On the subject of surface temperature, I happened upon this page recently. http://www.theregister.co.uk/2008/05/02/a_tale_of_two_thermometers/ If the link gets edited, google 'Is the earth getting warmer, or cooler?' for the article, or 'a_tale_of_two_thermometers' for more on the debate.
The article in theregister states
quote:
The UK Meteorological Office's Hadley Center for Climate Studies Had-Crut data shows worldwide temperatures declining since 1998. According to Hadley's data, the earth is not much warmer now than it was than it was in 1878 or 1941.
By contrast, NASA data shows worldwide temperatures increasing at a record pace - and nearly a full degree warmer than 1880.
The reality is that looking at their graphs and looking at the means:
HadCRU figures for 1878, 1941 and recent peak are: -0.2C, 0.0C and +0.5C NASA figures for the same three periods are -0.15C, 0.1C and +0.5C
To me they seem to agree. The "trick" is to put the HadCRU data on a different scale to make the warming look less.
As to the adjustments, I'd be certainly interested to know the justifications. I can think of a few reasons, but I don't know enough to speculate.
quote:
So when I hear that the science is settled do you agree? Are the 90%+ IPCC confidence statements reasonable?
As I've said before. The science is settled on the existence of a link between anthropogenic greenhouse gases and observed warming. Quite a lot else is up for debate. As to the confidence statements, there is a one-in-ten chance of a 90% "very likely" statement being wrong and a one in 3-or-4 of a "likely" statment being wrong. So I would say they are reasonable. I'm not aware of challenges, though in the area of impacts (eg. power and frequency of hurricanes") there is a lot to find out.
Quite a lot else is up for debate. As to the confidence statements, there is a one-in-ten chance of a 90% "very likely" statement being wrong and a one in 3-or-4 of a "likely" statment being wrong.
How's that for Orwellian double-speak?
Anyway, here is some interesting criticism of current temp measurements.
As I've said before. The science is settled on the existence of a link between anthropogenic greenhouse gases and observed warming.
Ah, I get it now "existence of a link" not "the main cause". Are we agreeing? I've never doubted a link.
quote:
The reality is that looking at their graphs and looking at the means:
So when the power goes off on my freezer I shouldn't worry about the actual temperature because the rolling average is still low enough to stop the food going off even though it has thawed? Or by the same logic the Grand Old Duke of York is on average higher after starting to march down than he was at the top of the hill.
I'd suggest the reality is looking at reality ie the actual temperature not a rolling average temperature and the rate of change which was positive and is now negative.
Are the graphs in the Register article numerically wrong?
So when the power goes off on my freezer I shouldn't worry about the actual temperature because the rolling average is still low enough to stop the food going off even though it has thawed? Or by the same logic the Grand Old Duke of York is on average higher after starting to march down than he was at the top of the hill.
The power on your freezer is probably off as much as it is on due to the thermostat
More seriously, we know that climate variability can shift heat from one part of the system to another such that the surface temperatures one year can be lower than another, for example, simply because more heat may have been absorbed by the ocean in one year. I would suggest that the running mean helps smooth out these variations.
So yes, in January and February may have been cold on the surface, but we know the oceans won't have changed significantly since December, hence the prediction that the surface temperatures would recover which they have.
quote:
Are the graphs in the Register article numerically wrong?
Yeah but no but... they are misleading for two reasons. 1. The 2008 value is based on January and February alone. See the latest version which includes March and moves 2008 up to the 14th warmest year - still not a fair comparison as you need 12 months to get a like-for-like. 2. The Hadley data does not include uncertainty. Uncertainty is present due to sampling error - we don't have global coverage - and measurement error - they need to account for station changes (which have associated uncertainties). The latter is where all the climateauditing is going on. You need to see the uncertainties to get a sense of whether a trend is real. The uncertainty is represented on the NASA graph by the green bars. For the HadCRU data the following link includes uncertainty:
All I want is for models to represent the physics as well as possible and that comparisons to real data are taken to account as part of an open and honest process of continual improvement.
I concur 100% with that remark SoM! This is a refreshing change from chaos theory.
We can already see that poster #144 has confused 'radiative forcing' from model runs with actual 'temperature observations' in their graph on the Climate Audit site. It's plain that the 'radiative' system of thermal accountancy is confusing to many people. Though, this is the system that the IPCC have decided to use, so I really don't see any way to make that system clearer without a massive educational initiative or disclaimers posted in big red letters wherever 'radiative' terminology is used and not a more generally understandable 'temperature' representation.
No! The radiative system 'is' too obtuse and shouldn't have been adopted, even if it does make modelling 'kinder' on computer time! Mostly because the 'kindness' requires a lot of data as a 'parameterised input', so it's 'loaded' and then 'run' through the iterations without any alterations that may occur within it (that is, not without a 'tweak' here and there when changes to that parameter are obviously needed). Though this is only my opinion, with a limited knowledge of modelling.
