In this post, we’re going to illustrate how poorly climate models used by the IPCC for their 5th Assessment Report simulate the polar amplification that data indicates took place during the early 20th Century warming period of 1916 to 1945.
This is part 2 of the post Global Mean Surface Temperature: Early 20th Century Warming Period – Models versus Models & Models versus Data (The WattsUpWithThat cross post is here). In that post, using the GISS Land-Ocean Temperature Index (GISS LOTI) dataset, we determined the 30-year period that ended before 1950 with the highest warming rate. It was 1916 to 1945. See Figure 1.
If a graph of 30-year trends is new to you, see the discussion that follows “A note for newcomers” in the post linked above.
In that earlier post, based on the climate model outputs of the ensemble members stored in the CMIP5 archive with historic and RCP8.5 forcings, for the period of 1916-1945, we also identified (1) an ensemble member that ran warmest (GISS-E2-H p3), (2) one that ran cool (IPSL-CM5A-LR EM-1), (3) the ensemble member with the highest trend (IPSL-CM5A-LR EM-2), and (4) the one with the lowest trend (CMCC-CMS), the last of which presented a cooling (yes, cooling) rate of -0.055 deg C/decade during a 30-year period when data indicate global surfaces warmed at a rate of about +0.15 deg C/decade.
AN OVERVIEW OF POLAR AMPLIFICATION
Polar Amplification refers to a natural phenomenon through which surfaces at high latitudes of the Northern Hemisphere warm [cool] at rates that are noticeably higher than the warming [cooling] trend for the rest of the globe.
Yes, data indicate that polar amplification will occur during a period of global cooling and result in excessive cooling in the high latitudes of the Northern Hemisphere, as was experienced in the mid-20th Century. See the 2012 post Polar Amplification: Observations Versus IPCC Climate Models (WattsUpWithThat cross post is here.)
THE SOURCE OF THE DATA AND MODEL OUTPUTS…
…presented in this post is the KNMI Climate Explorer.
MODEL-DATA COMPARISONS OF POLAR AMPLIFICATION
The best way I’ve found to illustrate the effects of polar amplification is by plotting the latitude-averaged surface temperatures trends for the period in question. See Figure 2. It is a graph that compares the observed global mean surface temperatures trends and those of the multi-model mean of CMIP5 climate models outputs, both on a zonal-mean basis. (Instead of latitude average, the climate science industry uses the term zonal mean.) The observations-based data are represented GISS Land-Ocean Temperature Index (GISS LOTI) dataset, and the models are represented by multi-model mean of the climate model simulations of global surface temperatures based on the models stored in the CMIP5 archive with historic and RCP8.5 (worst case) forcings.
The curve of the data during the early 20th Century warming period of 1916 to 1945 (solid grey curve) shows a classic example of polar amplification. After the mid-latitudes of the Northern Hemisphere, the warming rates climb higher and higher as the data near the North Pole…but then suddenly stop. On the other hand, the average of the climate model outputs (the multi-model mean, which is presented as the orangey dashed curve) shows little to no polar amplification during this period.
I present the model mean because it represents the consensus (groupthink) of the climate modeling groups for how surface temperatures should warm if they were warmed by the numerical representations of the forcings that are used to drive the models. Because the models couldn’t simulate the warming that took place during this period at most latitudes (other than the slight warming at about 60S latitude, where the model and data curves intersect) the warming must have occurred naturally.
NOTE: If a graph of near-surface temperature trends on a zonal-mean (latitude-average) basis is new to you, don’t worry…it’s easy to understand. The units for the vertical (y) axis are in deg C per decade (deg C/decade), so the data points represent the trends in surface temperatures (the warming and cooling rates). The horizontal (x) axis is latitude, with the South Pole at -90 (90S), the North Pole at 90 (90N) and with the equator at 0. To create a graph like the one shown in Figure 2, temperature data are downloaded in 5 degree latitude bands. For the graphs in this post, the trends start at -62.5 (62.5S), because there is no observations-based data available from Antarctica before the early 1950s. Those first data points at 62.5S are based on surface temperature data for the latitudes of 65S-60S. The next data points at 57.5S are based on the surface temperatures for the latitude band of 60S-55S. The process continues in 5-degree latitude bands northward toward the North Pole. [End note.]
Figures 3 through 6 present the model-data comparisons using the same observations-based data as Figure 2. They also include the 4 ensemble members we isolated in the post Global Mean Surface Temperature: Early 20th Century Warming Period – Models versus Models & Models versus Data. I’m not going to bother to comment on them individually. You’ll see why.
