2011-03-28 14:20:09Plot of past CO2 vs solar level over geological time scales
John Cook

john@skepticalscience...
124.185.238.238

I've been given carte blanche to hit any skeptic argument in tomorrow's Climate Show interview so I thought I'd hit a fav of mine - CO2 was higher in the past (which Chris is hitting in nice detail lately).

I thought the clearest way to show this in a single graph is to show a plot of CO2 going back hundreds of millions of years in the same graph as solar activity going back the same time. Probably best to show the radiative forcing from each so they can be compared on the same scale.

I don't suppose anyone would know where these kind of graphs could be found? This is a bit last minute - the interview is in 19 hours but if possible, would be great (would also be a useful graph for the Basic Rebuttal and I'd add it to http://sks.to/graphics

2011-03-28 14:59:25
Glenn Tamblyn

glenn@thefoodgallery.com...
124.176.252.221

John

No I don't know of such a graph but you could approximate it as a straight line I think. However that is misleading since Solar is linear and CO2 is logarithmic. The telling one might be from Royer, showing CO2 & Solar as forcings. However I think they show this as a combined figure. You might be able to get CO2 numbers from Berner and replot as forcing against Solar.

Also take a look at my comments to Chris's last draft post.

2011-03-28 15:07:31
Daniel Bailey
Daniel Bailey
yooper49855@hotmail...
97.83.150.37

Still looking for something I had going back to the PETM-ish, but this is all I can find for now:

GHG

http://ossfoundation.us/projects/environment/global-warming/natural-cycle/overview/

 

GHG2

http://upload.wikimedia.org/wikipedia/commons/f/f5/All_palaeotemps.png

2011-03-28 15:11:38Postpone this idea
John Cook

john@skepticalscience...
124.185.238.238

It's a bit last minute so I'll save this for another time.

You get around the logarithmic CO2/linear solar problem by considering the radiative forcing of both. Berner shows this by showing Sun + CO2 combined:


But what I want to show is CO2 forcing and solar forcing as two separate time series - to show CO2 going up as you go further in the past but sun going down, presumably at the same rate. To do this, I'd need the equation for solar luminosity over time (I think there's a linear equation in Royer 2006) and Royer's data from this graph:


Atmospheric CO2 through the Phanerozoic. Dashed line shows predictions of the GEOCARB carbon cycle model with grey shading representing uncertainty range. Solid line shows smoothed representation of the proxy record (Royer 2006).

2011-03-28 15:44:43
Glenn Tamblyn

glenn@thefoodgallery.com...
124.176.252.221

John

The caption from Fig 2 from Royer

"Radiative forcing through the Phanerozoic. Radiative forcing is derived following the protocol of Crowley (2000a) and the radiative transform expression for CO2 of Myhre et al. (1998) . For the calculation, the CO2 records from Fig. 1D are used and solar luminosity is assumed to linearly increase starting at 94.5% present-day values. Values are expressed relative to pre-industrial conditions (CO2 = 280 ppm; solar luminosity= 342 W/m2); a reference line of zero is given for clarity. "

2011-03-28 15:57:28Royer's solar figures
John Cook

john@skepticalscience...
124.185.238.238

So according to Royer, solar luminosity was 94.5% of present levels at 550 million years ago and increased linearly. I can work with that.

Now just need to find a time series of past CO2. Geocarb looks the way to go as it gives a continuous time series all the way back to 550 Mya

2011-03-28 16:40:45Quick & Dirty
Glenn Tamblyn

glenn@thefoodgallery.com...
124.176.252.221

John

Quick & dirty method if you can't get the geocarb data. Load Royer's fig 2 that shows the net forcing, digitise the dotted line based on GeoCarb and subtract the solar forcing from it to get the CO2 forcing.

I have an xls file that shows this. How do I upload it too you?

 

2011-03-28 17:00:28Can you email me the xls file to john@skepticalscience.com?
John Cook

john@skepticalscience...
124.185.238.238

Less trouble than uploading

What software do you use to digitise a figure?

2011-03-28 17:02:12
Glenn Tamblyn

glenn@thefoodgallery.com...
124.176.252.221

Sent by email

Rough method. Save to a picture like .bmp then just load Paint - the pointer tracks coordinates.

