Mike Jackson

You come to the point very well.

As you say, is the observed variation consistent with natural variation or is it necessary to include CO2?

I would say you need the CO2. The natural variable forcings have been either constant or becoming more negative since 1970, yet temperatures continue to increase. Even the combination of a weak solar maximum, above average vulcanism and increased air pollution only slowed the increase.

Taking account of natural variation only, temperatures should be cooling. Instead we get this tug of war in which the last decade or so has seen increased negative forcing almost cancelling the effect of increased CO2.

Have you chosen an onset date for the pause so that I can draw my graphs?

TinyCO2, Michael hart

Thae are the sort of questions I am hoping to test. Have you chosen an onset date and are you happy with my wording tfor the hypotheses?

Entropic man

I would say you need the CO2. The natural variable forcings have been either constant or becoming more negative since 1970, yet temperatures continue to increase. Even the combination of a weak solar maximum, above average vulcanism and increased air pollution only slowed the increase.

What makes you so certain of the bold section above and are you certain that the italic section isn't due to poor understanding of UHI and poor TOBs adjustments for instance?

EM this may seem strange coming from a sceptic but if the guys with billion pound computers can't make sensible predictions, anything else is a waste of time. I don't favour sceptic attempts to create climate models, I think climate is too complicated. I suspect there are a great many processes that aren't even known yet. I know the pause is problematic to climate modelling, simply because there are so many excuses why it's happened. Every month without significant warming is a further erosion of the original understanding behind the models.

You threw out those who don't trust the measurements. Well think on the basic metric of climate change - thermometer measurments. We are talking about a series of figures that not only change over time due to adjustments (right or wrong) but in total only span little over a degree. We are told that the series from 1850 is more or less right, which I can believe but for the kind of computer modelling they're trying to do, more or less isn't good enough. Each wiggle from hour to hour, let along year to year is part of the planet's programming. Every metre of the planet is part of it's matrix. Smoothing stuff out into years or hundreds of kilometres pretends those wiggles don't mean anything. It now looks like oceans have a lot to say on climate - we haven't got detailed records of how the planet moved heat about. We have no good evidence of what a warm period looks like. All our records start at the bottom of the LIA so our knowledge is based on one part of a cycle. Proxies are better than nothing but they're very sparse chains of evidence, with a fair bit of guesstimation involved. Perhaps storing energy in the ocean is how the planet staves off an ice age, releasing the heat when the oceans get below a certain level? Think about tiny errors in measurements that individually don't matter but magnified across time and space become big enough to change the way the models work. Any time spent on chaos theory should tell you that climate modelling is pretty much impossible, doing so without meticulous and lengthy data is completely impossible. If the modellers can't look at proxies and recreate past temperature changes exactly (not just ball park), then the model doesn't model.

A month doesn't seem to go by without several papers saying 'X climate event doesn't work quite the way we thought it did'. Whether it's the North Atlantic Conveyor or Antarctic ice. How do they know if one of those issues isn't a significant part of how the climate works? Another thing is the number of papers concluding 'this effect could account for X% of observed global warming'. They add up to very much more than 100% and while they can't all be right, what if some of then are partially right?

If back of fag packet caculations were enough to work out what the planet was doing then the models would be right. If the number of measurements we are taking were good enough then they'd know exactly where the missing heat is and/or why it's not where they thought it should be. They would also be able to explain previous situations (eg EU and US desuplurisation.) If you think natural processes are hiding the predicted warming, go look at Hansens' 1988 prediction. By now we should be seeing a half, to a degree more warming. If we subtract that warming from 1970s temperature, we'd not only be well on our way to another LIA but potentially on our way to a real ice age!

Look at these two images a mere year and a half apart.

http://www.ospo.noaa.gov/data/sst/anomaly/1999/anomnight.1.2.1999.gif
http://www.ospo.noaa.gov/data/sst/anomaly/1997/anomnight.7.1.1997.gif

Tell me a few ship's measurements could capture this kind of complexity before. We just don't have the data to understand how climate works. We don't even know yet what triggers an El Nino or a La Nina yet.

TinyCO2

You can set boundary conditions for some of these uncertainties. Tallbloke and I once independantly calculated that the amount of energy taken up and released during the 60 year AMO cycle could produce a temperature change of no more than 0.3C.

And if you have several cycles that don't release the stored energy? We're already 'missing' more that 0.3 C. If it can plunge into the deep ocean where we're not measuring it, then rapid warming of the 80s and 90s could be partly due to energy stored decades or even hundreds of years ago. We don't know, we weren't measuring it. If there is a built in, seemingly random time delay between energy being stored and released how can you know from the past if warming was due to input from the sun or from stored energy? How do you know if cooling was real cooling or just the oceans in a storage phase?

tinyCO2

There is one long term ocean heat storage system, the thermohaline circulation. Cycle time for some of the relatively warm saline water sinking off Greenland is 1000 years. Unfortunately for your hypothesis it doesn't carry enough heat. Similarly with ENSO, AMO and other cycles. They can create short term changes in the surface temperatures, but do not carry enough energy to power a long term trend.

There was a paper published recently which reckoned that there was no significant warming below 2000M. Above 2000M the ARGO floats are seeing warming consistent with expectations

The problem with "unknown" storage or heat transfer systems is that if they are significant their effect would show clearly in the system, measurably.

Applying Occam's Razor, there is no point postulating extra unknown processes when the existing ones successfully account for the energy budget.

Entropic man (& Raff)
You are using a similar premise to Raff in thinking that natural phenomena either confirm the CO2 has an effect or that AGW is falsifiable by using data for thimgs like sea level rise, glacier advance/ retreat.

