RealClimate

Climate science from climate scientists...

You are here: Home / Archives for Featured Story

Featured Story

The AMOC: tipping this century, or not?

by

A few weeks ago, a study by Copenhagen University researchers Peter and Susanne Ditlevsen concluded that the Atlantic Meridional Overturning Circulation (AMOC) is likely to pass a tipping point already this century, most probably around mid-century. Given the catastrophic consequences of an AMOC breakdown, the study made quite a few headlines but also met some skepticism. Now that the dust has settled, here some thoughts on the criticisms that have been raised about this study.

I’ve seen two main arguments there.

1. Do the data used really describe changes in AMOC?

We have direct AMOC measurements only since 2004, a time span too short for this type of study. So the Ditlevsens used sea surface temperatures (SST) in a region between the tip of Greenland and Britain as an indicator, based on Caesar et al. 2018 (PDF; I’m a coauthor on that paper). The basic idea starts with the observation that this region is far warmer than what is normal for that latitude, because the AMOC delivers a huge amount of heat into the area. The following chart which I made 25 years ago illustrates this.

Temperature deviation relative to the average along each latitude circle (i.e. the zonal mean). The northern Atlantic region air temperature is a lot too warm for its latitude, which (in models) largely goes away when the AMOC is stopped. From Rahmstorf and Ganopolski 1999.

If the AMOC weakens, this region will cool. And in fact it is cooling – it’s the only region on Earth which has cooled since preindustrial times. This is commonly referred to as ‘warming hole’ or ‘cold blob’.

We argued in Caesar et al. that the sea surface temperature there in winter is a good index of AMOC strength, based on a high-resolution climate model. (Not in summer when the ocean is covered by a shallow surface mixed layer heated by the sun and highly dependent on weather conditions.) We checked this across other climate models and found that our AMOC index (i.e. based on SST in the ‘cold blob’ region) and the actual AMOC slowdown correlated highly there (correlation coefficient R=0.95).

There are some other indicators, either using measured ocean salinities or using various types of proxy data from sediment cores, e.g. sediment grain sizes at the ocean bottom as indicators of flow speed of the deep southward AMOC branch. The key point to me is: these different indicators provide rather consistent AMOC reconstructions, as we showed in Caesar et al. 2021. The sediment data go back further in time but are likely not as reliable and don’t reach up to the present.

For recent decades there are potentially better approaches like ocean state estimates, and those are also consistent with the SST fingerprint – but these don’t go back far enough in time for the Ditlevsen type of study. The next graph shows a comparison of different reconstructions for the relevant time period used in the Ditlevsen study.

A comparison of direct observational AMOC data (RAPID) and two recent reconstructions to both the SST-based AMOC index (blue, used by the Ditlevsens) and two paleo-proxies that extend into the twenty-first century: the sortable-silt data and the marine productivity data. From Caesar et al. 2022.

Reconstructions based on salinity may also be good but they depend on precipitation, a notoriously variable quantity so it is rather doubtful whether analysing variance of salinity is doing any better than the SST signal.

The argument has been made that the ‘cold blob’ might not be caused by an AMOC decline but by heat loss at the ocean surface. That’s easy to check: if that were the case, then cooling in the area would be linked to increased heat loss at the surface. But if the AMOC is the culprit, then less heat should be lost, as a cooler ocean surface due to reduced ocean heat transport will lose less heat. The reanalysis data show the latter is the case.

This was shown by Halldór Björnsson of the Icelandic weather service and presented at the Arctic Circle conference 2016. I discussed this here in 2016 and also in my 2018 RealClimate article “If you doubt that the AMOC has weakened, read this”, together with possible other alternative explanations of the ‘cold blob’. We have recently repeated Halldór’s analysis at PIK and got the same results.

My conclusion: for the past century or so the SST data are probably the best AMOC indicator we have, and I don’t see concrete evidence suggesting that it’s unreliable.

2. The Ditlevsen study assumes that the AMOC follows a quadratic curve when approaching the tipping point.

That’s a more technical criticism. Their assumption follows from Stommel’s 1961 simple model of the AMOC tipping point. It results from the basic idea that (a) AMOC changes are proportional to density changes, and (b) the density change results from a balance between freshwater input and AMOC salt transport to the deep water formation (i.e. ‘cold blob’) region. Combined, these two assumptions lead to a quadratic equation.

These are very plausible basic assumptions, albeit using a linear equation of state, but we all know you can linearize things around a given point to get a first-order estimate. The argument that this is “too simple” doesn’t mean it’s wrong; rather this is correct at least to first order.