Glenn

 

2011-03-28 17:46:07Comparing solar forcing to CO2 forcing
John Cook

john@skepticalscience...
124.185.238.238

Glenn sent me the XLS of the CO2 forcing vs solar forcing and it's quite striking - tells a very vivid story. In fact, I can't believe noone's done this before (that I know of) as this graph really is the perfect answer in a nutshell to anyone who brings up the "CO2 was higher in the past" argument:

Glenn's methodology - digitised the combined forcing from Royer's pic, worked out solar forcing from Royer's equation and added it to the combined forcing to obtain the CO2 forcing.

However, just to be a pain in the butt, I think this graph would be more robust if we plotted it with CO2 forcing taken from the original GEOCARB values. If its possible to obtain these online anywhere, please post the URL here.

2011-03-28 17:57:01
Rob Painting
Rob
paintingskeri@vodafone.co...
118.93.204.4

Great work John, this what you're after?

2011-03-28 18:05:00
Glenn Tamblyn

glenn@thefoodgallery.com...
58.170.206.72

Agreed John.

This was the quick & dirty version. Definitely needs to be done from original CO2 data.

Its amazing. When I first plotted this I was gob-smacked. Some of those wiggles are Extinction Events. CO2 and Solar doing a pirouette over all that time. Talk about a picture being worth a thousand words.

Perhaps what is needed is John does his pod cast, Chris posts his second Solar Evolution post and we follow up with a post to tie them together with this. Doesn't need many words since a picture says it all.

2011-03-28 18:13:05Note re podcast
John Cook

john@skepticalscience...
124.185.238.238

I decided not to do the Climate Show interview on this subject - it was too last minute, the interview is first thing tomorrow morning. So instead, they said pick one of the top 10 SkS arguments and I realised we hadn't even done "it's the sun" yet. So tomorrow, we're doing the sun.

Rob, you da man, I think that's exactly it. Specifically, it looks like CO2 going back to 570 million years B.P. at:

ftp://ftp.ncdc.noaa.gov/pub/data/paleo/climate_forcing/trace_gases/phanerozoic_co2.txt

Too late today to have a close look at it but will hopefully look at it tomorrow. But the priority tomorrow is getting skeptic quotes for sks.to/skepticquotes - I have a tendency to get easily distracted by shiny objects, baubles or graphs so while I'm tempted to lose myself in this topic, I'm determined to stay the course! If I have time after doing that, will get onto the CO2 graph. Very cool! Look forward to adding this graph to sks.to/graphics and burying "CO2 was higher in the past" once and for all.

2011-03-28 18:19:58
nealjking

nealjking@gmail...
84.151.45.168

Are the proxies in the graphs proxying for CO2 or for temperature? I'm really unclear on what you guys are getting excited about.

In a word, what is the argument?

2011-03-28 18:32:18
Glenn Tamblyn

glenn@thefoodgallery.com...
58.170.206.72

Neal

The proxies are for CO2. This graph is the data from the Geocarb model shown in Royer 2006. That also includes other proxies that are the thick line in John's second graph. Royer showed the relationship between net forcing - Delta CO2 and Delta Solar over that time and that they correlate well with glacial periods. But we had never seen a graph that plots Delta Solar and Delta CO2 forcings separately. That is what this rough one is. Solar & CO2 have changed in amazing lockstep over the last 550 MYr. That wiggle at around 350-300 MYr is a major extinction event.

So Solar Evolution as in Chris's recent post, the weathering thermostat from his second post and the radiative properties of CO2 look like they have been acting in lockstep to keep things pretty stable for a very long time. This has been known but this graph just captures it all so clearly.

It impacts on the various 'CO2 was higher in the past' arguments, ' CO2 and Temperatures aren't correlated', arguments and 'CO2 is Weak/Saturated' arguments.

I think I may have to make a conversion and start supporting Intelligent Design.

2011-03-28 18:34:27The argument
John Cook

john@skepticalscience...
124.185.238.238

Neal, sorry, Glenn and I are getting a little giddy over this graph, aren't we? To get so excited over a graph is definite proof that you're a climate geek (or as we'd call it in Australia, a "climate tragic").

The proxy is for CO2 levels (or in Glenn's graph, the radiative forcing from CO2). The skeptic argument is "CO2 was higher in the past" - the world didn't boil away in the past when CO2 got up to 6000 ppm so that proves CO2 doesn't cause warming. Eg - look at this graph that shows CO2 getting to huge values 500 million years ago...

