The Azimuth Project
Does global warming lag or lead a rise in greenhouse gas concentration?


Does global warming lag or lead a rise in greenhouse gas concentration? refers to a comparison of earth temperatures to concentrations of greenhouse gases (GHG). This page documents a discussion about the corresponding findings.


In a comment in the Azimuth Forum, John Baez mentioned this article:

• Ole Humlum, Kjell Stordahl and Jan-Erik Solheim, The phase relation between atmospheric carbon dioxide and global temperature, Global and Planetary Change 100 (2013), 51–69.

which claims to show temperature changes generally precede changes in carbon dioxide concentration.

The article has a response on the blog RealClimate, which criticizes its findings. In particular its conclusions are criticized:

These results merely confirm already well-known facts, which ironically, they themselves hinted to in their paper (but they obviously did not make the connection):

changes in atmospheric CO2 appears to be initiated near or a short distance south of the Equator, and from there spreads towards the two Poles within a year or so.

The answer is of course: El Nino! A google scholar search with ‘”El Nino” AND CO2′ gives more than 20,000 hits, and Humlum et al. have rediscovered well-known facts which Keeling and Revelle discussed already in 1985.

See more at:

The connection to El Niño is indicated graphically in the RealClimate post. In a comment by John Baez links to charts displaying El Nino events are there given in more detail, a probable correlation between El Nino events and temperature is there more visible.


The article by Ole Humlum et al includes the following chart:

humlum co2 temperature phase

which was sort of reproduced by RealClimate:


Here diff12 seems to be defined as the forward difference of two time-dependent values differing by a year (here the values are the annual means of CO2 and temperature anomalies).

The US based “National climatic data center NOAA” provides the following chart in this context:

Temperature change and carbon dioxide change

NOAA chart temperature vs co2

which displays that co2 concentrations are sometimes lagging behind and sometimes precede temperature changes.

The data is here however not directly measured (since the data goes back to 450 000 ys), which may explain why the the “lagging behind” or “preceding” looks as being sometimes in the range of 10000-30000 years.

In this context it was pointed out that it would also be interesting to see charts of other green house gases vs. temperature. In particular the co2 concentration seems to be connected with the methane concentrations. On the wikipedia page: Atmospheric methane it was pointed out (23.02.2013) that:

Methane is created near the Earth’s surface, primarily by microorganisms by the process of methanogenesis. It is carried into the stratosphere by rising air in the tropics. Uncontrolled build-up of methane in the atmosphere is naturally checked — although human influence can upset this natural regulation — by methane’s reaction with hydroxyl radicals formed from singlet oxygen atoms and with water vapor. It has a net lifetime of about 10 years, (39) and is primarily removed by conversion to carbon dioxide and water.

where the reference (39) is:

see also Lifetime of Methane in the atmossphere in the Azimuth project.

Over half of the methane seems to be converted into co2, citation from p. 3 of Boucher et al:

This gives a lower bound of 0.51… and an upper bound of 1.0 for the fraction of methane that is converted to carbon dioxide.

However these findings are sofar somewhat in contradiction with information from other sources.

Methane and temperature

Following the above discussion about the decay of methane into CO2 I wanted to investigate the diff12 values of methane in comparision with temperature CO2. Unfortunately there is not much so much fine grain methane data available. In fact I found sofar only usable data from NOAA. In an email Ed Dlugokencky who has been in charge for the methane measurements at NOAA Mauna Loa pointed out that besides the systematic measurements of CH4 at the global scale from NOAA which started in 1983 there are measurements from UC Irvine, but which are infrequent and random spatial coverage (Blake et al.) and there is data from ice cores, firn air, and archive samples from Etheridge et al., 1998 JGR).