In a 1996 study I compared the results of a quadratic box model response to a fully-fledged 3D primitive equation ocean circulation model with nonlinear equation of state, the MOM model of the Geophysical Fluid Dynamics Lab in Princeton. It looks like this.

The AMOC strength (vertical axis) is shown as it depends on freshwater input (rain, meltwater) into the northern Atlantic. The box model equilibrium is shown as dotted parabola, the tipping point is S. By global warming we move from a past equilibrium toward the right – the box model run is the dashed line, the global ocean circulation model run is the solid line. Relevant is the upper branch, moving towards the right approaching the tipping point. From Rahmstorf (1996, PDF).

You can’t get a much better fit than that. A similar quadratic shape has also been found by Henk Dijkstra’s group at Utrecht University in a state-of-the-art global climate model, the CESM model (yet to be published). I have not seen any concrete evidence by the critics suggesting the shape may not be quadratic; that seems to be a purely hypothetical possibility. Also, if it is not exactly quadratic, the stated uncertainty range will be larger but it doesn’t fundamentally change the result.

What does it all mean?

An AMOC collapse would be a massive, planetary-scale disaster. Some of the consequences: Cooling and increased storminess in northwestern Europe, major additional sea level rise especially along the American Atlantic coast, a southward shift of tropical rainfall belts (causing drought in some regions and flooding in others), reduced ocean carbon dioxide uptake, greatly reduced oxygen supply to the deep ocean, likely ecosystem collapse in the northern Atlantic, and others. Check out the OECD report Climate Tipping Points which is well worth reading, and the maps below. You really want to prevent this from happening.

A figure from the recent OECD report Climate Tipping Points, showing how an AMOC shutdown after 2.5 °C global warming would change temperature (left) and precipitation (right) around the world.

We know from paleoclimatic data that there have been a number of drastic, rapid climate changes with focal point in the North Atlantic due to abrupt AMOC changes, apparently after the AMOC passed a tipping point. They are known as Heinrich events and Dansgaard-Oeschger events, see my review in Nature (pdf).

The point: it is a risk we should keep to an absolute minimum.

In other words: we are talking about risk analysis and disaster prevention. This is not about being 100% sure that the AMOC will pass its tipping point this century; it is that we’d like to be 100% sure that it won’t. Even if there were just (say) a 40% chance that the Ditlevsen study is correct in the tipping point being reached between 2025 and 2095, that’s a major change to the previous IPCC assessment that the risk is less than 10%. Even a <10% chance as of IPCC (for which there is only “medium confidence” that it’s so small) is in my view a massive concern. That concern has increased greatly with the Ditlevsen study – that is the point, and not whether it’s 100% correct and certain.

Would you live in a village below a dammed lake if you’re told there is a one in ten chance that one day the dam will break and much of the village will be washed away? Would you say: “Not to worry, that’s 90 % chance it won’t happen?” Or would you demand action by the authorities to reduce the risk? What if a new study appears, experienced scientists, reputable journal, that says it is nearly certain that the dam will break, the question is only when? Would you demand immediate attention to mitigate this danger, or would you say: “Oh well, some have questioned whether the assumptions of this study are entirely correct. Let’s just assume it is wrong”?

For the AMOC (and other climate tipping points), the only action we can take to minimise the risk is to get out of fossil fuels and stop deforestation as fast as possible. One major assumption of the Ditlevsen study is that global warming continues as in past decades. That is in our hands – or more precisely, that of our governments and powerful corporations. In 2022, the G20 governments alone subsidised fossil fuel use with 1.4 trillion dollars, up by 475% above the previous year. They aren’t trying to end fossil fuels.

Yet, as soon as we reach zero emissions, global warming will stop within years, and the sooner this happens the smaller the risk of passing tipping points. It also minimises lots of other losses, damages and human suffering from “regular” global warming impacts, which are already happening all around us even without passing major climate tipping points.

Links

For more on this, see my long TwiX thread with many images from relevant studies.

What is happening in the Atlantic Ocean to the AMOC?

If you doubt that the AMOC has weakened, read this

AMOC slowdown: Connecting the dots

And for even more, just enter “AMOC” into the search field of this blog!

What is happening in the Atlantic Ocean to the AMOC?

by

For various reasons I’m motivated to provide an update on my current thinking regarding the slowdown and tipping point of the Atlantic Meridional Overturning Circulation (AMOC). I attended a two-day AMOC session at the IUGG Conference the week before last, there’s been interesting new papers, and in the light of that I have been changing my views somewhat. Here’s ten points, starting from the very basics, so you can easily jump to the aspects that interest you.

