Monday
Apr012013
by 
Josh
Josh Introducing 'The Scythe' - Josh 213
Apr 1, 2013 Hockey Stick Josh Apr 1, 2013 Hockey Stick Josh 
Buy
Books
Click images for more details
Currently discussing
-
Dec 28 - derreg
on
Curiosidade na Madrugada -
Dec 28 - kekuss
on
Rushed Office Relocation!
Links
A few sites I've stumbled across recently....
-
Science blogs
-
Political blogs
-
Other blogs
Subscribe
Copyright © 2006-2013, Bishop Hill All rights reserved.
Reader Comments (84)
BitBucket Apr 2, 2013 1:57 AM
31 cm/ka is not a density but a sedimentation rate, cm is the thickness of deposited sediment, k means 1000, a is an abbreviation for annum.
Pharos, yes I know that. But the point is that 31cm/ka is the average - at the bottom of the core it may well be around that value. But at the top, before you start coring the seabed, the sludge is more likely to be a suspension with a lot of water. And when you start coring you are going to disturb that interface. My wild guess is that the top 3cm and maybe even 10cm is hopelessly mixed up so holds nothing useful. So can you answer my questions?
BB, you appear to be demonstrating that you've not read any of Steve Mc's postings on this issue. He quite clearly explalins why the Marcott redating issue for some of the cores presents a ridiculous result for exactly the reason that you've presented here.
To me it seems pretty obvious that you can't take a noisy time series that, in effect, has been lowpass filtered with a cutoff frequency of around 1/(400 yr), splice on a low noise section with a bandwidth around 400 times as great and about 1/100 of the duration and then make meaningful comparisons between the two sections, particularly in terms of the magnitude and rise time of spikes.
Because of the short duration of the high precision section (shorter than the response time of the lowpass filter), you can't even meaningfully reduce its bandwidth to equal that of the longer section.
If anyone wishes not to see this point, they can relax. No one is going to compel them to see it.
Ross McKitrick puts it succinctly:
The authors now defend their original claims by saying that if you graft a 20th-century thermometer record onto the end of their proxy chart, it exhibits an upward trend much larger in scale than that observed in any 100-year interval in their graph, supporting their original claims. But you can’t just graft two completely different temperature series together and draw a conclusion from the fact that they look different.
Does he really? I don't read him, as you say. Pharos quotes McI (yesterday 10:21PM) saying that a bomb-test spike 3cm down proves something or other. As I say, the top Ncm are likely to be hopelessly confused; and hence so is the assertion by Pharos/McI.
BB says "As I say, the top Ncm are likely to be hopelessly confused; and hence so is the assertion by Pharos/McI"
Equally, by your logic, the same would apply to Marcott et al.
With confused proxy dating they are not in a position to make "robust" judgements about 20th Century warming.
However I do not hear you making any criticism of Marcott et al.
The mark of a "true believer".
Re: Apr 2, 2013 at 2:58 PM | BitBucket
"Does he really? I don't read him, as you say. Pharos quotes McI (yesterday 10:21PM) saying that a bomb-test spike 3cm down proves something or other. As I say, the top Ncm are likely to be hopelessly confused; and hence so is the assertion by Pharos/McI."
In response to
"Apr 2, 2013 at 2:30 PM | steveta_uk
BB, you appear to be demonstrating that you've not read any of Steve Mc's postings on this issue. He quite clearly explalins why the Marcott redating issue for some of the cores presents a ridiculous result for exactly the reason that you've presented here."
Then why Bitbucket do you feel qualified to comment -
Apr 2, 2013 at 2:48 PM | BitBucket
"McI's post seems like sour grapes. He's just jealous that his post wasn't referenced by Revkin. There's a good reason for that: Tamino wrote a better article! Y'all might not have dared to venture over to Tamino's site, but his post on the tick makes the statistics clear in a way that McI with his tables and obsession with individual cores never could. T shows that processing differences instead of actual values can compensate for proxy drop-out and that a similar, if smaller tick is visible. If a reader actually wants to understand what is what, Tamino wins hands-down.
To suggest that two highly experienced statisticians cannot independently come to the same conclusions about a paper is foolish."
