by Frank Bosse
Probably not, despite the recent headlines.
A recent article in the journal Nature The deep ocean upsets the slowing and warming caused by melting water in Antarctica by Anh et al. (hereinafter referred to as E23) caused a stir in the media. BBC wrote:Antarctic ocean currents headed for collapse – report.
E23 built a model to describe the formation and behavior of deep water bodies around Antarctica. The Antarctic abyssal waters are important due to its effect on the reverse circulation (AOC) – the lower part of the Meridional Overturning Circulation (MOC) – which reverses heat, fresh water, oxygen, carbon and nutrients in the abyssal ocean. AOC directly affects warming and the availability of nutrients to support marine life near the ocean surface.
Here is a diagram of the global MOC:
Figure.1.: Global MOC, copy of Figure.1 of Marshall/Speer (2012). Antarctic Bottom Water (AABW) is shown on the left side in descending blue arrows.
E23 concludes:
“In particular, a slow-down in the deep ocean circulation of just over 40% is predicted to occur by 205.”
According to E23, this will also have some impact on the Atlantic Meridian Reverse Circulation (AMOC), which is responsible for much of the northward heat transport on earth:
“As meltwater from Greenland and Antarctica increases over time, both the AABW is reversed and the AMOC intensity weakens by 2050.” (AMOC is 19% shown in Figure 2).
This is due to additional meltwater from the Antarctic ice shelves, which has a widespread impact on the Antarctic Bottom Water (AABW):
“First, the amount of Antarctic melt expected to be replenished will cause unusual freshness. . . which produces fresher and less dense AABW, and ultimately reduces the mass of AABW after the 2030s.”
The main figure of E23:
Figure 2: Copies of Figures 3a, b in E23. Melting in Antarctica will result in a man-made AABW reduction (aka “Climate Crisis”) of AABW by 42%(a) by 2050, shown in red. Black: no compulsion.
In Figure 2 (b) AMOC shows a strong downtrend for the period 2004-2020; this is not the case in the observations of “Rapid” at 26.5N; there was a lot of volatility inside, with a drop in 2010 and then a slight recovery.
Figure: AMOC 2004 to 2020 observations of “Rapid” at 26.5°N: Source.
Let’s now see how the authors calculate the melt up to 2050, which is the key input to the model described for AABW. From the Methods section of E23:
“…and the multi-model overall mean of the CMIP6 models under the high anthropogenic emission scenario, Common Socio-Economic Roadmap 5-8.5 (SSP5-8.5), for the climate component. in the future from 2020 to 2050.”
in one twitter thread the lead author stated (and provided a “Share it” link to read full article, thanks for this):
“…our predictions are made under the ‘business as usual’ scenario. Deep and urgent emission reductions will give us a chance to avoid ocean collapse.”
Does SSP5-8.5 (or RCP 8.5 in IPCC AR5) have “Normal Operation”? Not so, stated this commentalso in “Nature”:
“Stop using the worst-case scenario of climate warming as the most likely outcome.”
Its projections of future greenhouse gas emissions are often admittedly unrealistic even on pessimistic assumptions.
Also: is the Multi-Model Synchronizer (MME) mean of the CMIP6 models suitable for this approach? No, the average value of the MME is highly skewed due to the high climate sensitivity caused by some models running too hot. Gavin Schmidt:
“The default behavior in the community should be to move away from treating the raw model group as meaningful.”
This leads to the urgent need for a discussion on the selection of SPS-8.5 and the significance of the CMIP6 grouping in E23. Unfortunately, the article doesn’t do this, so I’ll do it in this blog post.
What effect does the selection of projected temperatures in Antarctica between 2020 and 2050 have, in turn, on projected melting?
With the help of the KNMI Climate explorer, I investigated the projected trends, first the settings used in E23:
Figure.3: Linear temperature trends in and around Antarctica for SPS-8.5 and MMM of CMIP6, as estimated in E23.
For comparison, the MME significance of CMIP5 is not too biased for the more likely RCP4.5 scenario:
Figure.4: Linear temperature trends for RCP4.5 in and around Antarctica.
Note that the trend gradients in the key melt regions of western Antarctica (including the Ross Sea and the Weddell Sea) in Figure 3 are nearly 50% steeper than in Figure 4. This leads to warming in Figure 3. this region from 2020 to 2050 is 0.6 K (Figure 4) based on RCP 4.5 and the mean of CMIP5 MME; in Figure 3, the result is 1.3 K based on SSP5-8.5 and the mean of CMIP6 MME.
However, these are climate model simulation results. Compare the spatially resolved linear trends of the grid cells for the 60°S to 90°S region between 1990 and 2021, roughly the length of a 30-year period from 2020 to 2050 in E23 for observations (GISS) and the “not-so-hot case” of Figure 4:
Figure 5: Spatial trend gradient of plotted data for CMIP5 models with RCP 4.5 scenario (left) and observations, GISS (right) for the period 1990-2021 (“Consequence Forecasting) “)
The simulations have not only warmed so rapidly in Antarctica and surrounding regions over the past 30 years,
but the modeled warming correlates poorly with the observed warming in most of the grid cells (Fig.5)
In areas that are large and important for AABW formation, particularly over the coastlines (except for the Ross Sea at the bottom), the observed and modeled trends are quite different – in different perspectives. closely, these trends are close to zero.
The more realistic scenario CMIP5 RCP4.5 shows warming twice as fast in Antarctica (60°N-90°N) as observed and the average scenario CMIP 6-SSP5-8.5 shows warming almost 3 times faster in the “hindcast” 1990 to 2021 despite the fact that there are relatively small differences in greenhouse gas emissions and changes in other climate change drivers across scenarios SSP5-8.5 and RCP4.5 and observations during that period.
Given the uncertainty of spatially resolved trends in the observations, I use the relationship between trends for the entire Antarctic region, estimating that warming trends in the models will continue through 2050. This will lead to an additional warming of only 0.3K for 2020 to 2050 in the West Arctic, 23% estimated in E23.
Conclusion
Neither the single warming scenario nor the multi-model CMIP6 complex are meant to be used by E23 to estimate Antarctic melting through 2050. The resulting MME is meant to overestimate the potential for the surface to heat up and thus melt, making it “not meaningful” (see Gavin Schmidt quote) as input to the AABW model used in E23 .
For key regions, the trend slope from 1990 to 2021 in the observations is only about a third of the simulations used by E23. Furthermore, the projected future levels and emissions of greenhouse gases are unrealistically high in the SSP5-8.5 scenario used by E23. This suggests that future surface warming in and around Antarctica is likely to be much lower than assumed by E23, which means that the ice melts and hence the deep ocean churn rate. will be much slower than E23 projects.
Furthermore, E23 concluded that ocean cooling due to melting near areas in western Antarctica (namely the Ross Sea and the Weddell Sea) would lead to a reduction in AABW as described over 30 years. to 2050. I looked at observational data “Argo” , provided by “”sea map“. As of December 2021, there are no trends in the salinity data, here shown for average depths of 0-2000 m in the Weddell Sea:
Figure 6: “Argo” observations of ocean salinity near the Weddell Sea. In the Ross Sea area, there was also no trend (not shown). This figure was created with “Marine Atlas”.
The problems with this paper are: unreasonable dependence on the SSP5-8.5 emission scenario, synchronous use of CMIP6 multi-model means overheating, and failure to critically evaluate the model simulations by recent observations. The subsequent failures of Nature’s review and editorial processes, combined with exaggerated and uncritical media advertising, unnecessarily confounded the scientific community and the public. .
Recognize: I thank Nic Lewis for his very helpful comments on earlier draft versions.
