Saturday, September 30, 2006

My secret life

Those of you who know of me only as a climate scientist might be surprised to hear of my alter ego as an amateur engineer and forensic scientist, which I was reminded of via email this morning.

It relates to a design flaw with modern mountain bikes, which occasionally causes very serious crashes (the front wheel becomes detached, typically at high speed which results in the rider plumetting head-first into the ground). Like all intermittent faults, the frequency of the problem is debatable, but at least one person is now in a wheelchair for the rest of his life, and several more had serious (life threatening) head injuries which resulted in lengthy hospital stays (this guy in particular was very lucky to live). I won't bore you with the full technical details, which can be found on my disk brake and quick release page.

Anyway, the relevance of this to me is that I was the one who firstly worked out that there really was a problem, and then analysed the system sufficiently well to diagnose the underlying cause. I did this against a background of sceptical cyclists telling me that I was making it all up. (Front wheel retention has a long and chequered history and in the past failures could be correctly blamed on user error, but modern bicycles have introduced a crucial element into the mix.) It took me a few months to put the theory together in a complete and convincing manner, at which point most of the big names in the field (ie independent engineers with an interest in cycling) quickly endorsed it, but the industry has continued with implausible denial and obfuscation for several years since. (I've found the process to be useful training in dealing with both extremes in the climate science debate). Cannondale did some thoroughly fraudulent testing for the CPSC in the USA, "proving" that there wasn't a problem. Even the CPSC themselves got in on the act by writing a letter to tell me about an open meeting to discuss the issue, but not sending it until 2 days after the meeting had taken place (which entailed them sitting on the letter for several weeks, after emailing me a censored version which omitted the vital information). The cycling press are of course in the pay of the industry via advertising, and tried their best to downplay the issue and encouraged riders to not kick up a fuss.

So the story has been been rumbling along gently for the last few years, with continuing occasional reports of problems (eg here). I understand there are a couple of cases slowly working through the legal system, and there was an out of court settlement earlier this year.

In my in-box today:
Well just go the latest Dirt Rider [cycling magazine], Issue 124. And on page 13, there is an ad for the Fox Talas II with "Changed Dropout Slot Angle for Disk Brakes". And on P 59, under the Marzocchi Marathon Race review, "But kudos to Marazocchi (and other manufacturers) for identifying a potentially lethal disc brake and drop out combination..."
So, having discovered, investigated, and solved this problem essentially single-handedly in the face of widespread opposition, and after dragging the manufacturers kicking and screaming to the point where they actually start to take some action, I've now been airbrushed out of history.

Perhaps in a few years I can look forward to someone breathlessly announcing through the pages of Nature saying that climate sensitivity is very likely close to 2.5C...

Update: it's actually Dirt Rag, and their full review of the Marzocchi forks is available on-line here. This is the same Dirt Rag that wrote the following in 2003 (issue 102):
Once we heard about the problem, we contacted a few reputable sources who seem to agree that while Annan's concerns might be valid, those concerns are only in the rarest of cases-such as severe neglect or care for parts, poor or improper assembly or just plain stupid combination of parts (such as a light-weight cross country suspension fork on an off road tandem).

Friday, September 29, 2006

Ask me if I'm bovvered.

I see in the news that a new pressure group has sprung up with the goal of reversing the decline in students studying science at school and beyond. Follow the links at the bottom of that page for many more stories in a similar vein. According to the Great and Good, there aren't enough science teachers or enough scientists. As a result the economy will collapse, and we're all doomed.

Am I bovvered, though? Does my face look bovvered?

Science, by and large, is an increasingly poor career choice (by a number of measures), and I am really rather more surprised that so many people still do it at all, than that the number is declining.

