Global average temperatures: are there just two significant factors ...?

Question

Considering the period from 1850 to now; assuming that changes due to orbital cycles, ENSO and normal solar variation are not significant (too slow, too fast and too small respectively): Considering the period covered by the instrumental record; the HadCRUT4 Global mean, we can see that, as well as a general warming trend, there appear to be a regular cycle where the warming is either enhanced or diminished.

http://woodfortrees.org/plot/hadcrut4gl/mean:360/plot/hadcrut4gl/mean:240/plot/hadcrut4gl/mean:120

To me, this cycle seems to be about 64 years. So if we take a running mean of 64 years, this will tend to remove the cycle and we can see what is left.

http://woodfortrees.org/plot/hadcrut4gl/mean:768

Now the reason I am asking about this is that I am surprised by the smoothness of the curve. To me, this suggests that there are no other significant factors; just a 60 year cycle and an accelerating warming trend!

Now I agree that, ideally, we would have a couple of hundred years more data, so perhaps I m reading too much into this. But if it is reasonable, what are the two factors? The obvious explanation for the warming is increasing GHG s, but what about the cycle? ... the AMO perhaps?

http://woodfortrees.org/plot/hadcrut4gl/mean:360/plot/esrl-amo/mean:360/offset:-0.5

Finally, what does it suggest about future temperatures?

Answer 1

This is a really interesting question, but unfortunately the argument is flawed and the short answer is no. I'll have to do a bit of explaining ...



Let's suppose I take a few landmarks. Let's say the Statue of Liberty, the Mt. Rushmore monument, the CN Tower in Toronto, and the Epcot Center in Disney World, Florida. Now, could I find a selection of stars in the sky whose relative positions match those landmarks? Probably ... I dunno ... never tried. But the point is that the fact that I might find some correlation between the position of stars and position of certain landmarks doesn't mean that there is any actual physical or designed reason for it.



And this, in some sense, is the problem. It's a problem with a lot of the analysis done on Y/A. What has happened here is that you have looked at a set of data. You looked at it and saw what you think is a pattern in that data. You can fit that data reasonably well using a 60-ish year cycle combined with an upward trend. But this, like the stars and buildings, doesn't mean the two are linked. What other ways could you get a good fit to the data? Could you take, just as an example, a 5th order polynomial with a 3rd order polynomial and downward trend? Why would this be less satisfactory than your fit? Is it simply because it's more complicated? Maybe the more complicated fit is better! Could you take two cycles, with different time periods, start them at different points in history, and then find, when they combine around 1850 - 1970 that you get two humps that look like they're one 60 year cycle? Hopefully you can see what I'm getting at!



This is one of the main issues you encounter in data analysis - fitting curves to data is relatively easy. And there are an infinite number of ways of doing that. The problem is that those fits tell you *nothing* about the underlying process in the absence of a model because there is a multitude of ways I can fit the data!



If there was some reason we might expect a 60 year cycle, we had done some calculation as to how that would impact temperatures, and then applied our predictions as a fit to the data, and looked at the 'goodness' of that fit, and found it to be small, then that's a good scientific argument. That's a 'we had predicted a certain behaviour and found it in the data'. One explanation for the seeming 60 year cycle was the influence of Jupiter and Saturn on Earth but the problem with this explanation was that a) lots of things could possibly give you a 60 year cycle if you looked hard enough for physical phenomena that occur on 60 year cycles and b) no calculation of the effect of this influence on temperatures was ever attempted.



What you generally do not do is take data, find some combination of mathematical functions that give a fit to it, and then argue that because the fit is good, it means the fit belies some underlying physical mechanism. So this is why I say the analysis is flawed.



What's more, I can prove this. What I did was take the Hadcrut4 data and put it into graphing software. I've attached a link to that initial graph so you can have a look ...



http://i63.tinypic.com/123tb2r.jpg



Then what I did was take the Fourier Transform of it. What this shows is the periodicity of any cycles within the data. The FFT is in this graph ...



http://i66.tinypic.com/2ahi5fm.jpg



What you can see is that there is no real evidence of a cycle around 60 years duration. If you look at the FFT you'll see a peak at around 14 years, a peak around about 28, one around 42 and one around 56 ish. These are 'harmonics' caused by the upward trend (it's a roughly triangle function so you'd expect to see odd and even harmonics in the FFT as a result).