Figure 1. A very rough schematic of the AMOC: warm northward flow near the surface, deep-water formation, deep southward return flow in 2000 – 3000 meters depth. In the background the observed sea surface temperature (SST) trend since 1993 from the Copernicus satellite service, showing the ‘cold blob’ in the northern Atlantic west of the British Isles discussed below. Graph by Ruijian Gou.

1. The AMOC is a big deal for climate. The Atlantic meridional overturning circulation (AMOC) is a large-scale overturning motion of the entire Atlantic, from the Southern Ocean to the high north. It moves around 15 million cubic meters of water per second (i.e. 15 Sverdrup). The AMOC water passes through the Gulf Stream along a part of its much longer journey, but contributes only the smaller part of its total flow of around 90 Sverdrup. The AMOC is driven by density differences and is a deep reaching vertical overturning of the Atlantic; the Gulf Stream is a near-surface current near the US Atlantic coast and mostly driven by winds. The AMOC however moves the bulk of the heat into the northern Atlantic so is highly relevant for climate, because the southward return flow is very cold and deep (heat transport is the flow multiplied by the temperature difference between northward and southward flow). The wind-driven part of the Gulf Stream contributes much less to the net northward heat transport, because that water returns to the south at the surface in the eastern Atlantic at a temperature not much colder than the northward flow, so it leaves little heat behind in the north. So for climate impact, the AMOC is the big deal, not the Gulf Stream.

2. The AMOC has repeatedly shown major instabilities in recent Earth history, for example during the Last Ice Age, prompting concerns about its stability under future global warming, see e.g. Broecker 1987 who warned about “unpleasant surprises in the greenhouse”. Major abrupt past climate changes are linked to AMOC instabilities, including Dansgaard-Oeschger-Events and Heinrich Events. For more on this see my Review Paper in Nature.

3. The AMOC has weakened over the past hundred years. We don’t have direct measurements over such a long time (only since 2004 from the RAPID project), but various indirect indications. We have used the time evolution of the ‘cold blob’ shown above, using SST observations since 1870, to reconstruct the AMOC in Caesar et al. 2018. In that article we also discuss a ‘fingerprint’ of an AMOC slowdown which also includes excessive warming along the North American coast, also seen in Figure 1. That this fingerprint is correlated with the AMOC in historic runs with CMIP6 models has recently been shown by Latif et al. 2022, see Figure 2.

Figure 2. Correlation of SST variations (left) with AMOC variations (right) in historic runs with CMIP6 models, from Latif et al. 2022.

Others have used changes in the Florida Current since 1909, or changes in South Atlantic salinity, to reconstruct past AMOC changes – for details check out my last AMOC article here at RealClimate.

4. The AMOC is now weaker than any time in the past millennium. Several groups of paleoclimatologists have used a variety of methods to reconstruct the AMOC over longer time spans. We compiled the AMOC reconstructions we could find in Caesar et al. 2021, see Figure 3. In case you’re wondering how the proxy data reconstructions compare with other methods for the recent variability since 1950, that is shown in Caesar et al. 2022 (my take: quite well).

Figure 3. A compilation of 9 different proxy series for the AMOC evolution. Data locations are shown in the inset map, from Caesar et al. 2021.

5. The long-term weakening trend is anthropogenic. For one, it is basically what climate models predict as a response to global warming, though I’d argue they underestimate it (see point 8 below). A recent study by Qasmi 2023 has combined observations and models to isolate the role of different drivers and concludes for the ‘cold blob’ region: “Consistent with the observations, an anthropogenic cooling is diagnosed by the method over the last decades (1951–2021) compared to the preindustrial period.”

In addition there appear to be decadal oscillations particularly after the mid-20th Century. They may be natural variability, or an oscillatory response to modern warming, given there is a delayed negative feedback in the system (weak AMOC makes the ‘cold blob’ region cool down, that increases the water density there, which strengthens the AMOC). Increasing oscillation amplitude may also be an early warning sign of the AMOC losing stability, see point 10 below.

The very short term SST variability (seasonal, interannual) in the cold blob region is likely just dominated by the weather, i.e. surface heating and cooling, and not indicative of changes in ocean currents.

