Arguments
What caused early 20th Century warming?
What the science says...
| Select a level... |
Basic
|
Intermediate
|
Advanced
| ||||
|
Early 20th century warming is due to several causes, including rising CO2. |
|||||||
Climate Myth...
It warmed before 1940 when CO2 was low
"Of the rise in temperature during the 20th century, the bulk occurred from 1900 to 1940. It was followed by the aforementioned cooling trend from 1940 to around 1975. Yet the concentration of greenhouse gases was measurably higher in that later period than in the former. That drop in temperature came after what was described in the National Geographic as 'six decades of abnormal warmth'." (James Schlesinger)
The climate at any one time is affected by several factors which can act independently or together. The main factors include solar variability, volcanic activity, atmospheric composition, the amount of sunlight reflected back into space, ocean currents and changes in the Earth's orbit.
Before 1940, the increase in temperature is believed to have been caused mainly by two factors:
- Increasing solar activity; and
- Low volcanic activity (as eruptions can have a cooling effect by blocking out the sun).
Other factors, including greenhouse gases, also contributed to the warming and regional factors played a significant role in increasing temperatures in some regions, most notably changes in ocean currents which led to warmer-than-average sea temperatures in the North Atlantic. Does this mean that solar activity is also primarily responsible for late 20th century warming? In short, no. Solar activity since the 1950s has been relatively stable and therefore cannot explain recent trends. Similarly, increased volcanic activity may actually have had a cooling effect in recent decades. On the other hand, greenhouse gas concentrations, which were relatively low pre-1940, have increased considerably and are now dominating the climate system. This highlights the need to look at all factors when determining which factors are likely to be affecting climate at any one time.
In short, there's no reason to assume that because the sun was responsible for early 20th century, it is responsible for all warming. The evidence strongly suggests that current warming is mainly the result of increasing greenhouse gas levels.
Basic rebuttal written by Anne-Marie Blackburn
Update July 2015:
Here is a related lecture-video from Denial101x - Making Sense of Climate Science Denial
Last updated on 9 July 2015 by pattimer. View Archives
Klapper - I would suggest as a starting point following the link I previously provided for tables of forcings, and from there to the tropospheric and stratospheric aerosol measures. Those include listing multiple supporting references that you can read for questions you might have regarding those measures and uncertainties.
Again - if you are leaving out major forcings such as aerosols, you are basically guaranteed to have incorrect results.
Klapper @25: "They are the most recent I could find so I'll assume they are the best."
Why would you assume that? If they do not reflect the range of reasonable values that are supportable in the literature (remember, we have no direct satellite-based measurements of TSI from the 1910-1940 period to check the reconstructions against), then you are assuming that TSI for the period is known with greater accuracy than it is. "The best" in science isn't necessarily decided by "the most recent".
You just may be on your way to assuming your conclusions again.
Have you read and understood my example in #18? Do you agree with it, or do you find fault in the reasoning? IMO, you keep sending yourself off on the Bad Idea path...
Klapper: if you haven't already seen it, I suggest you read the comment here, posted on another thread. Look for the quote half way down, from Gavin Schmidt (RealClimate) on the 1910-1940 period.
@KR #26:
I checked the graph (a) from Hansen et al 2007 at the GISS website link you provided. In the period 1910 to 1945 the reflective aerosol and indirect aerosol effects combined look to almost cancel out the forcing from GHG's. There are postive forcings from BC and TSI in this period but these forcings estimates appear to be small compared to GHG's and aerosols.
To me this confirms what I've been saying all along: there is some forcing at play in the 1910 to 1945 period which the models have not captured. Since these forcings come from the GISS E model, I also checked the warming rate from a GISS Model E2-R model run of the 20th century from the CMIP5 database. The warming rate in the 1910 to 1945 period from that model is 0.043C/decade. Compare that to the actual observations of 0.14C/decade.
By including aerosols, the model/observations divergence just becomes that much worse. The error on the GISS model trend would have to be +/- .06C/decade to close the gap with the observations.
@Bob Loblaw #28
If Gavin thinks there's reason to believe the lapse rate, or ocean heat mixing were different in the period 1910 to 1945, then he should do a model run to see if that explains the model/observation discrepancy. As it is, I checked a GISS Model E2-R model run of the 20th century and it predicts a warming rate of less than 1/3 the warming rate of the observations in this period. There is no excuse the trend isn't long enough either. There is also no excuse that volcanoes muddled the picture.