So what my analysis tells me is that you have a lot of short period noise (in the 0 to 10 years cycles range) which is caused by the month-to-month fluctuations, an overall upward trend based on the harmonics I see, but no real evidence of a longer cycle time on the sorts of scales that you're talking about. And again, I stress, the fact that you think you see one, and that you can fit the data with curves that seem to match that cycle, isn't correct analysis. It ignores the fact that there could be lots of other cycles that conspire to give you such a shape!

Answer 2

It is not correct to say that solar variation is too small to influence temperature. Up to middle of the twentieth century, Earth's temperature correlated very well with solar activity. What has happened since then is that solar activity has dropped and Earth has warmed.

Answer 3

I'm not sure how meaningful it really is to take the mean of 64 years, and it's of course reasonable to expect the smoothness to increase with the mean period in general. The 64-year graph will be affected dramatically by the amount of data available. If you had 100,000 years of data the 64-year mean would have about the same visual appearance as the 1-month mean does with its ~150 years of data.



The graph will always smooth itself as your mean gets larger. Bring it up to 167 years and you're left with a single point.

http://woodfortrees.org/plot/hadcrut4gl/mean:2013



I'm not sure if there's much in the way of significance about 64 based on these results.

Answer 4

https://history.aip.org/history/climate/index.htm





I'm not familiar with the details of the scientific findings, but AFAIK, there are several different significant factors involved in the warming & cooling of the earth.



The factor that's most important NOW is human society's emissions of "greenhouse" gases that affect how much of the sun's incoming radiation is radiated back into space, and how much of it remains in the lower atmosphere to cause "global warming.



Another very important factor TODAY, from what I've read, is the "algedo" or reflectiveness of the Earth's surface and of the clouds in the upper atmosphere. The place where this is urgently important, according to many climate scientists, is the Arctic circle.



There the polar ice cap, because it's white, normally reflects most of the sun's energy that strikes it back into outer space. This helps to keep the planet cooler. But as the Arctic region has warmed over the past 25 years, the size of the ice cap in the summer months has shrunk dramatically. White, reflective ice has given way to darker green or blue sea water, which absorbs much more of the sun's incoming radiation.



This decreasing "albedo" -- this decreasing reflectiveness -- of the Arctic polar region should cause the planet to heat up more quickly. It should step on the gas and accelerate the process of global warming.



3. But there are other factors involved in warming & cooling the planet as well.



a. One is the shape of the earth's orbit around the sun, which varies over a period of many thousands of years. At some times, the orbit causes more of the sun's energy to hit the Arctic polar region in the summers, which accelerates global warming. At other times, a changing orbit causes less of the sun's energy to hit the Arctic polar region in the summers, which can bring on an ice age.



b. The "tilt" of the earth on its axis also goes through a cycle of changes over many thousands of years. When the tilt allows more of the Arctic to be exposed to summer sunlight, this encourages global warming. When the tilt exposes less of the Arctic to summer sunlight, we have global cooling and we normally head into an ice age.



c. The intensity of the sun also varies over -- I think it's an 11-year cycle, which corresponds to the cycle of sunspots appearing and disappearing. More intense sunlight coming into the earth's atmosphere causes warming conditions; less intense sunlight causes cooling conditions.



d. The circulation of deep ocean currents, which can carry heat from what part of the earth to other parts, also affects global warming and cooling. That's why some climate scientists believe the climate was significantly changed when drifting tectonic plates caused North America and South America to come together at the Isthmus of Panama many millions of years ago. This changed the major ocean currents and changed the climate as well.



d. The weathering of mountain rocks by erosion has an indirect effect on climate because it removes carbon dioxide from the atmosphere, combines it with minerals in the rocks, and then causes the minerals with carbon attached to them to be washed downstream off the mountains into rivers, and from rivers into the oceans and the seas. Over millions of years, this causes carbon that used to be present in the atmosphere -- and that has caused global warming -- to become buried in deposits beneath the ocean floor. Which should cool the planet.



e. The big cycles of ocean currents that you mention in your question also affect the average warmth of the atmosphere. I don't know if they do so in the way the articles you linked to suggest. But since a great deal of the "global warming" that has happened from greenhouse gas emissions since 1850 has added heat to the oceans, not to the atmosphere or the land, how the big ocean currents circulate that heat has a big influence on what the climate of coming centuries will be.



