Clever bees

I suppose this really belongs on the bee blog, but anyway... from Science: honeybees, who have 0.01% of the neurons that humans do, can recognize and remember individual human faces (thanks to JF for the tip).

Bryan Lawrence has a blog

I discover that Bryan Lawrence (head of BADC) has a blog: http://home.badc.rl.ac.uk/lawrence/blog. A mixture of climate and software... should be perfect for JF.

The Economist on Climate Change (sigh)

The last few issues of the Economist have seen a few climate change type articles. One of leaping penguins even made the front cover (headline "Don't Despair: grounds for hope on global warming"; however the grounds for optimism they find are thin: some grassroots action, and hints of voters changing their minds). The Economist (of course) isn't a very good source for the science of GW; its written by economist-types (oddly enough) not scientists. And they have their biases: mostly a free-market liberalism which makes them rather dislike the idea of anything that won't fit within that framework and which might badly strain it. I quite like their general tone usually: I have pinned up in my office two of their front covers arguing for greater immigration, just next to a nearly interchangeable one from Socialst Worker, which I found ironic; one of the first I saw was in favour of same-sex marriage: the Economist, whilst very free-market, is by no means std.right-wing.

As a side note, the most recent edition notes that Lee Raymond is going, which could well be good news.

The Economist has a good reputation in general, and is widely read by business-politician type folks, so we have to care what they say and how they say it. In particular, first paragraphs matter, because many people won't read past them. Which is why the 7th of Dec (or, in the paper version, 10-17th Dec; thanks CH) article is so bad:

THE climate changes. It always has done and it always will. In the past 2m years the temperature has gone up and down like a yo-yo as ice ages have alternated with warmer interglacial periods. Reflecting this on a smaller scale, the 10,000 years or so since the glaciers last went into full-scale retreat have seen periods of relative cooling and warmth lasting from decades to centuries. Against such a noisy background, it is hard to detect the signal from any changes caused by humanity's increased economic activity, and consequent release of atmosphere-warming greenhouse gases such as carbon dioxide.

This is std.septic.sh*t*. Not because its false, but because its misleading. Try this:

People die. They always have and they always will... therefore we shouldn't worry about whether to fund the health service or worry about cars on the roads or terrorists, its just more or less death.

A more honest intro would reflect the std.consensus: that the recent climate change is likely to be unusual and likely to have been caused by people.

The rest of the article isn't too bad: somewhat skeptic (note we've got the k back now) but not too bad. E.g.:

The third finding is the resolution of an inconsistency that called into question whether the atmosphere was really warming. This was a disagreement between the temperature trend on the ground, which appeared to be rising, and that further up in the atmosphere, which did not. Now, both are known to be rising in parallel.

Parallel is wrong, to be picky: the tropospheric trend should be larger, and is.

In case you're wondering, #1 was that its been warm recently, and #2 was the Arctic. #4 is detection of warming in the oceans; #5 is a bit dodgy in their words: The fifth is the observation in reality of a predicted link between increased sea-surface temperatures and the frequency of the most intense categories of hurricane, typhoon and tropical storm. If I were you, I'd read RC. #6 is the THC (again you want RC).

After a slightly dodgy solar bit, they continue with That the climate is warming now seems certain. And though the magnitude of any future warming remains unclear, human activity seems the most likely cause. The question is what, if anything, can or should be done. And thats a fair question. Too rapid or too great a warming, though, risks serious, unpleasant and in some cases irreversible changes, such as the melting of large parts of the Greenland and Antarctic ice caps. There is, to put it politely, a lively debate about how far the temperature can rise before things get really nasty and how much carbon dioxide would be needed to drive the process. Unfortunately, existing models of the climate are not accurate enough to resolve this dispute with the precision that policymakers would like. Again, pretty good, apart from that last bit (to me it implies that the poor old policymakers are just sitting there wondering when the GCMs will tell them what to do, which is nonsense: they all have agendas of their own).

Then lastly If greenhouse-gas emissions are to be capped, however, a mixture of political will and technological fixes will be needed. Seems fair to me, but we're heading out of my territory with that, so I'll just observe that political will seems distinctly missing, to me. I'm aiming for a post on Montreal soon.

Connolley has done such amazing work...