Don't you think it's somewhat foolish to claim that "Tamino wins hands-down" when you claim not to have read McIntyre's post?
The typical blinkered approach of the Warmist following his religion.
BB
The top part of the series in question was box cored specifically to overcome the problem you infer and recover a section of undisturbed soft sediment.
Here is an example of a box corer
http://www.uncw.edu/nurc/systems/toolmanual/boxcore.pdf
Stop snarking everyone.
There's a lot of techinical discussions going on here that go over the heads of mere mortals like me.
I'm only a layman but this is a sort of 'hide the decline' 2.0, isn't it? Where plots of different time series are fused together to give the appearance of a contiguous data set. I'm struggling to think of an example where that would be appropriate can anyone help?
@BitBucket (1:41) Maybe some Scents of Yuma might help.
Bit Bucket (2:58 PM)
From here, discussing a particular site:
FarleyR
I think its a series suppress/cherry pick type mischief rather than a splice but it's obviously crazy to compare proxies from largely ocean floor settings to surface instrumental measurements. It would be more reasonable to use correlatable time-series of instrumental readings from similar locations of the ocean floor, but I dont think they exist?
Icons in the Climate Wars ? Take your pick.
Lots of comment on my numbers, but nobody has addressed my main point.
The last century has seen a sustained temperature rise rate almost two orders of magnitude above the Holocene average change rate, and covering the whole Holocene temperature range.
I dont care whether you use Marcott et al, other proxies or the Word of God.
You cannot just claim that this has happened before without providing evidence. A mechanism would be nice too; something which can raise surface temperatures by 1C in 50 years and then drop them again to the original level without leaving any markers in any of the proxies.
Out all day, sorry to miss the fun:
Don Keiller: the same would apply to Marcott - of course. But nobody needs proxies to judge 20th C warming, do they now?
Marion: I need no qualifications to comment. Whether my comments make sense is of course arguable. There is also a difference between saying "I don't read McI" and "I have never read a McI post" but that is probably lost on you. I have read both posts on the 'Tick' and Tamino wins by a country mile.
Pharos: thanks for the link to the box core - nice to know how it is done. But I remain unconvinced. Just as the gas content in an ice core does not necessarily correspond exactly to the position in the core (because the gas diffuses in and out of the core over the centuries while the core is near the surface), my ignorant guess is that the top of a mud core, box or not, is similarly imprecise for similar reasons.
Simon Abingdon: no idea what that is, but if anything helps, go for it!
HaroldW: thanks for the data. The trouble I have with McI (and other similar commenters) is that they start from an assumption of bad faith or fraud and that permeates their whole approach. That makes it fundamentally nasty and I don't usually feel inclined to read it. There might be something interesting in this McI analysis, but as his motives are impure, his reasoning is tainted. Y'all probably think the same of my preferred sources, so there is no possible meeting of minds.
All: in conclusion, it looks likely that the temperature over the last 11000 years has followed a curve somewhat along the lines of Marcott. I've seen nobody disputing that. Whether it jiggled up and down over the whole time or followed a smooth course (or something in between) is not addressed by Marcott and remains a subject for conjecture. There is no apparent evidence that it has jiggled, but that is not proof. What is known for sure is that the last 100 years have see a big jiggle in an upward direction. People like me think that is something to worry about.
Icon = I con ?
EM -
Please refer to my Apr 2, 2013 at 2:53 PM comment.
The bandwidth of the proxy data is too small to reveal rapid rises and falls on a 50 year scale. The lowpass filtering effect removes all the high frequency content that would be needed to see such changes.
It makes sense to me to assume there have always been fluctuations in climate over short timescales. It would be amazing if such fluctuations only started when instrumental records were first kept.
Does that convince you? [I suspect not]
"BB, you appear to be demonstrating that you've not read any of Steve Mc's postings on this issue. He quite clearly explalins ..."
Apr 2, 2013 at 2:30 PM steveta_uk
"Does he really? I don't read him, as you say. ...."
Apr 2, 2013 at 2:58 PM BitBucket
"Marion: (...) There is also a difference between saying "I don't read McI" and "I have never read a McI post" but that is probably lost on you. I have read both posts on the 'Tick' "
Apr 3, 2013 at 12:53 AM BitBucket
BB - please be consistent in your stories about what you have and have not read.
