Guest Post by Willis Eschenbach
Among blue-water sailors like myself, the area of the ocean from 40° South to 50° South is called the “Roaring Forties” because of the strength of the winds that blow there so often.
And the next ten degrees south of that? They are called the “Furious 50’s”, followed by the “Screaming 60’s” (thanks to the commenters who corrected my names). There, the winds can blow unimpeded around the globe.
Figure 1. The cold pole.
Compared to the South Pole, the North Polar weather is far less complex. The North Pole is underlain by the Arctic Ocean, ice-covered most of the time. The South Pole, on the other hand, has a continent in the middle. And it’s mostly a high elevation permanently frozen chunk of rock. It sheds icy winds and glacial chunks into the surrounding ocean. Due to the constant winds and storms, the “mixed layer” around Antarctica is deeper than anywhere in the world.
Figure 2. Mixed layer depth.
And why is this of note? Well, I was asked to take a look at a recent peer-reviewed study called “Simulated Twentieth-Century Ocean Warming in the Amundsen Sea, West Antarctica“.
And where is the Amundsen Sea when it’s at home? It’s off of the coast of Antarctica, down below 60°S.
Figure 3. Oceanic warming trends, showing the Amundsen Sea.
A short digression. Roald Amundsen was one of my heroes when I was a kid. He was a famed polar explorer. He led the first expedition to reach the South Pole. He was also the first man to sail the Northwest Passage from the Pacific to the Atlantic over the top of Asia and Europe. As a boy I remember seeing his ship “Gjøa”, the one that he used for the Northwest Passage, up on blocks in Golden Gate Park in San Francisco. Even as a young inexperienced sailor, I was amazed at how small it is, only 70 feet (21 meters) long. Amundsen definitely had albondigas of pure brass … but I digress
The press release about the study is”Researchers demonstrate new link between greenhouse gases and sea-level rise“. Inter alia, it says (emphasis mine):
A new study provides the first evidence that rising greenhouse gases have a long-term warming effect on the Amundsen Sea in West Antarctica. Scientists from British Antarctic Survey (BAS) say that while others have proposed this link, no one has been able to demonstrate it.
Well … in a word, no. The study doesn’t provide any evidence at all, not one scrap. What it provides instead are the results of computer models using another computer model as input.
Sigh. Look, if computer models were “evidence”, I’d be a very rich man based on my 1980’s evolutionary-based computer model of the stock market … but computer model results are not evidence. They are merely the understandings and more importantly the misunderstandings of the programmers made solid.
The study itself says:
Our simulations are performed using the Massachusetts Institute of Technology general circulation model including components for the ocean, sea ice, and ice shelf thermodynamics. Here we build on the Amundsen Sea configuration, which has been updated and re-tuned to provide best agreement with observations when forced with the latest ERA5 atmospheric reanalysis.
The “ERA5 atmospheric reanalysis” is a computer climate model which is forced with whatever data is available. The model then fills in the blanks where there is no actual data … like say the Amundsen Sea, where people only venture very rarely, and even then only in modern times.
Let me take a heel-turn to the question of the performance of the climate models. This study is about sea levels and ocean temperatures. Folks keep telling me that the climate models have done well for decades. So I looked at the first IPCC Assessment Report. Here are their sea-level projections:
Figure 4. Sea level projections from the IPCC AR1 Report.
And here is a close-up of that figure, with the actual sea level rise overlaid on the graph.
Figure 5. Actual sea level rise, compared to the IPCC projection.
Moving to more modern models, here are model estimates of the post-1981 sea surface temperature (SST) rise from a number of CMIP6 models. These are the computer outputs of the sea surface temperature identified in the CMIP6 models as the variable “TOS”, the temperature of the ocean surface. Below I’ve compared them to the Reynolds OI SST observational dataset.
Figure 6. Modeled and observational sea surface temperature (SST) rise, 1981 – 2021
As you can see, the models are … well … let me call them “much less than accurate” and leave it at that.
To return to the Amundsen Sea, here are some of the complexities that affect the weather there.
Figure 7. Amundsen Sea.
As you can see, the sea temperature and the weather are affected by the Circumpolar Current, circumpolar winds, the Antarctic Slope Current, the Amundsen Sea Polyna (open water) in the middle of the Amundsen Bay sea ice, the varying ocean depths, and the ever-changing sea ice. It’s a most intractable and complex area to model.
To return to RealWorld from ModelWorld, here’s a look at just where and how the ocean has warmed and cooled since 1981.
Figure 8. Decadal trends in sea surface temperature (SST), Reynolds OI observational dataset.
As you can see, there’s been a bit of cooling in the SST in Amundsen Sea in this 41-year period.
Moving on, how well do their model results agree with the ERA5 reanalysis? Here’s their graphic:
Figure 9. Original Caption: “Timeseries of conditions in the Pacific Pacemaker Ensemble (PACE) simulations (blue) and the ERA5 simulation (red). Thinner blue lines show the 20 PACE ensemble members, while the thicker blue line is the ensemble mean. (a) Zonal wind (m/s) averaged over the shelf break. (b) Temperature (°C) averaged over the Amundsen Sea continental shelf between 200 and 700 m. (c) Basal melt flux (Gt/y) for ice shelves between Dotson and Cosgrove inclusive.
YIKES! Their own models don’t even begin to approximate the ERA5 reanalysis, which is the information that is used as an input to the models. And like the Reynolds OI SST, the ERA5 reanalysis shows slightly cooling temperatures in the Amundsen Sea, where the models claim warming … bad models, no cookies!
Oh, yeah, almost forgot. I did an analysis looking for what I call “weasel words” in their study. These are words that show their conclusions are weak. They include “could”, “may”, “might”, “possibly”, “plausibly”, “potentially”, “suggests”, and “assumed”. Those words are used a total of 34 times in their study … no bueno.
In closing, the study itself says (emphasis mine);
Rapid ice loss is occurring in the Amundsen Sea sector of the West Antarctic Ice Sheet. This ice loss is assumed to be a long-term response to oceanographic forcing, but ocean conditions in the Amundsen Sea are unknown prior to 1994. Here we present a modeling study of Amundsen Sea conditions from 1920 to 2013, using an ensemble of ice-ocean simulations forced by climate model experiments.
“Simulations forced by climate model experiments” for a location where ocean conditions are “unknown prior to 1994” … it’s models on top of models all the way down, what could possibly go wrong? It’s clearly destined to be a lead article in the “Journal of Irreproducible Results“.
I weep for the death of science.
Finally, let me mention that the Amundsen Sea is only 0.085% of the total ocean area … it appears that what we have here are scientists with models looking for funding. Not saying that’s a bad thing, just that it can easily lead to … well … this kind of study.
Here, I’ve just spent three days (well, part-days, I’m retired) mowing a couple of acres of steep hillside in our opening in the forest … and today, rain, marvelous rain. Life is good, what’s not to like?
My best wishes to you all,
AS USUAL: When you comment, please quote the exact words you are responding to, so we can all be clear on your subject.