6. The AMOC has a tipping point, but it is highly uncertain where it is. This tipping point was first described by Stommel 1961 in a highly simple model which captures a fundamental feedback. The region in the northern Atlantic where the AMOC waters sink down is rather salty, because the AMOC brings salty water from the subtropics to this region. If it becomes less salty by an inflow of freshwater (rain or meltwater from melting ice), the water becomes less dense (less “heavy”), sinks down less, the AMOC slows down. Thus it brings less salt to the region, which slows the AMOC further. It is called the salt advection feedback. Beyond a critical threshold this becomes a self-amplifying “vicious circle” and the AMOC grinds to a halt. That threshold is the AMOC tipping point. Stommel wrote: “The system is inherently frought with possibilities for speculation about climatic change.”

That this tipping point exists has been confirmed in numerous models since Stommel’s 1961 paper, including sophisticated 3-dimensional ocean circulation models as well as fully fledged coupled climate models. We published an early model comparison about this in 2005. The big uncertainty, however, is in how far the present climate is from this tipping point. Models greatly differ in this regard, the location appears to be sensitively dependent on the finer details of the density distribution of the Atlantic waters. I have compared the situation to sailing with a ship into uncharted waters, where you know there are dangerous rocks hidden below the surface that could seriously damage your ship, but you don’t know where they are.

7. Standard climate models have suggested the risk is relatively small during this century. Take the IPCC reports: For example, the Special Report on the Ocean and Cryosphere concluded:

The AMOC is projected to weaken in the 21st century under all RCPs (very likely), although a collapse is very unlikely (medium confidence). Based on CMIP5 projections, by 2300, an AMOC collapse is about
as likely as not for high emissions scenarios and very unlikely for lower ones (medium confidence).

It has long been my opinion that “very unlikely”, meaning less than 10% in the calibrated IPCC uncertainty jargon, is not at all reassuring for a risk we really should rule out with 99.9 % probability, given the devastating consequences should a collapse occur.

8. But: Standard climate models probably underestimate the risk. There are two reasons for that. They largely ignore Greenland ice loss and the resulting freshwater input to the northern Atlantic which contributes to weakening the AMOC. And their AMOC is likely too stable. There is a diagnostic for AMOC stability, namely the overturning freshwater transport, which I introduced in a paper in 1996 based on Stommel’s 1961 model. Basically, if the AMOC exports freshwater out of the Atlantic, then an AMOC weakening would lead to a fresher (less salty) Atlantic, which would weaken the AMOC further. Data suggest that the real AMOC exports freshwater, in most models it imports freshwater. This is still the case and was also discussed at the IUGG conference.

Here a quote from Liu et al. 2014, which nicely sums up the problem and gives some references:

Using oceanic freshwater transport associated with the overturning circulation as an indicator of the AMOC bistability (Rahmstorf 1996), analyses of present-day observations also indicate a bistable AMOC (Weijer et al. 1999; Huisman et al. 2010; Hawkins et al. 2011a,b; Bryden et al. 2011; Garzoli et al. 2012). These observational studies suggest a potentially bistable AMOC in the real world. In contrast, sensitivity experiments in CGCMs tend to show a monostable AMOC (Stouffer et al. 2006), indicating a model bias toward a monostable AMOC. This monostable bias of the AMOC in CGCMs, as first pointed out by Weber et al. (2007) and later confirmed by Drijfhout et al. (2011), could be related to a bias in the northward freshwater transport in the South Atlantic by the meridional overturning circulation.

9. Standard climate models get the observed ‘cold blob’, but only later. Here is some graphs from the current IPCC report, AR6.

Figure 4. Observed vs simulated historic warming (normalised to 1 °C). At this stage the ‘cold blob’ is not yet seen in the model average. Source: IPCC AR6
Figure 5. Simulated warming by the end of this century. Now the ‘cold blob’ appears in the CMIP6 models.

10. There are possible Early Warning Signals (EWS). New methods from nonlinear dynamics search for those warning signals when approaching tipping points in observational data, from cosmology to quantum systems. They use the critical slowing down, increasing variance or increasing autocorrelation in the variability of the system. There is the paper by my PIK colleague Niklas Boers (2021), which used 8 different data series (Figure 6) and concluded there is “strong evidence that the AMOC is indeed approaching a critical, bifurcation-induced transition.”

Figure 6. Early warning signals in four temperature and four salinity based AMOC reconstructions. Note that the tipping point is reached when the lines reach a lambda value of zero. From Boers 2021.