I was lucky enough to squeak through the system at the end of the "golden age" of grants (not loans), when the cost of two degrees was measured merely by the relative intangibles of a subsistence lifestyle and the opportunity cost of not getting on the salary ladder. On top of that, the current cohort face a hefty red figure in their bank balances in the form of accumulated tuition fees and loans. It's hardly a situation that would encourage a rational person to choose a low-paying job with poor prospects.

Ignoring the title, this page paints a pretty realistic picture, I think. I don't for a minute believe his thesis that women don't do science because they alone realise it's a crap choice, whereas us dumb men are stupid enough to fall for the fantasy of fame and fortune (for an alternative explanation, have a read of this - although IMO and IME it [fortunately] represents an extreme case). But as for its description of science as an underpaid, overworked, insecure choice with a high cost of entry and huge drop-out rate, I think it pretty much hits the spot. I hope all my readers understand that even the run-of-the-mill tenured professor is very much at the lucky and/or talented end of the bell curve of a group of people who were already generally at the top of the class before they even started trying to climb up this particular slippery pole. And even if you get the supposedly cushy tenured post, it's no defence when the Govt decides to downsize or just close your lab. I've blogged about CEH before, but those other two links concern firstly the lab where I had my first job, and then the one where my father worked for many years. Both closed on April 1st this year. I think in my 7 years of work in the UK (at 2 labs) there were about 4 rounds of redundancies in all, which hardly results in a conducive atmosphere for work even for those who were not personally threatened.

Perhaps I'm digressing a little. Of course there is a good side to scientific research, and I don't regret my decision at all. It's a great choice for the eccentrics and independently-monied :-) I'd prefer it if there were fewer rather than more of us, though. It's a question of supply and demand.

Of course the CBI wants more scientists. The fat cats who they represent stand to make fat profits on the slave labour of an army of underpaid post-docs (who they did not pay to train or educate, of course). If the supply dries up to the point at which they have to pay scientists higher salaries, I'm not going to shed any tears on their behalf. I'm sure that many of the worthies have their hearts in the right place, and think we need more scientists because science is generally a source of good, rather than because they like to build their empires up with hordes of post-docs with no job security. Nevertheless, oversupply drives down the price, and undersupply will drive it up again.

So basically, if people stop studying science, I'm not bovvered. In fact, I look forward to it. And if you are looking for career advice, please go and be an estate agent or hairdresser or journalist or...well, anything really. Just not a scientist!

Friday, September 22, 2006

Can we believe in high climate sensitivity?

You've seen the film, now read the book :-)

(Oh, there's a paper size issue which cuts the header line off the pages of the pdf - that would have made it clear that this is submitted to GRL.)

I guess you could see this as a re-writing of the Comment on Frame et al as a stand-alone paper, but it's designed as more of a general comment about the whole field (at least, a large part of it) and is based largely on the presentation I gave earlier ths summer. However, since (as far as I can see) Frame and Allen are the only ones actually specifically advocating uniform priors, it's hard to avoid a direct rebuttal of ther claims. The increased character limit means that as well as explaining the problems with other approaches, we can present some new results.

The paper is basically complementary to our previous multiple constraints paper. That considered the effect of combining different observations, this one looks in more detail at the prior and we hope has put some final nails in the coffin of the uniform prior. We show that this approach doesn't work at all, and even if it did, the results would not actually be of any use. If we'd thought more carefully about it, perhaps we could have rolled both halves of the argument up into one paper at the outset, but it's too late for that now. The new results we show are based on the Forster and Gregory analysis. I know I said I wasn't intending to publish a paper based on this, but their analysis is particularly useful due to its independence from climate models and forcing estimates. It now seems to me that an upper 95% probability limit for climate sensitivity of about 4C is easy to justify.

I hope we manage to get some referees who do not have too much of an axe to grind in this debate. Any meaningful comment from readers here is of course also welcome.