Back to wikipedia... Nature has an article on wikipedia vs Britannica. It was an interesting exercise, and as the most notable climatologist on wiki :-) they interviewed me, which lead to the sidebar article "Challenges of being a Wikipedian" (see the Nature article; click on the "challenges" link near the bottom). It contains the rather nice quote from Jimbo Wales "Connolley has done such amazing work and has had to deal with a fair amount of nonsense" (does Lumo still read this?).

What Nature did was to take a number (50; of which 42 came back usefully) of wiki and Britannica articles, and send them out to experts for review. There was a fairly severe constraint on this: that the articles had to be of comparable length in the two sources; which is why I think no climate change type articles were done. I strongly suspect that if you try to find anything about, say, the satellite temperature record in Britannica it will either be entirely missing or badly out of date. The list of articles is here.

There were 8 serious errors in both sources. Then we move onto more minor inaccuracies. The oddest thing about this is that the average number of errors in Britannica is 3 and wiki 4; and Nature (genuinely) expected us to be *pleased* about this, as though being nearly as good as Britannica was something to be happy about! I rather suspect that this may be due to the choice of articles to some extent. The GW articles don't contain many errors (except the septic cr*p, sadly we can't get rid of it all :-().

The most pleasing part, though, is the accompanying editorial which actively encourages scientists to contribute (James, are you listening?): Nature would like to encourage its readers to help. The idea is not to seek a replacement for established sources such as the Encyclopaedia Britannica [oh yes it is... WMC] , but to push forward the grand experiment that is Wikipedia, and to see how much it can improve. Select a topic close to your work and look it up on Wikipedia. If the entry contains errors or important omissions, dive in and help fix them. It need not take too long. And imagine the pay-off: you could be one of the people who helped turn an apparently stupid idea into a free, high-quality global resource.

[Update: there is a Nature blog here and this includes a list of the errors in the EB and wiki versions; see-also [[Wikipedia:External_peer_review/Nature_December_2005/Errors]]. The Nature blogs report on Jimbo's visit is interesting too. And [[Wikipedia:Requests_for_arbitration/Climate change dispute 2#Removal of the revert parole imposed on William_M._Connolley is nice to have...]

News from NZ...

I'm back. And to celebrate, here is a story from a local paper over there. My apologies to all the good folk of NZ, this is not a fair reflection of your country, but it is very funny, I'm thinking of sending it in to Private Eye.

Also, a joke: what do you call a woman who stands between goalposts? A: Annette. And by odd coincidence, Mt Annette was the peak I climbed from the Mueller hut. More on that in the photo-essay to follow "soon".

Also, a note: I'm switching comments to "only registered users" just as soon as I can work out how to do it. I'm a bit fed up with anonymous trolls, named trolls are so much better...

[Update: BL points out the obvious: that Cairns is in Australia (the "West Island") not NZ. Oops. I knew that... He also says that he would have posted that here, except I insist on only registered users. So for the moment, I'm turning that off again, since he is the second person to somewhat dislike that feature, and on reflection I don't need it]

Dunedin: sea ice conf

I haven't blogged much about this conf. Mostly because there is no wireless access (dinosaurs...) but also because much of it is deeply technical sea ice stuff of rather limited general interest. Which is in itself a point of interest: the amount of climate change related stuff is small. A few people have shown the std.pic of Arctic september ice, which shows decline (no-one has shown a similar for the Ant) but only a few people have done anything with it (trying to look at the changes in different ice types: first-yeay, multi-year; but then this is tricky). Also there are only a handfull of papers using climate models.

So what is it actually about? A major theme is sea ice depth (ice fraction is fairly easy (err, thouigh see my brilliant presentation deomonstrating that there are problems even there), depth is much harder). Lots of people are using satellites (radiometer on ERS-2; laser on ICEsat) or helicopter or ship or sfc bourne methods to estimate ice depth. The problem is that while its fairly easy to drill a hole in the ice and measure depth at a point, to get an area value is much harder. EM (electro-magnetic) sensors can detect the water level (on a ship they need to be only 3m above the ice; hung from a helicopter they can be 10 m above the ice, with the heli another 20 m higher up, which apparently makes for exciting flying). So those get you transects. From satellite, you can measure ice freeboard (radar) or top of snow (laser) if you can find enough leads to reference the values to a sea level; this is a major problem. Also measurements from underneath: the late lamented autosub; some stuff from military submarines (their CTDs didn't work so they used the entire sub as a billion pound CTD) which at true cost would be incredibly expensive, but since they are there anyway (not really clear what they *are* doing) they can do some science, even if their sounding kit is a bit dodgy for science.