Another study, this time using 312 paleoclimatic proxy data series going back a millennium, is Michel et al. 2022. They argue to have found a “robust estimate, as it is based on sufficiently long observations, that the Atlantic Multidecadal Variability may now be approaching a tipping point after which the Atlantic current system might undergo a critical transition.”

And today (update!) a third comparable study by Danish colleagues has been published, Ditlevsen & Ditlevsen 2023, which expects the tipping point already around 2050, with a 95% uncertainty range for the years 2025-2095. Individual studies always have weaknesses and limitations, but when several studies with different data and methods point to a tipping point that is already quite close, I think this risk should be taken very seriously.

Conclusion

Timing of the critical AMOC transition is still highly uncertain, but increasingly the evidence points to the risk being far greater than 10 % during this century – even rather worrying for the next few decades. The conservative IPCC estimate, based on climate models which are too stable and don’t get the full freshwater forcing, is in my view outdated now. I side with the recent Climate Tipping Points report by the OECD, which advised:

Yet, the current scientific evidence unequivocally supports unprecedented, urgent and ambitious climate action to tackle the risks of climate system tipping points.

If you like to know more about this topic, you can either watch my short talk from the Exeter Tipping Points conference last autumn (where also Peter Ditlevsen first presented the study which was just published), or the longer video of my EPA Climate Lecture in Dublin Mansion House last April.

Area-based global hydro-climatological indicators

by

The World Meteorological Organisation (WMO) Global Climate Observing System (GCOS) and Copernicus Climate Change Services (C3S) both provide sets of global climate statistics to summarise the state of Earth’s climate. They are indeed valuable indicators for the global or regional mean temperature, greenhouse gas concentrations, both ice volume and area, ocean heat, acidification, and the global sea level.

Still, I find it surprising that the set does not include any statistics on the global hydrological cycle, relevant to rainfall patterns and droughts. Two obvious global hydro-climatological indicators are the total mass of water falling on Earth’s surface each day P and the fraction of Earth’s surface area on which it falls Ap.  

[Read more…] about Area-based global hydro-climatological indicators

Back to basics

by

You can tell how worried the climate deniers are by how many fields of science they have to trash to try and have people not see what’s happening.

it will not have escaped most people’s notice that global temperatures are heading into uncharted territory. The proximate cause of this week’s headlines is the Climate Reanalyzer website at the U. Maine which provides a nice front end to the NOAA NCEP CFS forecast system and reanalysis and shows absolute daily temperatures in early July clearly exceeding the highest pre-existing temperatures from August 2016. It’s an arresting graphic, and follows in from the record high ocean surface temperatures that were being reported a month ago.

surface temperature as a function of day since 1979, showing 2023 exceeding the warmest temperatures seen in the previous record.

This is however a relatively new resource and was not online the last time that we set absolute temperature records (in summer 2016). So this has both salience and novelty – a potent combination!

The ultimate cause of these patterns is of course the ongoing global warming, driven almost entirely by human activities.

What are we looking at?

As we’ve explained before, all global temperature products are based on some kind of model – statistical, physical etc. There is no direct measurement of the global temperature – not from satellites, stations, or from the one random person who happens to be in most average place on Earth (where might that even be?). But that doesn’t mean the products aren’t useful!

In this particular instance we are looking at the output of a weather forecast model (NCEP CFS) that ingests multiple sources of in situ and satellite data every 3 hours which is then averaged over a day and over the surface of the planet. These calculations are precise reflections of what is in the model, but for multiple reasons this might not be a perfect reflection of what the real world is doing.

We looked at the coherence of different products, including the reanalyses, before and found that while they are highly correlated in terms of annual anomalies, they differ in their absolute magnitude (graphic from 2017).

Global mean temperature variations from different reanalyses.

Differences will depend on resolution – higher resolution models have better (and higher topography) and then will have slightly cooler temperatures (all else being equal – which it isn’t!), tuning, model structure etc. and can’t really be discriminated using the pure (sparse) observations.

Coherence at the monthly scale is also quite good (though a little noisier), and I haven’t (yet) seen a good comparison of the coherence of the different products at the daily scale (note that the standard products (like GISTEMP, HadCRUT5 and NOAAv5) don’t produce a daily product). One might anticipate that there is a similarity, but perhaps not a one-to-one correspondence on exactly which days were the warmest.

Monthly anomalies since 2010 from ten different products showing a broad correlation between products but with offsets in the global mean.

What are we seeing?