We gave Nature the chance to reject it first, which didn't take long. Of course I knew it would be a waste of time sending it there, but I think it's only fair to give them the chance to make amends if I'm going to criticise them. Also, it's amusing to pick the bones out of their excuses. In this case, it was because there is apparently nothing new in our work - this from the same Nature that puffed up Hegerl et al as "the best guide yet" and refused to consider our comment pointing out some rather obvious limitations (effectively the same points that we discuss in our new manuscript). It seems quite clear to me that their editorial filter acts to obstruct rather than enable scientific progress.

Wednesday, September 20, 2006

Sumo

More "science" coming soon, I promise...but first...

Monday was a holiday - "Respect for the Aged Day", to be precise - and the weather was grotty, so we set off for Tokyo to see the sumo which is currently in town. There are 6 major competitions each year, three of which are in Tokyo, but we've not been to one before. Each meeting runs for 15 days, 9am-6pm each day (and you thought cricket was boring?) with several divisions of fighters starting with the lower grades and culminating with the top dogs. It's neither a knockout nor a round-robin, instead everyone fights 15 times in a division of about 40 fighters (I don't know if they have set methods for each day's draw or whether it is all decided in smoke-filled rooms, it is clearly not a Swiss tournament, as the current joint leaders seem to be being kept apart in anticipation of a last day decider).

We had checked on the web site that tickets seemed to be available on the door, but thought we'd better turn up early to make sure we got in. So by 9:30am we were sitting in a nearly-deserted hall watching what looked for the world like two fat men in nappies slapping each other around the head.

It was surprisingly good fun, actually. We watched some of the lower ranks for a bit to get a feel for how it was arranged, and then went out to do some shopping, returning for a bowl of chanko-nabe (a traditional food for sumo wrestlers, quite a tasty soup with meat and vegetables in but nothing special) and to see the better fighters - the ticket allows a single exit and re-entry.

The top division have a skirt-lifting parade before they start (watch it if you don't believe me!):



and then the current champion does a moonwalk across the dohyou (ditto!):



Before the start of each bout, the caller sings their names in a high-pitched voice (this is from an earlier bout, as later on the crowd noise drowned it out:



Then after some posturing and stamping, the bout begins. Some of them were over in seconds, some took rather longer (can you see Baruto's foot slip out while he's trying to throw Ama?):



After the final bout, one of the wrestlers comes in to show off his bow-wielding skills:



And then we all went home. All in all, an interesting day out, but not one I will be desperate to repeat too often!

Saturday, September 16, 2006

Disaster!

But only a pretend one, today at least.

Japan takes disaster preparation very seriously, with an annual "Disaster Prevention Day" when various drills and events take place. Before you laugh at the concept of preventing disaster such as earthquakes and typhoons, bear in mind that the magnitude of the consequences has as much to do with our preparation and response as with the natural event itself (eg consider Katrina/New Orleans).

Today wasn't actually disaster prevention day, but for some reason there was an event put on in Yokohama, I think deliberately targetting the foreign resident population many of who will struggle to comprehend the official Japanese language information. There were plenty of fluent dual-language speakers around, including the lady explaining this earthquake simulation. It's a reconstruction of the Great Kanto Earthquake of 1923, which killed about 140,000 (mostly in fires in the crowded streets)



It was a lot shakier to experience it than it looks on camera! Jules started off standing up but soon saw the error of her ways...several stories up in a tall building I imagine it could be very scary indeed.

There was also a smoke-filled room to negotiate (no photos from that!) and fire engines and extinguishers to play with. We actually bought a fire extinguisher a couple of weeks ago - yes, I know we should have done it ages ago, but we do have plenty of bottled water and food stocked up. Beyond this sort of obvious preparation, it's pretty much a matter of crossing our fingers.

