RealClimate logo

Climate Change and Extreme Summer Weather Events – The Future is still in Our Hands

Summer 2018 saw an unprecedented spate of extreme weather events, from the floods in Japan, to the record heat waves across North America, Europe and Asia, to wildfires that threatened Greece and even parts of the Arctic. The heat and drought in the western U.S. culminated in the worst California wildfire on record. This is the face of climate change, I commented at the time.

Some of the connections with climate change here are pretty straightforward. One of the simplest relationships in all of atmospheric science tells us that the atmosphere holds exponentially more moisture as temperatures increase. Increased moisture means potentially for greater amounts of rainfall in short periods of time, i.e. worse floods. The same thermodynamic relationship, ironically, also explains why soils evaporate exponentially more moisture as ground temperatures increase, favoring more extreme drought in many regions. Summer heat waves increase in frequency and intensity with even modest (e.g. the observed roughly 2F) overall warming owing to the behavior of the positive “tail” of the bell curve when you shift the center of the curve even a small amount. Combine extreme heat and drought and you get more massive, faster-spreading wildfires. It’s not rocket science.

But there is more to the story. Because what made these events so devastating was not just the extreme nature of the meteorological episodes but their persistence. When a low-pressure center stalls and lingers over the same location for days at a time, you get record accumulation of rainfall and unprecedented flooding. That’s what happened with Hurricane Harvey last year and Hurricane Florence this year. It is also what happened with the floods in Japan earlier this summer and the record summer rainfall we experienced this summer here in Pennsylvania. Conversely, when a high-pressure center stalls over the same location, as happened in California, Europe, Asia and even up into the European Arctic this past summer, you get record heat, drought and wildfires.

Scientists such as Jennifer Francis have linked climate change to an increase in extreme weather events, especially during the winter season when the jet stream and “polar vortex” are relatively strong and energetic. The northern hemisphere jet stream owes its existence to the steep contrast in temperature in the middle latitudes (centered around 45N) between the warm equator and the cold Arctic. Since the Arctic is warming faster than the rest of the planet due to the melting of ice and other factors that amplify polar warming, that contrast is decreasing and the jet stream is getting slower. Just like a river traveling over gently sloping territory tends to exhibit wide meanders as it snakes its way toward the ocean, so too do the eastward-migrating wiggles in the jet stream (known as Rossby waves) tend to get larger in amplitude when the temperature contrast decreases. The larger the wiggles in the jet stream the more extreme the weather, with the peaks corresponding to high pressure at the surface and the troughs low pressure at the surface. The slower the jet stream, the longer these extremes in weather linger in the same locations, giving us more persistent weather extremes.

Something else happens in addition during summer, when the poleward temperature contrast is especially weak. The atmosphere can behave like a “wave guide”, trapping the shorter wavelength Rossby waves (those that that can fit 6 to 8 full wavelengths in a complete circuit around the Northern Hemisphere) to a relatively narrow range of latitudes centered in the mid-latitudes, preventing them from radiating energy away toward lower and higher latitudes. That allows the generally weak disturbances in this wavelength range to intensify through the physical process of resonance, yielding very large peaks and troughs at the sub-continental scale, i.e. unusually extreme regional weather anomalies. The phenomenon is known as Quasi-Resonant Amplification or “QRA”, and (see Figure below).

Many of the most damaging extreme summer weather events in recent decades have been associated with QRA, including the 2003 European heatwave, the 2010 Russian heatwave and wildfires and Pakistan floods (see below), and the 2011 Texas/Oklahoma droughts. More recent examples include the 2013 European floods, the 2015 California wildfires, the 2016 Alberta wildfires and, indeed, the unprecedented array of extreme summer weather events we witnessed this past summer.

The increase in the frequency of these events over time is seen to coincide with an index of Arctic amplification (the difference between warming in the Arctic and the rest of the Northern Hemisphere), suggestive of a connection (see Figure below).

Last year we (me and a team of collaborators including RealClimate colleague Stefan Rahmstorf) published an article in the Nature journal Scientific Reports demonstrating that the same pattern of amplified Arctic warming (“Arctic Amplification”) that is slowing down the jet stream is indeed also increasing the frequency of QRA episodes. That means regional weather extremes that persist longer during summer when the jet stream is already at its weakest. Based on an analysis of climate observations and historical climate simulations, we concluded that the “signal” of human influence on QRA has likely emerged from the “noise” of natural variability over the past decade and a half. In summer 2018, I would argue, that signal was no longer subtle. It played out in real time on our television screens and newspaper headlines in the form of an unprecedented hemisphere-wide pattern of extreme floods, droughts, heat waves and wildfires.

