Interesting analysis by DSP specialist of global temperature.
http://wattsupwiththat.com/2013/09/...mperature-data-suggests-global-cooling-ahead/
Unlike climatologists this guy knows how to analyse data.
The climatologists do have the advantage of knowing something about the processes generating the data presented.
Atmospheric CO2 levels stuck at about 280ppm from the end of the last ice age to the start of the Industrial Revolution, but they've risen exponentially since then, to the current 400ppm. The halfway point - 340ppm - happened around 1980.
This means that half the anthropogenic green-house warming we've seen is concentrated in the last thirty years. It's visible in the curves presented by Anthony Watt, but it's equally clear that the warming we've seen is of the same order as the noise imposed on the global temperature signal by the Atlantic Multidecadal Oscillation (AMO).
The DSP analysis pulls out a 57 year period. The AMO doesn't seem to be a stable oscillation, but its characteristic period - in so far as it can be said to have one - is something of the order of 70 years, which is close enough to 57 years for an analysis of 160 years of data, particularly when the the last thirty years is bent up by rapid anthropogenic global warming
Watt talks about a 170 year component, but that's just over-fitting the data - he's trying to fit a sine-wave to a hockey-stick. Extrapolating on the basis that a hockey stick is actually part of a sine wave isn't all that clever.
A serious data analyst wouldn't make that kind of mistake - the DSP analysis is clearly tongue in cheek - but Anthony Watt isn't into data analysis, but rather into giving his audience the kind of story that they will be happy to hear.
Interesting analysis by DSP specialist of global temperature.
http://wattsupwiththat.com/2013/09/...mperature-data-suggests-global-cooling-ahead/
Unlike climatologists this guy knows how to analyse data.
The climatologists do have the advantage of knowing something about
the processes generating the data presented.
[...]
Thanks for your masterly analysis.
It's nice to see you haven't lost
any of your religious fervour for CO2 induced warming despite a 17
year pause in warming.
I spent too long curve-fitting data to set up working curves to have many illusions left about the process. For the serious stuff I found out how to put confidences limits on the multiple parameters I was sucking out of the data, and if I tried to extract more parameters than the data would support, the confidence limits got remarkably large - as one parameter went up, another parameter could go down to keep the fitting curve more or less on the data.
Educational.
"Religious fervour"? Since you don't process rational arguments, they might look just like dogmatic assertions to you. Don't worry. God still loves you even if you can't make sense of his creation.
The 17-year "pause" in the warming looks rather like the other bobbles the Atlantic Multidecadal Oscillation imposes on the record.
Anthony Watt has very kindly given you 160 years of bobbly data to look at.
Admittedly, the bobbles included prolonged negative-going bits in the bits before 1900, while they now are now only "pauses" in a largely upward progression - so you've got to be a seriously a head-in-the-sand denialist like Anthony Watt to see them as some kind of negation of an underlying upward trend.
He'd get less support from the Heartland Foundation if his perceptions were a little better schooled. What's your excuse? Nobody is going to pay you for ventilating your ill-founded delusions, so you've got to be compensating for an insecure childhood or something equally unfortunate.
did you notice his graph 13?
2005 and beyond are cooling according to him - but the global averages
were amongst the recorded highest. Great model he came up with - stuffed
and proved wrong before he even posted.
So when Ben Santer said it needs 17 years to see global warming and
there hasn't been any for 17 years?
https://www.llnl.gov/news/newsreleases/2011/Nov/NR-11-11-03.html
/quote
The LLNL-led research shows that climate models can and do simulate
short, 10- to 12-year "hiatus periods" with minimal warming, even when
the models are run with historical increases in greenhouse gases and
sulfate aerosol particles. They find that tropospheric temperature
records must be at least 17 years long to discriminate between
internal climate noise and the signal of human-caused changes in the
chemical composition of the atmosphere.
/end quote
The RSS temperature record shows no warming for 16 years 8 months.
http://wattsupwiththat.com/2013/08/25/rss-flat-for-200-months-now-includes-july-data/
So far no sign of a "signal for human-caused changes".
Believing in CO2 warming when there is no warming is pretty weird.
