Center, Russian Academy of Sciences, and Moscow State University, with the use of 360 to 720 cores. The model calculated 6–10 years per 24 h in the above configuration.

Four scenarios were used to model the future climate: SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-5.8. The scenarios are described in [4]. The figure after the abbreviation SSP (Shared Socioeconomic Pathway) is the number of the mankind development path (see the values in [4]). The number after the dash means the radiation forcing (W m–2) in 2100 compared to the preindustrial level. Thus, the SSP1-2.6 scenario is the most moderate and assumes rapid actions which sharply limit and then almost completely stop anthropogenic emissions. Within this scenario, greenhouse gas concentrations are maximal in the middle of the 21st century and then slightly decrease by the end of the century. The SSP5-8.5 scenario is the warmest and implies the fastest climate change. The scenarios are recommended for use in the project on comparing CMIP6 (Coupled Model Intercomparison Project, Phase 6, [5]) climate models.  Each scenario includes the time series of:

• carbon dioxide, methane, nitrous oxide, and ozone concentrations;
• emissions of anthropogenic aerosols and their precursors;
• the concentration of volcanic sulfate aerosol; and
• the solar constant. 

One model experiment was carried out for each of the above scenarios. It began at the beginning of 2015 and ended at the end of 2100. The initial state was taken from the so-called historical experiment with the same model, where climate changes were simulated for 1850–2014, and all impacts on the climate system were set according to observations. The results of the ensemble of historical experiments with the model under consideration are given in [6, 7]. For the SSP3-7.0 scenario, five model runs was performed differing in the initial data taken from different historical experiments. The ensemble of numerical experiments is required to increase the statistical confidence of conclusions about climate changes.

[My Contextual Comment inserted Prior to Consideration of Results]

Firstly, the INM-CM5 historical experiment can be read in detail by following a linked post below, but this graphic summarizes the model hindcasting of past temperatures (GMT) compared to HadCrutv4.

Secondly, the scenarios are important to understand since they stipulate data inputs the model must accept as conditions for producing forecasts according to a particular scenario (set of assumptions).  The document with complete details referenced as [4] is The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6.

All the details are written there but one diagram suggests the implications for the results described below.

As shown, the SSP1-26 is virtually the same scenario as the former RCP2.6, while SSP5-85 is virtually the same as RCP8.5, the wildly improbable scenario (impossible according to some analysts).  Note that FF CO2 emissions are assumed to quadruple in the next 80 years, with atmospheric CO2 rising from 400 to 1000 ppm ( +150%).  Bear these suppositions in mind when considering the INMCM5 forecasts below.

Results [Continuing From Volodin and Gritsun]

Let us describe some simulation results of climate change in the 21st century. Figure 1 shows the change in the globally averaged surface air temperature with respect to the data of the corresponding historical experiment for 1850–1899. In the warmest SSP5-8.5 scenario, the temperature rises by more than 4° by the end of the 21st century. In the SSP3-7.0 scenario, different members of the ensemble show warming by 3.4°–3.6°. In the SSP2-4.5 scenario, the temperature increases by about 2.4°. According to the SSP1-2.6 scenario, the maximal warming by ~1.7° occurs in the middle of the 21st century, and the temperature exceeds the preindustrial temperature by 1.4° by the end of the century.

[My comment: Note that the vertical scale starts with +1.0C as was seen in the historical experiment. Thus an anomaly of 1.4C by 2100 is an increase of only 0.4C, while the SSP2-4.5 result adds 1.4C to the present].

The results for other CMIP6 models have not yet been published in peer-reviewed journals. However, according to the preliminary analysis (see, e.g.  https://cmip6workshop19.sciencesconf.org/ data/Session1_PosterSlides.pdf, p.29), the INM-CM5 model shows the lowest temperature increase among the CMIP6 models considered for all the scenarios due to the minimal equilibrium sensitivity to the CO2 concentration doubling, which is ~2.1° for the current model version, like for the previous version, despite new condensation and cloud formation blocks. [For more on CMIP6 comparisons see post Climate Models: Good, Bad and Ugly]