In a follow-up article just published in the AAAS journal Science Advances, we look at future projections of QRA using state-of-the-art climate model simulations. It is important to note that that one cannot directly analyze QRA behavior in a climate model simulation for technical reasons. Most climate models are run at grid resolutions of a degree in latitude or more. The physics that characterizes QRA behavior of Rossby Waves faces a stiff challenge when it comes to climate models because it involves the second mathematical derivative of the jet stream wind with respect to latitude. Errors increase dramatically when you calculate a numerical first derivative from gridded fields and even more so when you calculate a second derivative. Our calculations show that the critical term mentioned above suffers from an average climate model error of more than 300% relative to observations. By contrast, the average error of the models is less than a percent when it comes to latitudinal temperature averages and still only about 30% when it comes to the latitudinal derivative of temperature.

That last quantity is especially relevant because QRA events have been shown to have a well-defined signature in terms of the latitudinal variation in temperature in the lower atmosphere. Through a well-established meteorological relationship known as the thermal wind, the magnitude of the jet stream winds is in fact largely determined by the average of that quantity over the lower atmosphere. And as we have seen above, this quantity is well captured by the models (in large part because the change in temperature with latitude and how it responds to increasing greenhouse gas concentrations depends on physics that are well understood and well represented by the climate models).

These findings, incidentally have broader implications. First of all, climate model-based studies used to assess the degree to which current extreme weather events can be attributed to climate change are likely underestimating the climate change influence. One model-based study for example suggested that climate change only doubled the likelihood of the extreme European heat wave this summer. As I commented at the time, that estimate is likely too low for it doesn’t account for the role that we happen to know, in this case, that QRA played in that event. Similarly, climate models used to project future changes in extreme weather behavior likely underestimate the impact that future climate changes could have on the incidence of persistent summer weather extremes like those we witnessed this past summer.

So what does our study have to say about the future? We find that the incidence of QRA events would likely continue to increase at the same rate it has in recent decades if we continue to simply add carbon dioxide to the atmosphere. But there’s a catch: The future emissions scenarios used in making future climate projections must also account for factors other than greenhouse gases. Historically, for example, the use of old coal technology that predates the clean air acts produced sulphur dioxide gas which escapes into the atmosphere where it reacts with other atmospheric constituents to form what are known as aerosols.

These aerosols caused acid rain and other environmental problems in the U.S. before factories in the 1970s were required to install “scrubbers” to remove the sulphur dioxide before it leaves factory smokestacks. These aerosols also reflect incoming sunlight and so have a cooling effect on the surface in the industrial middle-latitudes where they are produced. Some countries, like China, are still engaged in the older, dirtier-form of coal burning. If we continue with business-as-usual burning of fossil fuels, but countries like China transition to more modern “cleaner” coal burning to avoid air pollution problems, we are likely to see a substantial drop in aerosols over the next half century. Such an assumption is made in the Intergovernmental Panel on Climate Change (IPCC)’s “RCP 8.5” scenario—basically, a “business as usual” future emissions scenario which results in more than a tripling of carbon dioxide concentrations relative to pre-industrial levels (280 parts per million) and roughly 4-5C (7-9F) of planetary warming by the end of the century.

As a result, the projected disappearance of cooling aerosols in the decades ahead produces an especially large amount of warming in middle-latitudes in summer (when there is the most incoming sunlight to begin with, and, thus, the most sunlight to reflect back to space). Averaged across the various IPCC climate models there is even more warming in mid-latitudes than in the Arctic—in other words, the opposite of Arctic Amplification i.e. Arctic De-amplification (see Figure below). Later in the century after the aerosols disappear greenhouse warming once again dominates and we again see an increase in QRA events.

So, is there any hope to avoid future summers like the summer of 2018? Probably not. But in the scenario where we rapidly move away from fossil fuels and stabilize greenhouse gas concentrations below 450 parts per million, giving us a roughly 50% chance of averting 2C/3.6F planetary warming (the so-called “RCP 2.6” IPCC scenario) we find that the frequency of QRA events remains roughly constant at current levels.

While we will presumably have to contend with many more summers like 2018 in the future, we could likely prevent any further increase in persistent summer weather extremes. In other words, the future is still very much in our hands when it comes to dangerous and damaging summer weather extremes. It’s simply a matter of our willpower to transition quickly from fossil fuels to renewable energy.