Apart from that, various things: properties of ice; today a pile of talks about the ice formation mechanism, which isn't really my thing: platelets and congelation and frazil and so on. Special mention for the chap running a molecular simulation of ice formation: with 1,500 molecules his simulation of the 9 ns it takes to freeze took 4 days processing; his value for the freezing temp is 271 (+/- 9) which he regards as extremely accurate. How ice freezes from underneath; new ice dynamics models; measurements from campaigns; etc etc.

Last night we had the conf dinner in Lanarch "Castle" a magnificent but truely fake building, more of a manor house or small chateau. And we had the piping in of the haggis and address to same. And the drinking till midnight.

Note to PH: yes the foxgloves are non-native. But they are lovely.

Road deaths and terrorism

Continuing an old theme, but this time with some actual numbers, via Nature:

114 deaths per million people occurred in road crashes in 29 countries in the developed world during 2001.

0.293 deaths per million people were caused by terrorism each year in the same countries in 1994–2003.

390:1 is the ratio of road deaths to deaths from terrorism.

Source: N. Wilson and G. Thomson Injury Prevention 11, 332–333 (2005).

And if I didn't mention it before, I'm currently in Dunedin because of this.

NZ: eternal sunset

I'm in NZ, Dunedin. And it turns out they do have internet here. To prove it, here are some pics of the flight over.

Heathrow to LA was good: we had a long slow sunset as we headed north, sunrise as we went due W, then sunset again into LA. Its a shame they don't have a better quality "photography" window in the back somewhere. We got to see greenland (briefly; the W side; the E side was in cloud) and sea ice over Hudson bay (see pix: this is the firt sea ice I've ever seen in real life; its from 11,000 m) and the vast expanses of frozen Canada. LA to Dunedin is 12+ hours; I managed to sleep much of it thankfully.

Trivia: on the flight into LA, we had all-plastic cutlery. Out of LA, we got metal forks and spoons.

Catherine Bennett can FOAD

As if the gulf stream stuff wasn't enough to wind me up, the Grauniad published Catherine Bennett Climate march, but will it work?: Going on the climate change protest this Saturday is like marching for niceness - and just as ineffectual. Though the first headline is only in the online edition. The article itself is just blather; she has nothing to say; I interpret it to mean that she has grown too old and fat to bother, and has nothing better to do than mock people who do care.

Anyway, on that temperate note, I'll sign off for the moment, and probably for the next two weeks, unless NZ is connected to the internet.

ps: thanks to those who commented on the poster, I corrected most of the typos.

"Alarm over dramatic weakening of Gulf Stream"?

By now you'll all have read the RC post: Decrease in Atlantic circulation? (see, thats where I got my question mark from); which is about Bryden et al. in Nature. RC, of course, has a nice science analysis; I just wanted to compare it to the Grauniad story: Alarm over dramatic weakening of Gulf Stream. Now the headline is nonsense, because the Nature paper itself sez: the northwards transport in the gulf stream across 25 oN has remained nearly constant. Something more complicated is going on (return flow shallower, hence warmer, hence overall heat transport N is less because more is coming back S), which I'm not going to explain because (a) its over at RC and (b) I haven't read the paper properly yet (I look forward to doing so tomorrow during my long flight). It looks to me like it was too complicated for the Grauniad sci writers.

There are I think various caveats to interpreting this: most notably (as RC notes), that if this really has already happened, you might expect some signal in the SSTs which doesn't seem to have been seen.

Also, the Nature article and the Nature commentary are a bit selective in their reading of GCM results to support this.

[Update: also see James Annan's take and wise words re Nature]

Reading the entrails: New Nukes?

The BBC says:

Blair says nuclear choice needed

Tony Blair says "controversial and difficult" decisions will have to be taken over the need for nuclear power to tackle the UK energy crisis.