But skeptics don't ask WHY was CO2 so huge in the past. And the reason is simple - because the further you go back into the Earth's past, the lower solar output is:

When you combine the effect of CO2 and sun, they pretty much cancel out. The reason is because of CO2 weathering. Chemical weathering removes CO2 from the atmosphere. When it gets hotter, weathering speeds up which brings down CO2 levels, cooling things down. Similarly, when it gets colder, weathering slows down and CO2 builds up in the atmosphere, warming things back up again. So CO2 weathering is like a natural thermostat, keeping the planet's climate in balance.

So when you consider sun AND CO2 over Earth's history, it's actually strong evidence FOR carbon dioxide as a powerful driver of climate. This is a classic example of skeptic cherry picking (looking only at CO2) leading to the exact opposite of the conclusion you get when you consider all the evidence (sun + CO2).

We've already explained this before on SkS but we've never had a succinct graph to communicate it so elegantly. And I'm not sure why we've never thought to do such a graph - it seems obvious in hindsight.

2011-03-28 20:34:28Who am I kidding, I can't resist a shiny new graph!
John Cook

john@skepticalscience...
124.185.238.238

Okay, this is crazy but I just couldn't resist doing this graph. But I would welcome someone checking my work to make sure I got it all 100% before this sees the light of day. I've uploaded my spreadsheet to:

http://www.skepticalscience.com/docs/Royer_Forcings.xls

So I downloaded the data from NOAA and plotted it in Excel. I couldn't find any documentation on the NOAA site about what the RCO2 numbers meant but I guessed they were the amount relative to pre-industrial levels of 280ppm. So I plotted that and compared it to Royer's GEOCARB graph (my plot is the blue line):

Seems fairly consistent. Next I plotted the solar levels - worked out the luminosity of the sun so that it hit 94.5% present levels at 550 million years ago. Then to calculate solar forcing from that, I just took the difference between past solar luminosity to present levels and divided by 4. So calculating CO2 forcing using the usual 5.35 x log (C/Co) and plotting it with solar forcing, I got substantially different values to Glenn. This is what I got:

The dotted line is the combined effect of CO2 + Solar forcing. A useful test would be to compare this to Royer's graph of the same metric:

It all seems fairly consistent. But if anyone would like to go over my spreadsheet with a fine toothcomb, I'd welcome any feedback.

2011-03-28 21:55:03
Glenn Tamblyn

glenn@thefoodgallery.com...
124.181.124.17

John

My original digitisation had a factor of 10 error which made the curve look way to flat. And I have made the corrections to your calcs as well. Now pretty strong agreement. Still good but not as dramatic as the original.

Note to self. Doing important calcs while trying to deal with 6 other things at work can lead to errors.

Will post the new XLS tomorrow - off to bed now before the salt mines tomorrow.

 

2011-03-28 21:57:13
nealjking

nealjking@gmail...
84.151.45.168

a) In the third graph from the top: What are the vertical dark-gray bars?

b) There is supposed to be an 800-year lag between temperature increase and CO2 increase, so at least initially, high-T drives CO2. Does that affect the picture in any way?

2011-03-28 22:12:07Errors in my excel
John Cook

john@skepticalscience...
124.185.238.238
Glenn, hmm, does that mean I underestimated my solar forcing by 4? But then the resultant net forcing will be strongly negative. Something ain't right. Feel free to tinker with my XLS.

Neal, the dark vertical bars in the net forcing graph are periods of glaciation. Note they all coincide with periods where net forcing drops below zero (approximately). The only exception is in the late Ordovician around 450 million years ago but this is only because the CO2 proxy is not of high enough resolution. Other more high resolution proxies show CO2 dropping below the 'glaciation threshold' around this time.

The CO2 lag is not relevant at these time scales - we're talking hundreds of millions of years here.

2011-03-28 22:47:44
Rob Painting
Rob
paintingskeri@vodafone.co...
118.93.204.4

Nealstradamus, the GEOCARB models only resolve down to 10 million year time scales IIRC. 

2011-03-29 01:24:48This graph from Hansen tells a different story
James Wight

jameswight@southernphone.com...
112.213.158.190

Forcings 65 My to 0 BP

2011-03-29 02:43:40
Chris Colose

colose@wisc...
64.188.12.126

You can use the equation in my first post from Gough (1981) to create a reasonable curve of solar irradiance.  CO2 is much trickier because there's a lot of uncertainly once you move past the ice core record, especially many hundreds of millions of years ago, and as someone suggested the relation between pCO2 and forcing is non-linear.  The Royer curve is a good reference for what you are looking for I think.