Cracking the Climate Change Case

I have an op-ed in the New York Times this week:

How Scientists Cracked the Climate Change Case
The biggest crime scene on the planet is the planet. We know the earth is warming, but who or what is causing it?
Emilia Miękisz

Many of you will recognise the metaphor from previous Realclimate pieces (this is earliest one I think, from 2007), and indeed, the working title was “CSI: Planet Earth”. The process description and conclusions are drawn from multiple sources on the attribution of recent climate trends (here, here etc.), as well the data visualization for surface temperature trends at Bloomberg News.

There have been many comments about this on Twitter – most appreciative, some expected, and a few interesting. The expected criticisms come from people who mostly appear not to have read the piece at all (“Climate has changed before!” – a claim that no-one disputes), and a lot of pointless counter-arguments by assertion. Of the more interesting comment threads, was one started by Ted Nordhaus who asked

My response is basically that it might be old hat for him (and maybe many readers here), but I am constantly surprised at the number of people – even those concerned about climate – who are unaware of how we do attribution and how solid the science behind the IPCC statements is. And judging by many of the comments, it certainly isn’t the case that these pieces are only read by the already convinced. But asking how many people are helped to be persuaded by articles like this is a valid question, and I don’t really know the answer. Anyone?

Pre-industrial anthropogenic CO2 emissions: How large?

Filed under: — mike @ 11 October 2018

Guest article by William Ruddiman

Fifteen years after publication of Ruddiman (2003), the early anthropogenic hypothesis is still debated, with relevant evidence from many disciplines continuing to emerge. Recent findings summarized here lend support to the claim that greenhouse-gas emissions from early agriculture (before 1850) were large enough to alter atmospheric composition and global climate substantially.

Marine isotopic stage (MIS) 19 is the closest orbital analog to the current MIS 1 interglaciation (Tzedakis et al., 2012), with similarly small changes in precession (εsinω) and nearly synchronous peaks in sin and obliquity (Fig. 1a, b). MIS 11 was once claimed to be the closest MIS 1 analog (for example, Broecker and Stocker, 2006), but that claim is now rejected because obliquity and precession peaks in MIS 11 were far offset.

Figure 1 Comparison of (a) obliquity and (b) precession (εsinω) trends during MIS19, (green), MIS11 (black) and MIS1 (red). Based on Tzedakis et al. (2012). (c) CO2 trends during MIS19 (black) and MIS1 (red). CO2 data for MIS 19 are from Dome C (Bereiter et al. 2015). CO2 data for MIS 1 are from Law Dome (MacFarling Meure et al. 2006) and Dome C (Monnin et al. 2001, 2004) for MIS1.


With MIS 11 eliminated as an analog, the focus is on MIS 19. The CO2 signals early in MIS 1 and MIS 19 (Fig. 1c) reached nearly identical peaks of 270 and 269 ppm, after which the MIS 1 value fell for 4000 years but then rose by 20 ppm to a late pre-industrial 280-285 ppm. In contrast, the MIS 19 CO2 trend continued downward for more than 10,000 years to 245-250 ppm by the time equivalent to the present day. This value is consistent with the 240-245 ppm level proposed in the early anthropogenic hypothesis for a natural Holocene world (with no human overprint). The 35-ppm difference between the two interglaciations is close to the 40-ppm Holocene anomaly inferred by Ruddiman (2003).

A GCM simulation of the MIS 19 time equivalent to today by Vavrus et al. (2018) indicates that the low CO2 values would have caused year-round snow cover (indicative of incipient glaciation) in the Canadian Archipelago and over Baffin Island (an area roughly the size of Greenland), as well as other Arctic regions (see also Ganopolski et al., 2014).

Ruddiman (2003) estimated pre-industrial carbon emissions of 300-320 Gt, based on a back-of-the-envelope compilation of the incomplete forest clearance histories then available (Table 1). [One Gt is one billion tons]. That estimate was for a while rejected as too high by a factor of 5 to 10 (Joos et al., 2004; Pongratz et al., 2008; Stocker et al., 2011. However, Kaplan et al. (2011) found that those estimates had been biased downward because they assumed much smaller early per-capita clearance than the large amounts shown by actual historical data. Those estimates also ignored areas that had been cleared and were not in active agricultural use, but had not yet reforested. Adjusting for these factors, Kaplan and colleagues estimated pre-industrial emissions of 343 GtC.