The prime minister told the Liaison Committee, made up of the 31 MPs who chair Commons committees, any decision will be taken in the national interest.

He is said to believe nuclear power can improve the security of the UK's energy supply and also help on climate change.

A government review of energy options is expected to be announced next week.

I like the any decision will be taken in the national interest. This fails the try-negating it test: any decision will be taken against the national interest is unsayable. So ItNI means "prepare for an unpopular decision".

Although there is always a techno-industrial lobby in favour of Nukes, I'd guess that and also help on climate change may be quite accurate. Blur has been talking about Kyoto options and as I noted I think the govt has realised we're (they're?) not going to hit our targets. So he needs to pull something out of the hat. These nukes won't do it: they won't be onsteam by 2012 unless they arrive rather fast; but they could probably be folded into the plans if pushed.

So... is this a runner? Lots of people don't like nukes: Greenpeace protesters have disrupted a speech used by Tony Blair to launch an energy review which could lead to new nuclear power stations in the UK. Two protesters climbed up into the roof of the hall where Mr Blair was due to address the Confederation of British Industry conference. After a 48-minute delay, Mr Blair made his speech in a smaller side-hall. Forcing Blur off into a side-hall is a success, and will have annoyed him a lot.

But the "debate" about their (de)merits is as poor as ever: at least judging from radio 4 this morning. We had someone who doesn't like nukes, and then Bernard Ingham who does (I think, like Bellamy, out of an unstated assumption that its Nukes or Windfarms and he doesn't like windfarms). The green chap said Nukes are uneconomic; BI said they are. I rather suspect that they aren't, under current conditions: our present Nukes barely manage to stay afloat even with all their building costs written off; and I don't see piles of commercial applications waiting to be built. Of course some of this is due to the endless wrangling which costs; and how to cost the long term storage is obviously a bit of a poser since no-one yet knows how it will be done.

One of the arguments that the Green side is starting to push is that Nukes aren't that good for CO2: that over their lifecycle, they emit lots, comparable with coal/gas. I rather doubt that makes sense. I've never seen the figures. If it *is* true then it would account for the economics being so bad. If anyone has them, do please leave a comment.

You'll have noticed that I haven't explicitly given my opinion on this, though which side I lean should be clear enough. I excuse this by it being far from my expertise: I'm not sure why you should want my opinion. If offer this observation, though: that through the years on sci.env I have observed that the people in favour of Nukes invariably know more about them, and those against know little. Blur is likely to be an exception to this, though.

Sea ice: what I do in my spare time

Fairly soon now I'm off to NZ (oh dear, my CO2 burden...) to present some sea ice work. The poster part of it is nz-hadcm3.pdf. I have a day or two left, so feel free to point out typos and gross scientific errors.

The theme of the work is upgrading the sea ice dynamics in HadCM3, which has occurred just in time for it to be replaced by HadGEM. Never mind, we learnt a lot in the process. Mostly we learnt how hard it is to force the sea ice to behave itself in a coupled model.

The poster (in theory) says it all, so I won't explain at length here: but feel free to ask questions...

Topping Punts

There is an air of "tipping points" about. This is an idea (possibly coined by Schellnhuber) where "the balance of particular systems has reached the critical point at which potentially irreversible change is immenent, or actually occurring". That quote, somewhat bizarrely, comes from the Books and Arts section of Nature (here), which is a slightly dodgy regular section where they make a feeble stab at pretending the "two cultures" ever talk to each other.

And so the piccy is S's attempt to find an "icon" for climate change. But (ibid) "the issues surrounding climate change are extraordinarily complex. Can an image be found that is both simple and good science? Given the contentious nature of the debates, particularly in the United States, it is unwise to offer hostages to fortune by parading vulnerable predictions". I don't think the image is simple: is it good science?

But first of all, what about the "tipping points" concept anyway? I've previously pushed the idea that the climate is stable (in the absence of perturbation). You could argue, quite plausibly, that we shall soon have emitted enough CO2 to raise the T enough that we will be committed to melting Greenland. Perhaps that counts as a tipping point. But its slow. Its on the map as "instability of the greenland ice sheet" which is an odd way of phrasing it but has the "virtue" of implying speed.