I suggest avoiding unpleasant condescension in such circumstances.
Martin A: Lol. In fact, BB, we prefer consistency without condescension. But if has to all or nothing, let's at least have some consistency.
Martin, have a gold star - you are a champion nit-picker. You clearly missed your true vocation; Auditor. Maybe you and Marion think I just guessed at his use of a table and individual cores for his tick post when I said, "... his post on the tick makes the statistics clear in a way that McI with his tables and obsession with individual cores never could".. Childish stuff.
You might not be able to admit it, but I am sure even 'sceptics' come away from reading Tamino's post understanding exactly how to compute the curve without exaggerating the tick. Not so with McI, as his agenda is not to explain but to accuse.
I don't get it BB. I read all the McIntyre posts, in close to real time, and didn't find them difficult to follow. Why don't you tell me what I missed by not reading Tamino. What did he add?
Did you read it? You can see clearly that using a differencing method of calculating the curves gives a much better impression of the data (it shows the up-tick to be really quite small but not an artefact).
BB: thanks but I just wrote "Why don't you tell me what I missed by not reading Tamino." This implies I didn't read it. I think some comprehension training might be in order.
You are getting too focused on the uptick at the end of Marcott et al's data and not on the larger context.
After the hockey stick and the scythe I give you the wheelchair (and I bet Josh will have fun with that one).
http://neven1.typepad.com/.a/6a0133f03a1e37970b017ee9bca1cf970d-pi
This was compiled by Jos Hagelaars.
http://ourchangingclimate.wordpress.com/2013/03/19/the-two-epochs-of-marcott/
The first part, the green curve from Shakun et al, starts 20,000 years ago during the warming that preceded the end of the last glacial period and ends at the transition to the Holocene 11,000 years ago. This period begins with ice sheets covering Canada and most of the UK and ends with a climate resembling 1900AD.
The second part, the blue curve from Marcott et al begins with the onset of the current interglacial and continues to the 20th century. It includes the peak and the gradual decline expected as an interglacial declines towards the following glacial period.
The third (red) curve is Hadcrut4, the Met Office instrument record for the 20th century and the present.
The fourth curve is IPCC A1B, the middle-of the-road forecast to 2100.
This is why Bitbucket, myself and many others are concerned. What started as a normal interglacial is going in a most unusual direction.
Entropic man:
So true. And can you show us three places where you have made precisely this point to those who hyped the same uptick in terms such as "We're Screwed" and "Climate Game-Changer"? We're on a consistency drive now and none of us should be exempt from that.
Richard Drake
Then focus on the evidence and stop point scoring.
I'm taking that as a no: no interest whatsoever in correcting the record where it matters most. We agree on the most important evidence I'm sure: that Marcott's original uptick was completely specious. I have no idea therefore why you're not out putting the record straight in 300 existing publications that got this wrong and misled ordinary people about the threat of current warming trends. I only asked for three examples and that wasn't to score points - it was game over if you hadn't been bothered even to that extent.
Richard Drake, are you afraid of reading Tamino? Perhaps you fear that your hero McI will be found wanting. Nobody is being misled as there really is an up-tick - your pal Mr Guenier makes it plain that all the sceptics he has ever come across (and that must include you) believe that temperatures have risen since the 19th C. And you can see it from the instrumental record. The only debate is whether such rapid up-ticks have occurred in the historic record and I'm not seeing any evidence of that from you.
Under scrutiny, Marcott now admits his hockey stick paper - differing from his thesis - is bollox ('not robust').
Tamino, though, thinks it is. As does BitBucket.
Weird eh?
Before being so smug, take a look at Tamino's post on the likelihood of up-ticks in the 11000 year reconstruction. He finds that if there had been an up-tick such as that in the 20th C (which we know has occurred, independent of Marcott), they would show up in the Marcott curves.
So no Marcott uptick-away-smoothing then .. ?
Go and read the article if you don't understand
Pleading the Fifth, eh?
This is the REAL hockey stick:
http://stevengoddard.files.wordpress.com/2013/03/screenhunter_256-mar-02-06-55.jpg