Erb et al. (2017) averaged 7 estimates of the amount of carbon that would currently be stored in Earth’s potential natural vegetation had there been no human activities (910 GtC) compared to the 460 GtC carbon actually stored there today. They attributed the difference of 450 GtC to cumulative vegetation removal by humans (mostly deforestation). With ~140 GtC of clearance having occurred during the industrial era, that left an estimated 310 GtC as the total removed and emitted to the atmosphere during pre-industrial time. In a similar analysis, Lorenz and Lal (2018) estimated pre-industrial carbon emissions of ‘up to’ 357 GtC.

Studies in other disciplines have begun adding direct ground-truth evidence about early clearance. Analyses of pollen in hundreds of European lake cores (Fyfe et al., 2014; Roberts et al, 2018) show that forest vegetation began to decrease after 6000 years ago and reached near-modern levels before the start of the industrial era (Fig. 2). In China, compilations of over 50,000 archaeological sites by Li et al. (2009) and Hosner et al. (2016) show major increases of farming settlements in previously forested areas beginning 7,000 years ago. These extensive compilations support the above estimates of large early anthropogenic clearance and C emissions.

Figure 2. Evidence of early forest clearance in Europe. (A) Locations of cores in the European pollen database. Cores used for pollen summary in B are shown in red (Fyfe et al., 2015). (B) Changes in forest, open, and semi-open (mixed forest and open) vegetation plotted as ‘pseudobiome’ sums.


As this wide-ranging multi-disciplinary evidence has emerged, some scientists continue to reject the early anthropogenic hypothesis. Most of the opposition is based on a geochemical index (δ13CO2) measured in CO2 contained in air bubbles trapped in ice cores. The δ13CO2 index shows the relative balance through time between the amount of 12C-rich terrestrial carbon from the land and 13C-neutral carbon from the ocean. The small 13C decrease in atmospheric CO2 during the last 7000 years has been interpreted as indicating minimal input of 12C-rich terrestrial carbon during that time (Broecker and Stocker, 2006; Elsig et al., 2009). In a July 20, 2018 post, Jeff Severinghaus estimated the early human contribution to the observed CO2 rise as “1 to 2 ppm at the most”, or just 5-10% of the recent estimates reviewed in Table 1.

Other scientists (Stocker et al., 2018; Ruddiman et al., 2016) have pointed out that the δ13CO2 index cannot be used to isolate the amount of deforestation carbon unless all significant carbon sources and sinks are well constrained. The compilation by Yu (2011) indicating that ~300 Gt of terrestrial (12C-rich) carbon were buried in boreal peats during the last 7000 years shows that this constraint had not been satisfied in previous studies. Burial of ~300 GtC in boreal peats requires a counter-balancing emission of more than 300 GtC of terrestrial carbon during the last 7000 years, and the discussion above summarizes evidence that pre-industrial deforestation can fill that deficit. Even now, however, carbon exchanges (whether sources or sinks) in non-peat permafrost areas and in river floodplains and deltas during the last 7000 years remain poorly known.

Scientists trying to make up their minds about this still-ongoing debate can now weigh wide-ranging multi-disciplinary evidence for large early forest clearance against reliance on the as-yet poorly constrained δ13CO2 index.


Bereiter, B., S. Eggleston, J. Schmitt, C. Nehrbass-Ahles, T. F. Stocker, et al. (2015), Revision of the EPICA Dome C CO2 record from 800 to 600 kyr before present, Geophys. Res. Lett., 42, 542–549.

Broecker, W. S. and T. L. Stocker (2006), The Holocene CO2 rise: Anthropogenic or natural? EOS Trans. Amer. Geophysical Union 87, 27.

Erb, K.-H., T. Kastner, C. Plutzar, C., A. L. S Bais, N. Carvalhai., et al. (2018), Unexpectedly large impact of forest management on global vegetation biomass. Nature 553, 73-76.

Elsig J., J. Schmitt, D. Leuenberger, R. Schneider, M. Eyer, et al. (2009), Stable isotope constraints on Holocene carbon cycle changes from an Antarctic ice core. Nature 461, 507-510.

Fyfe, R. M., J. Woodbridge, and N. Roberts (2015), From forest to farmland: pollen-inferred land cover changes across Europe using the pseudobiomization approach. Global Change Biology 20, 1197-1212.

Ganopolski, A., R. Winkelmann and H. J. Schellenhuber, (2014), Critical insolation-CO2 relation for diagnosing past and future glacial inception. Nature 529, 200-203.