But enough quibbling. The one I reacted badly to was "Antarctic ozone hole". Its an envoronmental icon, but hardly a tipping point: as far as its known its reversible, and on a long slow trend to being reversed (err, as long as GW doesn't cool the stratosphere too much...).

As to all the rest... I dunno, its a bit vague isn't it? I'm not sure I'm too keen on this search for an icon stuff.

A couple of BTW's to finish off: (1) I'm down to wiggly worm, so it looks like status is based on snapshot rather than accumulated - must get posting again. (2) I'm off conferencing for a while at the end of the week, so will be dropping further down. (3) I may get assimilated by the Borg in the near future anyway... Mark seems to have self-assimilated.

Campaign against Climate Change - march Dec 3rd

A foray into explicit politics: promotion for the Campaign against Climate Change and the London march on Dec 3rd.

The Parker Paper

The Parker UHI paper (see [[Urban Heat Island]]) from Nature 2004 (and the Peterson 2003) strengthens the TAR contention that the UHI isn't important; and perhaps negligible. Now RP Sr has taken a shot at it. Unfortunately his paper is... difficult. You can take his word for what it says if you like, but I'd rather not. Happily, RP is so confident of his position that he has followed up with a whinge about Nature rejecting him, which includes the reviewers responses: Pielke has failed to adequately assess whether there are any trends in windiness in the Parker data set. Parker stratified by wind conditions, both at rural and urban sites, so any trends in windiness (even if this were possible in a stratified data set) would occur both at rural and urban sites. To suggest that there would be different turbulent mixing at rural and urban sites would then require differences in trends in temperature to be found, which is exactly what Parker found not to be the case. The logic presented in Pielke’s comment is circular and incorrect is the briefest.

One day I may actually read it, or meet someone who has. Until then I don't have a good way to assess it.

Its cold and Scott Adams gets whacked by Dogbert

Today we had the first (and who knows, maybe the only) snow of winter. Just a flurry; nothing settled, sadly.

Meanwhile, although I really like Dilbert, it looks like Scott Adams needs a whack from Dogbert to chase out the demons of stupidity aka ID/Creationism: via some rather circuitous routes I found Stein and Wolfgang.

And while I'm here, there is nice blog starting by Robert Friedman about his trip to the South Pole. Take the virtual tour!

Grauniad: Sea level rise doubles in 150 years

Yes, back to the familiar old topic: bashing science coverage in the papers. This time that old lefty favourite, the Grauniad, which has an article on Sea level rise doubles in 150 years. Who have discovered that Global warming is doubling the rate of sea level rise around the world... The oceans will rise nearly half a metre by the end of the century... Scientists believe the acceleration is caused mainly by... fossil fuel burning... during the past 5,000 years, sea levels rose at a rate of around 1mm each year, caused largely by the residual melting of icesheets from the previous ice age. But in the past 150 years, data from tide gauges and satellites show sea levels are rising at 2mm a year.

To which the obvious reply is "is this supposed to be news"? Slightly garbled of course (satellites say 3mm/y; the longer time tide gauge record is ~2 m/y; see the wiki [[Sea Level Rise]] page and the refs to the IPCC therein). The other interesting bit of garbling is the 1mm/y over the last 5kyr... the TAR says Based on geological data, global average sea level may have risen at an average rate of about 0.5 mm/yr over the last 6,000 years and at an average rate of 0.1 to 0.2 mm/yr over the last 3,000 years. So, *if* they haven't garbled it, they story is that the folk from Rutgers University have upped the estimates of SRL over the last 5kyr. But I'd bet on garbling myself. The abstract from Science is here but I can't read the full contents (I had an offer of a subs for $99/y and am considering taking it up...) but it seems to be more interested in the Myr timescale.

The Grauniad also covers the latest EPICA stuff, but thats much better covered over at RC so you should go there for that.

[Update: thanks to the kindness of not one but two readers, I now have a copy of the article from Science. In true blog style, I've quickly skimmed it far enough to discover that the 1 mm/y over the last 5kyr is a bit of a sideshow, and fortunately for you, its in the supplementary online material freely available. They say "Sealevel rise slowed at about 7 to 6 ka (fig. S1). Some regions experienced a mid-Holocene sealevel high at 5 ka, but we show that global sea level has risen at È1 mm/year over the past 5 to 6 ky." So I must apologise to the Grauniad: no garbling.