For the global temperature, it’s well established that the maximum is during the Northern Hemisphere summer. This sometimes comes as a surprise to people (why doesn’t the opposing seasonality in the Southern Hemisphere cancel this out?), but it relates to the fact that there is a lot more land in the Northern Hemisphere. Since the seasonal cycle over land is much larger than over the ocean (smaller heat capacity, and less evaporation), that means that the seasonal variations in the north outweigh the variations in the south.

Thus the months of July and August are generally the warmest in the year, and consequently we expect the warmest days during those months – and this is reflected in the CFS output (and in the ERA5 output also). The monthly variations are also reflected in the GISTEMP product which allows you to see the shifts from 1880 onward (about a 1ºC warming in each month since the late 19th C):

The station-based products are a little delayed with respect to the reanalyses, but they generally reflect the same patterns – thus one should expect the June temperatures in NOAA, GISTEMP and HadCRUT5 to be the warmest June on record. Given too, that these temperatures are being driven by persistent warming in the oceans, increasingly juiced by the growing El Niño event in the tropical Pacific, records in July and August are also likely. This is of course increasing the odds for 2023 to be a record year (I would estimate about 50% at this point).

But the WSJ Opinion page says that there’s no such thing as the global temperature!

Well, they would say that wouldn’t they. [Narrator: there is, in fact, a perfectly well defined global mean of any two-dimensional field defined on the sphere, including temperature].

More generously, one might think that their argument (such as it is), is that the global mean isn’t directly relevant for anyone. That is, no-one lives in the global mean, all impacts are local and driven by weather variations. But we’ve known for decades that the global mean change is a really good predictor (not perfect, but pretty good) of local impacts on heat waves, intense rainfall, drought intensity etc.

But let’s be honest, it’s basically pure distraction and attempts to complicate something that is pretty basic:

The climate is warming, records are being broken, and we are increasingly seeing the impacts.

I know why the WSJ doesn’t want you to realise this, but it’s not hard to see past their obfuscation.

Turning a new page[s]

by

The world is full of climate dashboards (and dashboards of dashboards), and so you might imagine that all datasets and comparisons are instantly available in whatever graphical form you like. Unfortunately, we often want graphics to emphasize a particular point or comparison, and generic graphs from the producers of the data often don’t have the same goal in mind. Dashboards that allow for more flexibility (like WoodForTrees) are useful, but aren’t as visually appealing as they could be. Thus, I find myself creating bespoke graphics of climate and climate model data all the time.

Some of these are maintained on the Climate model-observations comparison page but many of the graphs that I make (often to make a point on twitter) aren’t saved there and often their provenance is a bit obscure. Given that twitter will not last forever (though it might be around for slightly longer than a head of lettuce), it’s probably useful to have a spot to upload these graphics to, along with some explanation, to serve as a reference.

I have therefore created a couple of ‘pages’ (in wordpress speak) with fixed URLs where I will be curating relevant graphics I make (and findable at the bottom of the page under “DATA AND GRAPHICS”). The first is focused on the surface temperature records. I often update relevant graphics associated with this in early January (when we get another dot on the graphs), but there are associated graphs that I’ve made that don’t make it into those updates, so this is a place for them too. This includes the impacts of ENSO, comparisons across different platforms, or the impact of homogenization.

Comparison of four instrumental records which all coherently show warming since 1880.

The second page is bit more eclectic. These are graphs that are relevant to some trope or talking point that often pops up, and my graphs are an attempt to provide context (usually), or to debunk it entirely. This is where you’ll find maps of where the climate is warming faster than the global average, time-series of river ice break-up dates, and an example of sensible scaling of CO2 changes and temperature.

Map showing all the areas where trends from 1971-2022 are greater than the global mean trend. Almost all of the northern hemisphere landmass, and much of the SH land too.

To start with, I’m just going to upload some graphs I’ve made recently (with any updates that are needed), and I’ll add content as I make something new. If there are any other ideas (that aren’t too involved!), I’ll be happy to look at adding those too. Let me know if this is useful.

Evaluation of GCM simulations with a regional focus.

by

Do the global climate models (GCMs) we use for describing future climate change really capture the change and variations in the region that we want to study? There are widely used tools for evaluating global climate models, such as the ESMValTool, but they don’t provide the answers that I seek.

I use GCMs to provide information about large-scale conditions, processes and phenomena in the atmosphere that I can use as predictors in downscaling future climate projections. I also want to know whether the ensemble of GCM simulations that I use provides representative statistics of the actual regional climate I’m interested in. 

[Read more…] about Evaluation of GCM simulations with a regional focus.