Using data available on the web for sea levels (back to 1700) it appears to me that the top 10 44 year periods (4 sunspot cycles to give a period of reasonable length) with greatest rise in sea level are the periods ending

1817
1904
1972
1961
1903
1971
1901
1836
1891
1900

Note the most recent year is over 40 years ago. This suggest to be that there is no problem with sea level rise accelerating, and for Raff's OR test AGW is falsified.

SandyS

I'mhaving trouble understanding your point. Could you unpack a little and give a link back to the data?

EM I' m having trouble understanding what your trying to prove. Is it that the earth has warmed and that it's cause is human emissions? Or is it that the earth has warmed and that its cause is human emissions and there'll be catastrophic consequences?

Geronimo

I'm doing what I enjoy, playing with the science. Unfortunately I keep getting distracted by you lot.

Martin A

I' m in Ireland. They know here that leprechauns cause it all. :-)

It is easy to dream up hypotheses from the likely to the bizarre. The trick is to find those which can be usefully tested.

Entropic man

If you read back through this thread for Raff's comments you'll see that he claims that single symptom claimed to be affected by AGW which can be shown to be more extreme for want of a better expression (I've had a long day appointment at rhumatologue at 09:00 meant a 06:00 wake up now 23:00) prior to the current period. I take the period of so called AGW as being about 40 years. Since our discussion on sea level a few months I had collected a couple of references when they turn up on the web.

The data as far as I can tell shows more extreme rises in the late 19th century. This is prior to any major increase in CO2 and if sealevel rises are in major part due to sea temperature rises then this must be due to something other than CO2 induced warming. Raff should now have joined the sceptic camp ;-).

Sources:

http://www.psmsl.org/products/reconstructions/jevrejevaetal2008.php#sthash.TMszKs2b.dpuf

data

http://www.psmsl.org/products/reconstructions/gslGRL2008.txt

You can check it out yourself in Excel, I spliced the missing years from the CU data you referenced last time we discussed this.

"The data as far as I can tell shows more extreme rises in the late 19th century."

Was there not quite significant worldwide glacier retreat (ie Glacier Bay for example) during the 19th century? This and it's causes would have an effect on sea level changes.

Rob Burton
But did the glaciers melt because of CO2 induced warming?

Martin A

Actually it does. You make a prediction and then test the prediction against reality. If they are in good agreement it improves your confidence in the hypothesis. If there is a big difference you change the hypothesis.

For hypotheses testable by laboratory controlled experiments, you do so. If this is not possible, for example in a planetary atmosphere, you compare prediction with field observations.

Remember, you can never "prove" a hypothesis, only disprove it.

SandyS

Nice reference, thank you. I fear you are seeing patterns in the noise.

Entropic man
Not me, I only looked when people including your good self said that sea level rises were unprecedented and/or due to man made CO2. If I'm seeing patterns in the noise. what the heck were you and others doing? Being sceptical means checking for yourself, from what I can see the sea level rise cause by CO2 theory/hypothesis is a dead duck.

Earlier I wrote two alternative pause related hypotheses.

From my side: The rate of warming from 1970 to the onset of the pause represents the long term trend.

From the sceptics: The rate of warming from the onset of the pause to 2013 represents the long term trend.

I have come up with a ready reckoner to compare these two alternatives.

This graph shows HadCRUt4 from 1970 on.

I have added a 1970-1998 linear trend in green and a 1998-2013 linear trend in blue. I printed out the graph and extended both trends to the year 2040.

HadCRUt4 annual averages have 95% confidence limits of +/- 0.1C. I drew boundary lines above and below each linear trend to match.

The two 95% confidence ranges no longer overlap after 2032, so it will probably take until then to make a statistically significant distinction between them.

I intend to tuck this away and come back to it annually to see which trend is looking more probable. This is the sort of homebrew science that helps me crosscheck the professionals and the propaganda.

TinyCO2

While doing the ready reckoner above, it occured to me that HadCRUt4 may be showing one of the storage/release phenomena we discussed. The temperatures followed the 1970-1998 trend line quite well until the 1997/1998 El Nino and then jumped upwards. Perhaps a rapid release of stored heat from the Pacific. The temperatures may then have increased more slowly as a series of La Ninas mopped up the excess heat.

Recent temperatures are now back on the 1970-1998 trend line. It will be interesting to see where they go next.

SandyS

It may be my own stupidity, but I still can't reconstruct the 40 year extreme changes in rate you describe. Could you show your working.

Entropic Man
Unfortunately I haven't saved my workings but from memory I repeated and without messing too much to exactly replicate this is the result. The method is the same not the period.

I used Excel, Down loaded the data from the links I gave you. Added the last few years using Colorado data (I know it's not the same but it was the best I had) This gave me 3 columns of data; column A was the year and column B the sea level. anomaly. and 316 rows, one of which was the header. I plotted a chart to eyeball what I had against the link I posted as a sanity check. Both appear to cross the zero anomaly in the late 1930s.

I decided on a purely arbitrary 46 year period. The data starts at 1700, so at the row for 1746 I inserted the formula =(B48-B2)/(A48-A2) at the first row and I used a calculated value for the number of years in case any data was missing. I used Excel to replicate the formula for all the rows from 1745 to 2014. I then copy and pasted values (not formula) to a new worksheet. Then ordered by the 46 year value largest to smallest, using this repeated process I get the top 10 periods end in the years

1850
1834
1903
1975
1964
1904
1907
1980
1971
1835

There is no sign of recent decades appearing in the list.

If you can repeat the exercise and get different results. post here as I'd be interested. Looking at my chart I reckon that a 70 year period might give the last decade as showing 70 year periods with greatest increases as there are two pauses in the post 1860 data when the steady rise begins, The first ends in about 1938 so choosing somewhere round there to present to give the measurement period should give the first decade of the 21st century as the worst case.

The current sea level situation according to the data is not unprecedented not even unusual.