Grinding out the causes of CO2 variations is still harder, and I wouldn't give the impression it's all a response to solar increases over time.  There's a lot of geologic factors involves.

 

By the way, "RCO2" is defined in the Berner paper as the ratio of mass of CO2 in the atmosphere at time t divided by the mass at present (which they suggest can be loosely taken to be 300 ppm).  That's a good paper to show the importance of CO2 over geologic time in terms of global te

2011-03-29 03:16:53good stuff
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

Interesting stuff so far.  It would be good to figure out where your "Present" is, and then add in the more recent solar and CO2 data.  CO2 since 1850 can be found here.  TSI is plotted over the past several thousand years here.

2011-03-29 09:02:56
Riccardo

riccardoreitano@tiscali...
93.147.82.157

Myhre fits his simplified formula in the 300-1000 ppmv range, I don't think it's safe to extrapolate that much.
In the calculation of CO2 forcing you should use the natural logarithm.
Solar forcing is (K3-342)*(1-albedo), not (K3-342)/4.

2011-03-29 14:35:53Elaboration of the graph I posted above
James Wight

jameswight@southernphone.com...
112.213.158.190

I got that graph from Hansen 2008. Hansen’s greenhouse gas forcing is based not on proxy measurements but on the estimated total forcing over that period (derived from deep ocean temperature). I posted it because it doesn’t agree with the graph John has plotted based on CO2 proxies.

Eyeballing Hansen’s graph, the difference in GHG forcing (presumably mostly CO2) between the Holocene and 65 Myr is at least 7 W/m2 (with an even larger difference between the LGM and 50 Myr). But in John’s graph it’s more like 3 W/m2. The solar forcing is about the same.

Hansen 2008 compares its reconstructed GHG forcing to proxy estimates. They match reasonably well in the flat period of the last 35 Myr, but earlier proxies vary widely.

Hansen’s purpose was to estimate a safe CO2 level, so maybe his methods don’t suit our purposes. Still, I’m curious as to the reason for the discrepancy.

2011-03-29 15:21:54Source of the discrepancy
John Cook

john@skepticalscience...
124.185.238.238

My dodgy Excel work is the cause of the discrepancy - Riccardo points out how I got the solar forcing wrong and I think I botched using the log function in Excel. I think my end result was close to correct more from good luck than good management. If I get time today, I'll revise the Excel - otherwise, it looks like Glenn has cleaned up my mess so look forward to seeing his results.

2011-03-29 16:29:17
Glenn Tamblyn

glenn@thefoodgallery.com...
124.181.123.12

Riccardo, you are right about Myrhe - 300 to 1000. Any suggestions for who/what might give a passable extrapolation out to say 6000 ppmv?

I had forgotten about Albedo. In which case a better representation might attempt to apply Albedo for different temp ranges - Ice Free, Current and Full Glacial based on paleo temp records.. Does anyone have any sources for any of this?

 

2011-03-29 20:09:46
Riccardo

riccardoreitano@tiscali...
192.84.150.209

I checked the logarithmic dependence up to 9000 ppm with modtrans and apart from the coefficient (4.8 instead of 5.35) it turns out to hold quite well.

I don't know how to manage albedo changes, though it shouldn't be a big issue given that most of the time there's not much ice around.

Overall, even though the calculations are crude, I think that the message is clear enough.

2011-03-29 20:45:43
Riccardo

riccardoreitano@tiscali...
192.84.150.209

This is what i get:



CO2 forcing is calculated as 5.35*ln(CO2/280) and CO2 is CO2 concentration taken from column I in John's excel file.
The sun forcing is (1-albedo)*0.0342*x with albedo=0.3

2011-03-30 02:54:04caveat
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

Would be worth making specific note that we're ignoring albedo changes in our solar forcing.  And I think it would definitely be worthwhile to show recent data to put the CO2 forcing increase into perspective.

2011-03-30 05:23:26
Riccardo

riccardoreitano@tiscali...
93.147.82.172

It would be a barely visible small vertical line, less than 2 W/m2. Maybe plotting only co2+sun and projections up to year 2100 could make sense.

2011-03-30 06:07:23or narrowing
dana1981
Dana Nuccitelli
dana1981@yahoo...
64.129.227.4

That, or perhaps doing a second plot narrowing in on a shorter timeframe, like the past million years or something.