CMIP6: Not-so-sudden stratospheric cooling

by

As predicted in 1967 by Manabe and Wetherald, the stratosphere has been cooling.

A new paper by Ben Santer and colleagues has appeared in PNAS where they extend their previous work on the detection and attribution of anthropogenic climate change to include the upper stratosphere, using observations from the Stratospheric Sounding Units (SSUs) (and their successors, the AMSU instruments) that have flown since 1979.

[Read more…] about CMIP6: Not-so-sudden stratospheric cooling

References

  1. B.D. Santer, S. Po-Chedley, L. Zhao, C. Zou, Q. Fu, S. Solomon, D.W.J. Thompson, C. Mears, and K.E. Taylor, "Exceptional stratospheric contribution to human fingerprints on atmospheric temperature", Proceedings of the National Academy of Sciences, vol. 120, 2023. http://dx.doi.org/10.1073/pnas.2300758120

A NOAA-STAR dataset is born…

by

What does a new entrant in the lower troposphere satellite record stakes really imply?

At the beginning of the year, we noted that the NOAA-STAR group had produced a new version (v5.0) of their MSU TMT satellite retrievals which was quite a radical departure from the previous version (4.1). It turns out that v5 has a notable lower trend than v4.1, which had the highest trend among the UAH and RSS retrievals. The paper describing the new version (Zou et al., 2023) came out in March, and with it the availability of not only updated TMT and TLS records (which had existed in the version 4.1), but also a new TLT (Temperature of the Lower Troposphere) record (from 1981 to present). The updated TMT series was featured in the model data comparison already, but we haven’t yet shown the new TLT data in context.

[Read more…] about A NOAA-STAR dataset is born…

References

  1. C. Zou, H. Xu, X. Hao, and Q. Liu, "Mid‐Tropospheric Layer Temperature Record Derived From Satellite Microwave Sounder Observations With Backward Merging Approach", Journal of Geophysical Research: Atmospheres, vol. 128, 2023. http://dx.doi.org/10.1029/2022JD037472

Some new CMIP6 MSU comparisons

by

We add some of the CMIP6 models to the updateable MSU [and SST] comparisons.

After my annual update, I was pointed to some MSU-related diagnostics for many of the CMIP6 models (24 of them at least) from Po-Chedley et al. (2022) courtesy of Ben Santer. These are slightly different to what we have shown for CMIP5 in that the diagnostic is the tropical corrected-TMT (following Fu et al., 2004) which is a better representation of the mid-troposphere than the classic TMT diagnostic through an adjustment using the lower stratosphere record (i.e. TMT_{corr} = 1.1 TMT - 0.1 TLS).

[Read more…] about Some new CMIP6 MSU comparisons

References

  1. S. Po-Chedley, J.T. Fasullo, N. Siler, Z.M. Labe, E.A. Barnes, C.J.W. Bonfils, and B.D. Santer, "Internal variability and forcing influence model–satellite differences in the rate of tropical tropospheric warming", Proceedings of the National Academy of Sciences, vol. 119, 2022. http://dx.doi.org/10.1073/pnas.2209431119
  2. Q. Fu, C.M. Johanson, S.G. Warren, and D.J. Seidel, "Contribution of stratospheric cooling to satellite-inferred tropospheric temperature trends", Nature, vol. 429, pp. 55-58, 2004. http://dx.doi.org/10.1038/nature02524

How not to science

by

A trip down memory lane and a lesson on scientific integrity.

I had reason to be reviewing the history of MSU satellite retrievals for atmospheric temperatures recently. It’s a fascinating story of technology, creativity, hubris, error, imagination, rivalry, politics, and (for some) a search for scientific consilience – worthy of movie script perhaps? – but I want to highlight a minor little thing. Something so small that I’d never noticed it before, and I don’t recall anyone else pointing it out, but it is something I find very telling.

The story starts in the early 90’s, but what caught my eye was a single line in an op-ed (sub. req.) written two decades later:

… in 1994 we published an article in the journal Nature showing that the actual global temperature trend was “one-quarter of the magnitude of climate model results.”McNider and Christy, Feb 19th 2014, Wall Street Journal

Most of the op-ed is a rather tired rehash of faux outrage based on a comment made by John Kerry (the then Secretary of State) and we can skip right past that. It’s only other claim of note is a early outing of John Christy’s misleading graphs comparing the CMIP5 models to the satellite data but we’ll get back to that later.