The diff12 of methane values can be compared in an interactive diagram at our personal website with the diff12 values of CO2 and Temperature anomalies from HADCRUT. The findings of Humlum et al. are recovered. Moreover it looks to me that methane diff12 values precede temperature diff12 values (and thus diff12 CO2 values, which could be due to the decay of methane into CO2). That is I find it looks indeed a bit as if most of the temperature diff12 peaks lag behind the methane peaks by approximately 1/2-1 a year 1). It would be good to compare this also with other temperature data, in particular with raw local temperatures and other locations. And apart from the fact that there might be a bug in the code - it is of course scientifically rather shaky to infer causal relations from such findings and thus it would be way too unsolid to infer from the above that methane might be a bigger driving force of global warming than CO2 - however it is suggestive to look for further evidence into that direction.

1)update 06122013: There was actually a bug (!) in the offsets in the drawing routines in the interactive applet which had been corrected by now. The lag doesn’t look anymore that big, but it still looks like there is a lag. Furthermore the new methane measurements are now included they go now until 12,2012.

Here a current screenshot of the interactive applet, where the red bar had been placed onto the most distinct peak for comparision:


Further discussions about this topic can be found at the Forum thread Temperature vs GHG concentrations and methane. In particular the role of precipitation and wetlands for cooling and methane emissions has been raised there.

On the interactive diagram at our personal website it was already mentioned that the temperature values seem to reoccur with a two year periodicity, this is rather good visible if one uses the slider. In a comment to the Azimuth blog it was mentioned that since the temperature periodicity looks rather precise and connected to the earth year it would be good to look wether it is connected to some planetary motion (even if a biannual cycle appears to be an odd phenomenom within the solar system), but may be one should look at a longer time-scale first, on the other hand it seems temperature data before the fifties may anyway be not precise enough. One should at least first check wether other temperature data displays the same regularity. The page Is there an exact biannual global temperature oscillation? deals with this question in more detail.

Since as pointed out -in contrast to CO2- the lag for the methane looks not so clear (in fact as said in it “looks” as if it precedes temperature values, where this might be true only in some cases) there could be a complicated interaction between the causes for elevated methane and temperature values. It was mentioned in this comment that methane may -similar to CO2- be released due to the upheavel of nutrient waters during trade wind phenomena similar to El Nino (thats as the author has sofar understood the common opinion on the causes for a correlation of CO2 values with El Nino) or on the contrary that strong trade winds may occur due to methane induced heat (islands?) or both at the same time. The latter possibilities appear to be suggestive not only because of the missing lag of methane behind temperatures but also because of other possible on-a-first-sight-correlation of methane and temperature values.

The author doesn’t know about how precisely the global warming potential of methane had been determined. In discussions at the forum it was therefore asked to which extend the absorption peak of methane in the ultraviolet has been taken into account.

The following image displays what is meant with the regular 2 year temperature oscillations.


In the image every peak which lies within the grey biannual interval is indicated by a red dot. A peak which is not in the interval is yellow. On the diff12(TEMP) curve there are 2 yellow dots (out of 26 dots altogether) and the two yellow dots lie almost within an interval, which is why the author thinks that this pattern is very regular. A peak which is not really a peak but just some kind of bump is indicated by a dot with no color filling, there are three dots of this type.

The peaks are best visible in the diff12 temperature curve, however they are also visible in the mean temperature curve (dark green line on temp anom). Each peak in this mean temperature curve seems to be accompagnied with a rather rapid oscillation in temperature values, which mean results in that peak. The temperature values have also this rapid oscillations in between those years, here however the mean seems to be no peak or an antipeak. Sometimes however there is also a small peak this is indicated by blue dots. So in some sense it looks as if there are annual spikes of rapid oscillations, which however on average sum to a peak in one year and to an antipeak in the consecutive year. So in some sense the oscillation may eventually be due to events which are connected to the solar year, like - this is just a very bold guess - eg different mean temperatures in the two hemisspheres.

It should be mentioned that the temperature data is partially based on the CRUTEM4 station data. However upon stronger scrutiny it turned out that the station data has gotten rather uncomplete in recent years. A discussion about this is summarized under How good is climate science temperature data?