-- democratic socialist / retired environmental journalist

Answer 5

no

Answer 6

It is the Sun and its cycles. Without the Sun there would no ocean currents or life on Earth.



The Sun is currently transitioning into an inactive cycle called a Solar Grand Minimum. Sun spot numbers and TSI are at 150 years lows similar to that of the Maunder Minimum. Another tell tale sign is the increase in Galactic Cosmic Rays causing more cloud cover into the mid latitudes and influencing Global Cooling.



CO2 is non-factor, the Greenhouse Effect is a myth, NASA, NOAA and East Anglia are climate frauds.



It is the Sun and Nature, Humans cannot control either and Nature cannot be fooled.

Answer 7

NO!



First, you have a "TERRIBLE" surface temperature record that you are using and too many "TERRIBLE" scientists who back up the "TERRIBLE" surface temperature record. They are stuck in "CO2 Warming Mode".



Second, you act as if you are oblivious of surface ocean temperature and its affects on overall land surface temperatures. ARGO data is correlating with surface temperature swings with ocean oscillation temperature swings magnificently. Lots of "cause" with "effects".



How do you explain the current lack of warming in the Arctic Region along with the exceptional COLD records being set?



I just read a 2011 article about how ARGO Floats are changing ALL of the past ocean modeling. There's BAD correlation with past scientific studies and their conclusions based on research before the advent of ARGO.



There are (at least) 3 factors.



I'm still curious about the parameters set in the temperature records that make up the "1850 data". There has to be a lot of "assumptions and conjecture" in order to gain an accurate accounting, because there isn't any actual mechanical temperature data over most of the planet, so they have to rely on BAD proxy data to get a result.



Alarmism has squandered many opportunities for scientific honesty.



You are perpetuating it by suggesting that there are only 2 variables to surface temperature changes.



The 1920 to 1950 ERA was a significant warming period that was nullified due to the replacement of surface proxy data and ocean proxy data.



Let's stick with the influx of ARGO data for awhile. We already know that surface temps are immediately related to temperature fluctuations of the ocean surface. Satellite data is being shown to be highly divergent.



Science still doesn't have a handle on it IMO.



Sunday Night Football (11/12/2017) score is New England 14 and Denver 3 with 8:19 left in the first quarter.



Additionally :



Proxy data is usually taken from below the surface area. Maybe that's a good spot to measure Earth's surface.



Above surface data is always measured "in the air", which simply doesn't apply to Earth's actual surface temperature. We already know that CO2 barely affects surface air temperature. A 40% increase in CO2 level has barely moved our surface air temperature over the past 350+ years.



Don't we measure our own temperature from within?

Answer 8

You are not the first to discover this relationship. It has been published. It is the PDO which is principally responsible for the observed 60 year oscillation. The problem comes in when you base your models on the 30 year positive phase and extrapolate from there. Certainly CO2 plays a factor, forcing the overall trend up. The big question is "up how much?"



Edit: Three TD's so far. Alarmists do hate this article and the graph.

Answer 9

Fake data. Crap in and crap out!

Answer 10

CO2 is certainly important

Answer 11

I doubt the majority do, but I wouldn't be surprised if a few actually want to do so.

Answer 12

I cant see what your graphs are measuring

Both PDO and AMO have cycles, and I dont see how you can say solar is not significant, if solar can shrink or expand the height of our atmosphere as NASA says it does I dont think you can say it is insignificant.

There are certainly more than 2 factors what about thermohaline circulation.