If the error boundaries overlap, absolving everyone from having to think about this discrepancy, they probably only barely overlap. I don't think I'm going to convince you this period is worthy of more detailed investigation, and may pose a problem for model assumptions, and I don't think I'll try. I've done enough back of the envelope calculations on TSI (and also aerosols, using some guesstimates for forcing from Hansen et al 2007, chart (a)), to believe the model/observations gap is larger than the posting above would have you conclude. Please no comments on "back of the envelope" after all that's all the above calculations are too.
Klapper:
All the model runs in the world will not improve the accuracy of the observations. Gavin's quote clearly talks about the lack of suitable observations in the period in question. You cannot "improve" a model to the point where it is more accurate than the observations you are testing it with. Trying to fit the model to errors in the data will just introduce errors into the model. It is becoming increasingly obvious that you cannot see past this barrier to your learning.
Now you're just creating strawmen. I never said the period wasn't worthy of investigation. There is a limit to what can be done, however, due to the limitations of the observations that are available to test the model. Unless the observations (or proxy reconstructions) can be improved, doing more model runs won't change that. It is far more productive to continue to work on comparing the models to more recent periods where observations are more complete.
Given the lack of any response to my question regarding my example at #18, I will conclude that you either haven't bothered reading it, don't understand it, or are just unwilling to deal with its implications. Until you provide an answer to my question (in #27), instead of avoiding it, there seems little point in watching you chase your tail.
@Bob Loblaw # 31:
Bob, your posts are too long. Don't waste your time bloviating about sociology/politics, just stick to what's wrong with my numbers. I have read your #18 post but I'm not going to respond to non-qualitative commentary (more or less, I might change my mind but for now let's just stick to numbers/quantitative type commentary).For example, if my delta TSI number of 0.09W/m2/decade is wrong, then show me a recent peer-reviewed number that is higher, and give me the reason it is better.
Klapper:
I have already told you why a reconstruction using nothing but sunspot numbers is poor/limited, and KR has pointed you to literature disucssing sunspot number in more detail.
That you dismiss a mathematical discussion - of how limited accuracy in observations means limited ability to improve models - as "bloviating about sociology/politics" tells me pretty much all I need to know about your scientific skills.
It is up to you to demonstrate why your back of the envelope calculation is better than what is in the literature. I won't hold my breath waiting.
Whenever I show this plot in a talk, someone invariably asks about the cause of the maximum around 1945. The only info I can find, including your posting above, talks about the general increase from about 1910 to 1945, followed by aerosol cooling until around 1975.
This 1935-1950 feature certainly looks real so the only thing I can say to the questioner is that I am not aware of a specific answer from climatologists, but I suppose that the pronounced bump is an accident of natural variability. The questioner usually appears as unsatisfied as I feel. Do any of you have a better answer?
[DB] Resized image breaking page formatting.
tcflood
These graphs, also from GISS, show different regions - Norther exa-tropics, tropics and southern exa-tropics. Quite different pictures.
Also this paper might be interesting, particularly figure 11.
tcflood: The narrow WWII peak itself is almost certainly spurious. It only shows up in the SST data, and particularly in ERSSTv4. There are very substantial known changes in SST measurement practices during WWII, which are hard to correct for completely. Hansen discusses it here.
Kevin C @36, this is the relevant image from the article to which you link:
The right hand series of panels is ERSST v4-v3b. In it land temperatures are cancelled out and consequently not shown. The middle column of panels (ERSSTv4) however, does show land temperatures. In particular, it shows a very warm Chile plus Argentina, more or less adjacent to the unusually warm south east Pacific temperatures. Further, checking the 250 km resolution, meteorological station only map for the 44-45 period at Gistemp shows an off shore (island) meteological station of the coast of Namibia or Angola which again shows unusual warmth. Finally, the west coast of Australia is warm, although not exceptionally so in the 250 km resolution GISTemp map. These correlations to the warm pools in the SH in that period suggest those warm pools are not, or at least not entirely artifacts.
@35 - 37 in response to 34
Thanks for the references and comments. Now I can give a more informed answer to the question.