So how do we reconcile that to the pic I show (which is from wiki, not the Miller paper)? Both show a rise of about 15m over the last 8 kyr. The wiki pic has that very steep (15 to 4-) from 8 to 7 kyr; then much shallower. The Miller article fig S1 starts a bit deeper and has a much more uniform slope. Since the Miller data is almost entirely from one area and appears to contradict what I think I already know, I'll stick with wiki and the TAR for now. But informed comment is welcome. I do find it a teensy bit surprising that the Miller paper doesn't comment on the discrepancy between their Holocene results and "accepted wisdom": its possible I have the AW wrong.

Update 2 (minor): switch href on the figure to the wiki page]

Stability in a control run of HadCM3

One of the things I do is port [[HadCM3]] to new platforms (although I shouldn't over emphasise my role in that: much of the hard work of portabilising it was done at the Hadley Centre; nonetheless new platforms throw up new problems). HadCM3 was written for a Cray T3E; its known to be stable when run without forcing for thousands of years on that platform. There is a portable version of the model, which requires a little bit of effort to make it run on new platforms. The first thing to do is make it compile; the second to make it run through the first timestep; the third through the first meaning period; and then hopefully all that remains is to check that it is stable.

Which brings in this picture. Black is a 200 year control run, with the g95 compiler on a 4-processor Opteron system (using 3 procs for most of the time). Blue is a rather older run on an Athlon system under the antique fujitsu/lahey compiler. Red is an in between run on Opteron with the Portland Group compiler (pgi). All are seasonal data, differenced from 100 year means of an "official" control run. What you'll notice is that the red run has a distinct climate drift, which is enough to make it unusable. Blue looks OK; black has been run out long enough to be sure its OK. The grey shaded bit is some kind of 95% confidence limit based on the variability of the 100 year "official" run.

Quite why the Opteron/pgi runs drifts I don't know. Its 99.999% the same code as the other runs (differing only in whatever it took to make the compiler accept it). Most likely there is some compiler bug in there; but I will probably never know.

By eye, the 200 year run has no drift. By line fitting, the results are:

0- 50: [ 0.00168505, 0.00451390]
0-100: [ 0.00052441, 0.00159977]
0-150: [-0.00050959, 0.00009494]
0-200: [-0.00060853,-0.00015253]

where I've shown the (95%) confidence intervals for a line fit over the first 50, 100, 150 and 200 years. Which shows up the internal variability quite nicely. If I'd just taken the first 50 years I might have believed in a drift of 0.3 oC/century, which is small but not perhaps totally negligible. By 100 years the "drift" has a central value of 0.1 oC/Century which would be negligible. Out to 150 years there is no statistical trend. Out to 200, a trivial cooling. Note, BTW, that all these sig estimates are rather thrown-together and should be a bit wider to take proper account of autocorrelation.

The mirror world

RP has what I regard as a posting full of mistakes: Reflections on the Challenge (my post The Big Picture refers). And he doesn't get any better in the comments.

One of mine got through. This one, below, got stopped for "questionable content" - judge for yourself - so since I have my own blog I'll post it here.

Roger - you're still getting it wrong; Tom Rees is essentially right.

You say "So your position now is that the hockey stick was in 2001 a key study in making the case for attribution. That is, that without the hockey stick the case for attribution in 2001 would have been somewhat weaker? I disagree."

No. I didn't say *key*. But I *do* say that without MBH the attribution case in the TAR would have been *somewhat weaker* (but not *very much weaker*). [Good grief], you can just read the thing (surely youre familiar with it): http://www.grida.no/climate/ipcc_tar/wg1/007.htm. Which makes it clear that MBH is part of, but by not means the whole of, the attribution case.

Yes, MBH wasn't in the SAR, but then as the TAR sez "Since the SAR, progress has been made" and MBH was part of that progress.

If you want to position yourself as some kind of referee in this [bizarre] process, you need to be much clearer about the structure of things.

The words in []'s are ones I experimented with deleting in the hope of getting past the content filters. No such luck.

Stoat related fun!

There is some good stoat-related fun over here (thanks to Jim E).