First though, let’s dig into that line. The 1994 article is a short correspondence piece in Nature, where Christy and McNider analyzed MSU2R lower troposphere dataset and using ENSO and stratospheric volcanic effects to derive an ‘underlying’ global warming trend of 0.09 K/decade. This was to be compared with “warming rates of 0.3 to 0.4 K/decade” from models which was referenced to Manabe et al. (1991) and Boer et al. (1992). Hence the “one quarter” claim.

But lets dig deeper into each of those elements in turn. First, 1994 was pretty early on in terms of MSU science. The raw trend in the (then Version C) MSU2R record from 1979-1993 was -0.04 K/decade. [Remember ‘satellite cooling’?]. This was before Wentz and Schabel (1998) pointed out that orbital decay in the NOAA satellites was imparting a strong cooling bias (about 0.12 K/decade) on the MSU2R (TLT) record. Secondly, the two cited modeling papers don’t actually give an estimated warming trends for the 1980s and early 90s. The first is a transient model run using a canonical 1% increasing CO<sub>2</sub> – a standard experiment, but not one intended to match the real world growth of CO2 concentrations. The second model study is a simple equilibrium 2xCO2 run with the Canadian climate model, and does not report relevant transient warming rates at all. This odd referencing was pointed out in correspondence with Spencer and Christy by Hansen et al. (1995) who also noted that underlying model SAT trends for the relevant period were expected to be more like 0.1-0.15 K/decade. So the claim that the MSU temperatures were warming at “one quarter” the rate of the models wasn’t even valid in 1994. They might have more credibly claimed “two thirds” the rate, but the uncertainties are such that no such claim would have been robust (for instance, just the uncertainties on the linear regression alone are ~ +/-0.14 K/dec).

This image has an empty alt attribute; its file name is mcnider55-253x600.png
Replication of the Christy and McNider calculation and figure from 1994 but using the UAH v5.5 data.

But it gets worse. In 2014, McNider and Christy were well aware of the orbital decay correction (1998), and they were even aware of the diurnal drift correction that was needed because of a sign error introduced while trying to fix the orbital decay issue (discovered in 2005). The version of the MSU2R product at the beginning of 2014 was version 5.5, and that had a raw trend of -0.01 K/decade 1979-1993 (+/- 0.18 K/dec 95% CI, natch). Using an analogous methodology to that used in 1994 (see figure to the right), the underlying linear trend after accounting for ENSO and volcanic aerosols was…. 0.15 K/dec! Almost identical to the expected trend from models!

So not only was their original claim incorrect at the time, but had they repeated the analysis in 2014, their own updated data and method would have shown that there was no discrepancy at all.

Now in 2014, there was a longer record and more suitable models to compare to. Models had been run with appropriate volcanic forcings and in large enough ensembles that there was a quantified spread of expected trends. Comparisons could now be done in a more sophisticated away, that compared like with like and took account of many different elements of uncertainty (forcings, weather, structural effects in models and observations etc.). But McNider and Christy chose not to do that.

Instead, they chose to hide the structural uncertainty in the MSU retrievals (the TMT trends for 1979-2013 in UAH v5.5 and RSS v3.3 were 0.04 and 0.08 +/- 0.05 K/dec respectively – a factor of two different!), and ignore the spread in the CMIP5 models TMT trends [0.08,0.36] and graph it in a way as to maximise the visual disparity in a frankly misleading way. Additionally, they decided to highlight the slower warming TMT records instead of the TLT record they had discussed in 1994. For contrast, the UAH v5.5 TLT trends for 1979-2013 were 0.14± 0.05 K/dec.

But all these choices were made in the service of rhetoric, not science, to suggest that models are, and had always been, wrong, and that the UAH MSU data had always been right. A claim moreover that is totally backwards.

Richard Feynman often spoke about a certain kind of self-critical integrity as being necessary to do credible science. That kind of integrity was in very short supply in this op-ed.

References

  1. J.R. Christy, and R.T. McNider, "Satellite greenhouse signal", Nature, vol. 367, pp. 325-325, 1994. http://dx.doi.org/10.1038/367325a0
  2. F.J. Wentz, and M. Schabel, "Effects of orbital decay on satellite-derived lower-tropospheric temperature trends", Nature, vol. 394, pp. 661-664, 1998. http://dx.doi.org/10.1038/29267
  3. J. Hansen, H. Wilson, M. Sato, R. Ruedy, K. Shah, and E. Hansen, "Satellite and surface temperature data at odds?", Climatic Change, vol. 30, pp. 103-117, 1995. http://dx.doi.org/10.1007/BF01093228