# # #
# # #
# # #
A COUPLE MORE ENSEMBLE MEMBERS
There are 4 ensemble members at the KNMI Climate Explorer for the runs of the IPSL-CM5A-LR model. When asking for the outputs of a model in anomaly form, the KNMI Climate Explorer provides the outputs of all ensemble members in sequence, so they are easily isolated from one another. As a result I had two more ensemble members of the IPSL-CM5A-LR model on file in MS EXCEL. They are identified at the KNMI Climate Explorer as IPSL-CM5A-LR EM-0 and IPSL-CM5A-LR EM-3. Their simulations of global mean surface temperature trends on a zonal-mean basis for the period of 1916-1945 are presented below in the format of this post as Figures 7 and 8.
# # #
Figure 9 is a spaghetti graph that shows the curves of all of the ensemble members presented in this post. I’ve included it to counter any erroneous impressions created by the model-mean presented in Figure 3. The only thing consistent about the ensemble members of the CMIP5 archive shown in this post is their inconsistency when attempting to simulate global mean surface temperatures and resulting polar amplification during the early 20th Century warming period of 1916 to 1945.
NEXT IN THIS SERIES
The next two posts in this series will examine climate model outputs and compare them to data during the 30-year mid-20th Century cooling period with the lowest linear trend.
In the following, I’ve initially repeated the closing comments from the earlier post in this series.
For the early 20th Century 30-year warming period of 1916-1945, climate models are consistently horrible and consistently inconsistent at simulating the primary metric of human-induced climate change, which is global mean surface temperature.
And, surprisingly, based on those horrendous excuses for climate models, we’re supposed to believe their crystal-ball like prognostications of future global mean surface temperatures and other climate metrics!!?? Fat chance of that happening with anyone who has a spark of common sense. If only more persons understood how poorly climate models simulated global mean surface temperatures—the primary metric of human-induced climate change—the human-induced global warming scare might just disappear into the past like the Y2K scare. Then again, the global warming/climate change scare has nothing to do with science; it is simply global politics at its worst, masquerading as science.
The IPCC couldn’t be a scientific entity. No scientific entity would set its foundation on models that perform as badly as this. The climate models stored in the CMIP5 archive should be presented as examples of failed attempts to simulate Earth’s climate, not used for government policy.
That’s it for this post. Enjoy yourself in the comments and have a wonderful remainder of your day.
STANDARD CLOSING REQUEST
Please purchase my recently published ebooks. As many of you know, last year I published 2 ebooks that are available through Amazon in Kindle format:
- Dad, Why Are You A Global Warming Denier? (For an overview, the blog post that introduced it is here.) Also available in paperback.
- Dad, Is Climate Getting Worse in the United States? (See the blog post here for an overview.)
And please purchase Anthony Watts’s et al. Climate Change: The Facts – 2017.
To those of you who have purchased them, thank you. To those of you who will purchase them, thank you, too.
Thanks again, Bob!
While I appreciate WUWT, its too bad that more people don’t follow your blog. I have to go to WUWT for comments.
I keep trying to get WUWT-ers to visit Climate Observations. The Warmistas malign you, but cannot refute your observation-based postings.
There is probably a simple explanation for the polar amplification: the surface area of a circular zone on a sphere decreases with increasing latitude. This drives the ocean currents (Gulf Stream and Japan Current) deeper. The surface to volume ratio changes and there is less surface evaporation to cool the ocean water. Here is a reference that describes the process.
Alexander, M. A.; M. S. Timlin, and J. D. Scott, Progress in Oceanography 49 41-61 (2001), ‘Winter to winter recurrence of sea surface temperature, salinity and mixed layer depth anomalies’
(I have a .pdf I can send you if have trouble finding the reference)
Also, there is some useful information on the Woods Hole web site. Note the increase in wind speed/evaporation in the winter months and the polar ‘hot spot’ from the Gulf Stream.
Reblogged this on Climate Collections.
Hi Bob, Steve Goddard has a lot of posts showing GISS adjustments that seem to regularly cool the past. 1. Do you agree that they seem to cool the past? 2. If so, does this LOTI data set exhibit the same adjustments. 3. If so, could they have inadvertently caused the early 20th trend to go higher? Thanks so much for an informative post explained simply.
dellwilson, I’ve seen Steve Goddard’s comparisons but haven’t looked into it myself recently. Regarding your last question, I can’t answer without examining the older data, which is not on my list of things to do.
Well done, Bob. You will be interested in reading
if you haven’t already. Mathematician/scientists from the London School of Economics and the Mathematical Institute of Oxford discuss the problems associated with optimizing models and ignoring real world data.
The caution is that the map is not the territory. This needs to be read and widely distributed. It’s an excellent discussion of the Ludic Fallacy in logic, and vitiates the current efforts to refine and improve models in climate-like problems (though not in weather-like problems).
OT, hi bob, just saw this, published today:
Article | Published: 06 May 2019
Higher frequency of Central Pacific El Niño events in recent decades relative to past centuries
Thanks, Alec. In other words, because Central Pacific El Niño events are weaker than East Pacific El Niño events, then, overall, El Niño events have become weaker.