Hosner, D., M. Wagner, P. E. Tarasov, X. Chen, and C. Leipe (2016), Spatiotemporal distribution patterns of archaeological sites in China during the Neolithic and Bronze Age: An overview. The Holocene 26, 1576-1583.

Joos F, Gerber S, Prentice IC, et al. (2004) Transient simulations of Holocene atmospheric carbon dioxide and terrestrial carbon since the last glacial maximum. Global Biogeochemical Cycles 18. DOI: 10.1029/2003GB002156.

Kaplan J. O, K. M. Krumhardt, E. C. Ellis, W. F. Ruddiman, C. Lemmen, et al. Goldewijk (2011), Holocene carbon emissions as a result of anthropogenic land cover change. The Holocene 21, 775-792.

Li, X., J. Dodson, J. Zhou, and X. Zhou (2008), Increases of population and expansion of rice agriculture in Asia, and anthropogenic methane emissions since 5000 BP. Quat. Int. 202, 41-50.

Lorenz, K. and R. Lal (2018), Agricultural land use and the global carbon cycle. In: Carbon sequestration in agricultural systems, p. 1-37.

MacFarling Meure, C., D. Etheridge, C. Trudinger, P. Steele, R. Langenfelds, et al. (2006), Law Dome CO2, CH4 and N2O ice core records extended to 2000 years BP. Geophys. Res. Lett., 33, L14810, doi:10.1029/2006GL026152.

Monnin E., A. Indermühle, A. Dällenbach, J. Flückinger, B. Stauffer, et al. (2001), Atmospheric CO¬¬2 concentrations over the Last Glacial Termination. Science, 291, 112-114.

Pongratz, J., C. Reick, T. Raddatz, and M. A. Claussen (2008), A reconstruction of global agricultural areas and land cover for the last millennium. Global Geochemical Cycles 22, GB3018m doi:10.1029/2008GLO36394.

Roberts N, R. M. Fyfe, J. Woodbridge, et al. (2018), Europe’s forests: A pollen-based synthesis for the last 11,000 years. Nature Scientific Reports. DOI: 10.1038/s41598-017-18646-7
Ruddiman, W. F. (2003), The anthropogenic greenhouse era began thousands of years ago. Climatic Change 61, 261-293.

Ruddiman, W. F., D. Q. Fuller, J. E Kutzbach, P. C. Tzedakis, J. O. Kaplan et al. (2016), Late Holocene climate: Natural or anthropogenic? Rev. of Geophys. 54, 93-118.

Stocker, B. D., K. Strassmann, and F. Joos (2011), Sensitivity of Holocene atmospheric CO2 and the modern carbon budget to early human land use: analyses with a process-base model. Biogeosciences 8, 69-88.

Stocker, B.D., Z. Yu, and F. Joos (2018), Constraining CO2 emissions from different Holocene land-use histories: does the carbon budget add up? PAGES 26, 6-7.

Tzedakis, P. C., J. E. T. Channell, D. A. Hodell, H. F. Kleiven, and L. K. Skinner (2012), Determining the length of the current interglacial. Nature Geoscience 5, 138-141.

Vavrus, S. J., F. He, J. E. Kutzbach, W. F. Ruddiman, and P. C. Tzedakis (2018), Glacial inception in marine isotope stage 19: An orbital analog for a
natural Holocene. Nature Scientific Reports 81, doi:10.1038/s41598-018-28419-5.

30 years after Hansen’s testimony

Filed under: — gavin @ 21 June 2018

“The greenhouse effect is here.”
– Jim Hansen, 23rd June 1988, Senate Testimony

The first transient climate projections using GCMs are 30 years old this year, and they have stood up remarkably well.

We’ve looked at the skill in the Hansen et al (1988) (pdf) simulations before (back in 2008), and we said at the time that the simulations were skillful and that differences from observations would be clearer with a decade or two’s more data. Well, another decade has passed!

More »


  1. J. Hansen, I. Fung, A. Lacis, D. Rind, S. Lebedeff, R. Ruedy, G. Russell, and P. Stone, "Global climate changes as forecast by Goddard Institute for Space Studies three-dimensional model", Journal of Geophysical Research, vol. 93, pp. 9341, 1988.

The Alsup Aftermath

The presentations from the Climate Science tutorial last month have all been posted (links below), and Myles Allen (the first presenter for the plaintiffs) gives his impression of the events.
More »