Interesting analysis by DSP specialist of global temperature.
http://wattsupwiththat.com/2013/09/...mperature-data-suggests-global-cooling-ahead/
Unlike climatologists this guy knows how to analyse data.
The climatologists do have the advantage of knowing something about
the processes generating the data presented.
[...]
You mean like it's an opportunity to inject prejudice and obfuscation?
Even if Watt's analysis may have problems, in that it seems to
expose some periodic signals that are suspiciously close to the
length of the sample and its harmonics, the analysis should in
principle be objective, impartial, transparent and public.
The official climatologists' story is none of that. Their
arguments and data are tainted by money and politics.
Even
Watt's raw input data can't be trusted because of where it
came from!
That was a political and not an engineering comment.
Because recent temperatures are high doesn't stop there being a
downward trend since 2005:
http://www.woodfortrees.org/plot/hadcrut3vgl/from:2005/plot/hadcrut3vgl/from:2005/trend
Clear downward trend on HADCRUT3 which is the dataset he analysed.
Projecting future climate by regression analysis, over a short timebase, is
nonsense. Doing it by computer simulation, as most climatologists do, is just
more-compute-intensive nonsense.
"At least 17 years". Back in 2003, the significance of the Atlantic Multidecadal
Oscillation hadn't been fully grasped - that press release talks about the El Nino/La Nina oscillation, which is the same kind of thing, but cycles a lot more rapidly.
It used engineering data to make a political point.
But if you use a little more of the data set, it becomes an upward trend. If you go back to 1850, it becomes a totally unambiguous upward trend.
It's all nitpicking if you take the record back for a million years or so - as we now can - and can see the alternation between ices ages (CO2 levels of about 180ppm) and interglacials (CO2 levels around 280ppm).
The current scinetific understanding of what has been going on over the past 2.6 million years (which Raveninghorde has yet to get his head around) provides a useful perspective on today's temperature excursions.
So the science was settled but the alarmists hadn't grasped the
significance of the AMO.
I corrected a factual error by David. Don't see that as an attack on
your religion.
You didn't correct a factual error. You indulged your usual enthusiasm for picking the time interval over which to estimate the "trend" in order to be able to present the result you thought that David should have seen.
This is a fairly serious error in technique. The planet wasn't created in 2005, and if you want to work out what's going to happen to it over the next century or so, you need to understand what been happening over the past few million years, rather than concentrating on the last eight or 17 years.
There's loads of evidence available on the subject - ice cores, layered lake and ocean sediments and so forth. Take a look at it sometime. It won't be a revelation - science lets you make up your own mind about the message - but even you might find it educational.
Bill you need new glasses.
David said:
/quote
did you notice his graph 13?
2005 and beyond are cooling according to him - but the global averages
were amongst the recorded highest. Great model he came up with -
stuffed and proved wrong before he even posted.
/end quote
I pointed out thst David was factually wrong as HADCRUT3 temperatures
have been falling since 2005. Graph 13 was not invalidated.
No. You need a new brain.
No. You ignored what he wrote - which was that the global averages were amongst the higher ever recorded (note the plural "averages") - and told us to pay attention to your "fact", which was that the trend line through the period was decreasing.
Huh???
Just because temperatures are amongst the highest recorded does not
mean there is not a cooling trend since 2005 anymore than the fact
that a sine wave can be near it's highest voltage and the voltage is
reducing.
The point is David claimed: "Great model he came up with - stuffed and
proved wrong before he even posted." when the model is consistant with
the facts.
John Larkin sees climatologists doing computer simulations in the same way that he uses LTSpice, running the simulation for just long enough to produce the behaviour he wants to see then stopping the simulation before the circuit can move off in a direction he doesn't fancy.
Climatologists run lots of different simulations, using a wide range of different models. Nobody thinks that any one model predicts physical reality, but by comparing lots of simulations they do get some kind of feel for what's going on and can deliver a rough idea of the way the climate is going to evolve.
Crude and imprecise as their predictions are, they are a lot more reliable than John Larkin's wishful thinking, and provide a persuasive case for us to trade off a bit of economic growth in favour of getting more of our energy from renewable sources.