Another thing that grips my curiosity from your link to Climate Audit is; how does one make observations of surface temperatures from a satellite?
I note that Steve McIntyre concocted a graph in the main article (the one I thought I could see 'phase change' between two traces). Do you know how he was able to observe surface temperature from satellite data? I thought this was only 'observable' within the infrared spectrum, but satellites are 'blinded' at surface levels in infrared! Aren't they?
quote:
Wow, the search must be on for a robust defence of NASA.
I think that NASA will just quote that 'they needed to make satellite observations conform to land based observations' and consider their actions 'justifiable' (another 'modelling' side-step).
Do you know how he was able to observe surface temperature from satellite data?
Its in Wiki.
I assume it uses Fourier transforms (when it talks of inversion) and focusses on relative frequencies that are emitted by the surface but not absorbed by the atmosphere + lots of calibration but that's only a guess.
Steve_M
quote:
More seriously, we know that climate variability can shift heat from one part of the system to another such that the surface temperatures one year can be lower than another, for example, simply because more heat may have been absorbed by the ocean in one year.
But that returns us to my question from a few days ago - where did the heat go when the surface cooled but ther's been no measurable change in ocean heat for several years?
quote:
The Hadley data does not include uncertainty. Uncertainty is present due to sampling error - we don't have global coverage
So if they don't have global coverage they have an average of where some of the weather is. So if the weather is moving around then variability is inevitable depending which way the wind's blowing (in a manner of speaking).
quote: The UK Meteorological Office's Hadley Center for Climate Studies Had-Crut data shows worldwide temperatures declining since 1998. According to Hadley's data, the earth is not much warmer now than it was than it was in 1878 or 1941.
By contrast, NASA data shows worldwide temperatures increasing at a record pace - and nearly a full degree warmer than 1880.
The reality is that looking at their graphs and looking at the means:
HadCRU figures for 1878, 1941 and recent peak are: -0.2C, 0.0C and +0.5C NASA figures for the same three periods are -0.15C, 0.1C and +0.5C
To me they seem to agree. The "trick" is to put the HadCRU data on a different scale to make the warming look less.
The 'x' against 'y' 'scale' of all graphs is expected to be representational to any 'interpretation', 'interpolation' or 'representation' of any data included within the graph! Do you suggest that these graphs are 'disproportionately' presented?
quote:
As to the adjustments, I'd be certainly interested to know the justifications. I can think of a few reasons, but I don't know enough to speculate.
Steve, there is no 'speculation' needed because this only leads to 'policy' and 'politics' doesn't lay within this posting domain.
Post script of my last post missed a word and punctuation mark. It should have read: Steve, there is no 'speculation' needed because this only leads to 'policy' and 'politics', and doesn't lay within this posting domain.
Its in Wiki as follows satellite temp measure in Wiki
I assume it uses Fourier transforms (when it talks of inversion) and focusses on relative frequencies that are emitted by the surface but not absorbed by the atmosphere + lots of calibration but that's only a guess.
This is 'obviously' 'under debate'!
That doesn't change the fact that any satellite 'observation' in the IR band that enables temperature 'observations' of 'surface temperature' can only be an 'assumed' temperature and not an 'observed' temperature'!
That doesn't change the fact that any satellite 'observation' in the IR band that enables temperature 'observations' of 'surface temperature' can only be an 'assumed' temperature and not an 'observed' temperature'!
What of 'real physics'? Smile
Does my post make sense? Smile
Yes, make sense to me. Satellite measurement is like any other proxy measurement. When you use a mercury thermomater to measure temperature you have callibrated the scale based on your knowledge of expansion of mercury. So it is no less real physics.
The 'x' against 'y' 'scale' of all graphs is expected to be representational to any 'interpretation', 'interpolation' or 'representation' of any data included within the graph! Do you suggest that these graphs are 'disproportionately' presented?
Both graphs have been obtained from the data providers site, so I'm certainly not saying they've been fiddled. I've always noticed that the NASA plot looks like the temperature rises are greater than in the HadCRUT3 plot, but when you look more closely they are reasonably consistent particularly given that they're measuring the anomalies using different methods.
SoM
quote:
But that returns us to my question from a few days ago - where did the heat go when the surface cooled but ther's been no measurable change in ocean heat for several years?
I think I said I don't know.
Temperatures cooled very quickly between December and January and now appear to be recovering quite quickly after February. Based on basic physics I'm assuming transfer between atmosphere and ocean is too slow to account for this. So I would suggest that the cooling blip was due to more heat going into space than was received by the sun - possibly due in part to the high snow cover reflecting more solar energy away.
The influence of ocean current changes is more gradual because the ocean moves slower.
But if a transfer of energy from atmosphere to the top 100 metres of ocean took place such as to cool the atmosphere by 0.1C (equivalent to the difference between 2001 and 2007 in the southern hemisphere), then the top 100m of ocean would warm by just 0.002C which is much less than the error bars of the measuring systems
This link points to the atmosphere profiles measured by the MSU satellites. The measurements of the lower troposphere appear to represent from surface up to 3 or 4 kilometres.