2011-03-30 12:33:39Albedo forcing
James Wight

jameswight@southernphone.com...
112.213.158.190

As I understand it, albedo is more or less a “matching funds” feedback whenever there are ice sheets on the planet. In other words, whenever the total forcing drops below approximately where it was 35 million years ago.

2011-03-30 14:44:42Some more playing around
Glenn Tamblyn

glenn@thefoodgallery.com...
60.230.35.98

This has a few additions.

Original 3 curves with Albedo of 30%, Solar = 5.35 Ln(C/Co)

Then curve of Solar with very rough variable Albedo used based on climate at the time. 0.2 in a warm world, 0.3 for the current, 0.4 for deep ice age.

Then curve of CO2 & Solar with var Albedo.

Last, plotted against this, paleo Temps, simply taken by eyeball from the Geocraft diagram based on Scotese. Scale for Temps adjusted to roughly align it with the combined forcing curves.

Interesting. The Ordovician doesn't show up since the GeoCarb data is too course. Temps lag the Ice Age at 350 mYr bp. And at around the time of the Jurassic temp drop, CO2 climbs and the two don't start to move in synch again till around the PETM at 55 Myr bp.

So CO2/Solar & Temps agreeing quite well except for those two periods.

What was happening then? Have a look at these maps of paleo continents ate different periods http://www.scotese.com/earth.htm

At 350 Myr bp, the sea ways between Gondwana and Euramerica are still open. By 300 Myr bp they have closed, shutting off equatorial currents and forming Pangea. And Gondwana being over the Sth Pole is a good base for Glacial conditions. Maybe this explains to 350 to 300 lag, the change of ocean currents enhanced the potential for Ice Age conditions which is also considered a factor in our more recent entry into Ice Age 30 Myr bp.

Then at 195 Myr bp Pangea is still intact and no equatorial currents. Then by the Late Jurassic, 150 Myr bp, Pangea is breaking up and equatorial currents become possible again. This continues subsequently, only starting to be reversed as Nth/Sth America join and Africa becomes a barrier across the Equator.

Basic premise. CO2 is strongly correlated with Temps when we take continental configuration into account. Even without considering the continents, there are long periods of good correlation.

Caveats. This is very coarse temp data, we need a better source. Variations in Albedo are also coarse.

Are these observation strong enough for a speculative post? What papers can we use to support this?

 

2011-03-30 21:42:05Suggested approach
John Cook

john@skepticalscience...
124.185.238.238

I believe Royer 2006 supports this. But what do you think of this approach?

  1. Confirm that we've got the graph right. If you like, I can email Royer - I think I've corresponded with him in the past - back then, I was curious how he calculated the solar levels in the past (he just pointed me to the part of the paper where he explained it - a bit embarrassing).
  2. Write up a more detailed post explaining the methodology of how the graph was put together, explaining some of the physics along the way, as part of an Advanced Rebuttal of "CO2 was higher in the past".
  3. I'd like to write the Basic Rebuttal of "CO2 was higher in the past" using a simplified version of Glenn's graph - I have a pretty clear picture in my head of how I'd construct the post
  4. I'd move the existing basic rebuttal of "higher past co2" to the "co2 was high in the Ordovician" argument which is more suitable

Sound like a plan?

2011-03-30 22:41:51Sounds like a plan.
Glenn Tamblyn

glenn@thefoodgallery.com...
124.180.194.29

What would be good would be a better source for paleo Temp's. Anyone know of one? Also linking it to Chris's Solar posts would be good.

I can start drafting the Advanced rebuttal since I had already claimed it but intended to add the Solar evolution/ Thermostat stuff. With Chris having covered them I can simply link to that.

Question for Royer is whether his calc's included Albedo.

Any papers covering the Jurassic would be good.

 

2011-03-31 04:33:44
Chris Colose

colose@wisc...
144.92.130.207

Glenn,

This quote from Berner's GEOCARB III paper might help:

/"This type of modeling
is incapable of delimiting shorter term CO2 fluctuations (Paleocene-Eocene boundary,
late Ordovician glaciation) because of the nature of the input data which is added to
the model as 10 my or longer averages. Thus, exact values of CO2, as shown by the
standard curve, should not be taken literally and are always susceptible to modification.
Nevertheless, the overall trend remains. This means that over the long term there is
indeed a correlation between CO2 and paleotemperature, as manifested by the
atmospheric greenhouse effect."//