The changes in the surface air temperature are similar for all scenarios; therefore, we analyze the difference between temperatures in 2071–2100 and 1981–2010 under the SSP5-8.5 and SSP1-2.6 scenarios (Fig. 2). The warming is maximal in the Arctic; it reaches 10° and 3°, respectively. Other features mainly correspond to CMIP5 data [8], including the INMCM4 model, which participates in the comparison. The warming on the continents of the Northern Hemisphere is about 2 times higher than the mean, and the warming in the Southern Hemisphere is noticeably less than in the Northern Hemisphere. The land surface is getting warmer than the ocean surface in all the scenarios except SSP1-2.6, because the greenhouse effect is expected to weaken in the second half of the 21st century in this scenario, and the higher heat capacity of the ocean prevents it from cooling as quickly as the land.

The changes in precipitation in December–February and June–August for the SSP3-7.0 scenario averaged over five members of the ensemble are shown in Fig. 4. All members of the ensemble show an increase in precipitation in the winter in a significant part of middle and high latitudes. In summer, the border between the increase and decrease in precipitation in Eurasia passes mainly around or to the north of 60°. In southern and central Europe, all members of the ensemble show a decrease in precipitation. Precipitation also increases in the region of the summer Asian monsoon, over the equatorial Pacific, due to a decrease in the upwelling and an increase in ocean surface temperature (OST). The distribution of changes in precipitation mainly corresponds to that given in [6, Fig. 12.22] for all CMIP5 models.

The change in the Arctic sea ice area in September, when the ocean ice cover is minimal over the year, is of interest. Figure 5 shows the sea ice area in September 2015–2019 to be 4–6 million km2 in all experiments, which corresponds to the estimate from observations in [11]. The Arctic sea ice does not completely melt in any of the experiments and under any scenario. However, according to [8, Figs. 12.28 and 12.31], many models participating in CMIP6, where the Arctic ice area is similar to that observed at the beginning of the 21st century, show the complete absence of ice by the end of the 21st century, especially under the RCP8.5 scenario, which is similar to SSP5-8.5.

The reason for these differences is the lower equilibrium sensitivity of the INM-CM5 model.

Note that the scatter of data between experiments under different scenarios in the first half of the 21st century is approximately the same as between different members of the ensemble under the SSP3-7.0 scenario and becomes larger only after 2070. The sea ice area values are sorted in accordance with the radiative forcing of the scenarios only after 2090. This indicates the large contribution of natural climate variability into the Arctic ice area. In the SSP1-2.6 experiment, the Arctic ice area at the end of the 21st century approximately corresponds to its area at the beginning of the experiment.

Climate changes can be also traced in the ocean circulation. Figure 6 shows the change in the 5-year averaged intensity of the Atlantic meridional circulation, defined as the maximum of the meridional streamfunction at 32° N. All experiments show a decrease in the intensity of meridional circulation in the 21st century and natural fluctuations against this decrease. The decrease is about 4.5–5 Sv for the SSP5-8.5 scenario, which is close to values obtained in the CMIP5 models [8, Fig. 12.35] under the RCP8.5 scenario. Under milder scenarios, the weakening of the meridional circulation is less pronounced. The reason for this weakening of the meridional circulation in the Atlantic, as far as we know, is not yet fully understood.

Conclusion

Numerical experiments have been carried out to reproduce climate changes in the 21st century according to four scenarios of the CMIP6 program [4, 5], including an ensemble of five experiments under the SSP3-7.0 scenario. The changes in the global mean surface temperature are analyzed. It is shown that the global warming predicted by the INM-CM5 model is the lowest among the currently published CMIP6 model data. The geographical distribution of changes in the temperature and precipitation is considered. According to the model, the temperature in the warmest summer month will increase faster than the summer average temperature in Russia.

None of the experiments show the complete melting of the Arctic ice cover by the end of the 21st century. Some changes in the ocean dynamics, including the flow velocity and the meridional stream function, are analyzed. The changes in the Hadley and Ferrel circulation in the atmosphere are considered.

Resources:

Climate Models: Good, Bad and Ugly

2018 Update: Best Climate Model INMCM5

Temperatures According to Climate Models