Great link Steve, thanks (it's now in a 'favourites' folder).
quote:
I've always noticed that the NASA plot looks like the temperature rises are greater than in the HadCRUT3 plot, but when you look more closely they are reasonably consistent particularly given that they're measuring the anomalies using different methods.
There is a reason for that and I'm surprised that you didn't notice, but they both measure different altitudes of the atmosphere as well.
The Hadley Centre take measurements at the surface of the planet (lower boundary layer of the tropo) and encounter more 'noise' (weather) in their readings from the hydrosphere, such as cooling rainfall and evaporating water.
NASA however, take measurements at a higher altitude. The lowest MSU measurement they make (from your link) is their TLT wad of the atmosphere in the lower tropo. Needless to say there is different hydrosphere activity here and they probably try to filter out any interference from the boundary layer, as this would only be 'doubling up' on a Hadley Centre product unnecessarily. Inclusion of a boundary layer measurement within this would also 'pollute' the measurement for the lower tropo by including 'noise' from hydrosphere activity at the surface.
I think that both products are good tools, but each tool has a different application for use. Though, both could probably find a use somewhere in both climate and weather applications, but one can not 'substitute' the other in an application as they are not different measurement systems of the same observation.
I can understand now why I though I could see a 'phase shift' in a graph of these two systems that I recently saw. It was probably there, but lost in clutter and lack of definition.
NASA however, take measurements at a higher altitude. The lowest MSU measurement they make (from your link) is their TLT wad of the atmosphere in the lower tropo. Needless to say there is different hydrosphere activity here and they probably try to filter out any interference from the boundary layer, as this would only be 'doubling up' on a Hadley Centre product unnecessarily. Inclusion of a boundary layer measurement within this would also 'pollute' the measurement for the lower tropo by including 'noise' from hydrosphere activity at the surface.
Sorry, the problems of trying to have two conversations at once. The remss link related to the discussion of satellite measurements of near surface temperature.
The NASA GISS data discussed in theregister piece and compared with the Hadley data is surface station data.
I thought there was probably miscommunication here!
quote:
Sorry, the problems of trying to have two conversations at once.
No, I'm sorry Steve. I'm the one that introduced the 'cross talk'! Then it seems I lost track and doubled back again.
quote:
The NASA GISS data discussed in theregister piece and compared with the Hadley data is surface station data.
Now I'm totally confused.
'The Register' compares the NASA GISS graphic detail of '1999' data with their '2007' data, with reference to Had-Crut and satellite data as a coincidental comparison (though the 'coincidental comparison' is, I think, unfair).
Good link SoM. Perhaps the IPCC should include 'hind casts' in future evaluations and this does seem to be an 'Empirical Falsification', of sorts. However, this site offers 'other interest' that 'piques' my attention, but it's all in 'Greek' and I can't understand it!
It looks like you not only get an urban heat island effect but also an urban rain effect. So I guess as we build more cities they'll wring more water vapour out of the atmosphere, cause catastrophic cooling as well as increased flooding. Also it references our friends the aerosols again. More adjustments to the models required?
"This makes future climate projections not credible. "
Good old Gavin Schmidt of realclimate sets fire to the old Trojan horse soon as he looked at it!
QUOTE Response: With all due respect to the authors, they do not appear know very much about either TAR or AR4. Looking at the statistics of local temperature and precipitation is useful but picking just a few long records and comparing to the nearest individual grid cells is not sensible. The differences in topography an local micro-climates are probably large and will make a big difference. A better approach would have been to look at aggregated statistics over larger areas. This has in fact been done though - for instance Blender and Fraedrich (2003), and there was a recent paper that looked the AR4 models (in GRL maybe? - I can’t quickly find the reference). The most curious aspect of this paper’s reception in the blogosphere is that the authors use the surface station records which in all other circumstances the cheer squad would be condemning as being horribly contaminated. Just saying. - gavin] UNQUOTE
It's a bit dated, but it follows on well and on remembering 'urban heat islands' I realised something that isn't mentioned in either article. From page 3.
quote:
When searching for ship tracks, Coakley's team looked at the near-infrared light (light which has a wavelength longer than that of red light in the visible color spectrum) coming off the clouds. At this wavelength many ship tracks appear as bright lines that can be distinguished from the surrounding, uncontaminated clouds.
Surely, if the ship tracks are 'bright' in the near-infrared spectrum they must be 'exothermic' and hotter than the surrounding atmosphere. If this is the case then in both articles the 'culprit' looks to be 'black carbon' or 'soot', yes? And we all know what an abundant variety of pollutants black carbon can carry, as well as generating heat and ozone in the atmosphere!
I don't know why we're posting these here though, this is the antithesis of this thread. Sorry Batch!
C4 Forums 