Tuesday, September 12, 2006

Blogapause

Not much going on here but the rain...at least it seems like summer has ended. I caught the first glimpse of higanbana in the verge near our house - this is the "equinox lily" whose appearance I've been looking forward to for about 3 months :-)

I did another practice JLPT test on Sunday and scraped over the magic 60% "pass" threshold this time - 241 points out of 400 to be precise, up 3% from the previous one a couple of weeks earlier. That was despite the handicap of doing the papers back-to-back in a noisy cafe with a caffeine overdose and plenty of distractions - loud Americans are harder to tune out than polite Japanese! I'm on the home stretch as far as learning the kanji and vocab goes, but still pretty awful at the listening comprehension (in fact my score even went down on this compared to how I'd done in the previous test). People who live in Japan and actually use Japanese to communicate seem to generally find this the easy part of the test, but it's only 25% of the overall score so not too crucial. There are still nearly 3 months left, so assuming I don't get lazy (and even if these tests are a bit on the easy side compared to the real thing) things seem to be on course. Perhaps it would be tempting fate to say that 70% is coming within range...

Wednesday, September 06, 2006

Tuesday, September 05, 2006

A modern Difference Engine?

The Difference Engine was a mechanically-powered device which could perform a limited (but useful) set of calculations. As Wikipedia says: "This machine used the decimal number system and was powered by cranking a handle."

I've blogged briefly about my Zaurus before. It remains very useful, mostly for Japanese learning, email/web access and occasionally writing (typically blogs, more rarely scientific papers). However, it has a limited battery life of at most 8h, and much less when using the web. So I've investigated a number of options for recharging it when I go on longer trips away from civilisation - especially long mountain trips. (Yes, I know what you are thinking - why does he need to take that up the mountains? But the email doesn't work up there, it is only used for Japanese language and other amusements in the long afternoons spent idly at the huts.) Previously, I've often carried 4 AA NiMH cells which are an adequate substitute for the proper charger, together with a solar charger to top these up (the solar power is too much of a low trickle to directly power the unit reliably). But although this works, it's a bit of a hassle to set up and also annoyingly heavy given its relative "luxury" status for a walking holiday.

Recently I found a small hand-cranked dynamo, sold as an emergency mobile phone charger. It was only 5 quid, although the postage to Japan was as much again. And voila - the modern Difference engine - turn the handle to power the machine!




It claims an output of up to (limited to) 6V, and when applied to the load of the Zaurus (which has a 3.7V 1800mAh battery) it drops to around 4V or so. I haven't accurately checked the current, but I'd be surprised if I could get it above 500mA - perhaps 200mA is more realistic. It weighs about 70g, which is less than 3 AA cells, let alone the solar charger. I haven't actually used it in anger on a long trip yet, but it seems to work just fine in short tests. I guess it should also hook up with no problems to all sorts of USB-powered (=5V) gadgets with an appropriately-wired plug, so long as their current draw isn't too high. But follow my example at your own risk...

Monday, September 04, 2006

Super computer!

JAMSTEC's new supercomputer was switched on today, and there was a seminar introducing it. It certainly seems super to me! It's an SGI ALTIX 4700, with a total of 1280 dual-core processors. That linked press release talks about availability from late August, so having it up and running by the 4th September is very impressive indeed! Assuming I've done my sums right (not by any means a sure thing) it would slot in somewhere in the top 30 of the latest top500 list. It'll be further down the pecking order by the time the next list is drawn up, of course.

The only drawback I can think of is that it may prove rather more popular than the Compaq Alpha which it replaces (there must be someone's law: "all computers are the same speed", or perhaps: "jobs expand to fill the CPU available"). The Compaq Alpha was our favourite computer as it was good for highly parallel work (EnKF) but relatively lightly used, being a rather different architecture from the Earth Simulator which most code development here is aimed at. However, as well as the Altix there is also a new 24-processor SX8 which was just switched on today too, which may (I hope) be more popular for those who are more focussed on sharing code with the ES.

These two JAMSTEC computers are separately managed from the Earth Simulator, which I understand is also due for an upgrade in a few years (this is still officially in the planning stage, with funding not yet formally agreed).

I've grumbled a bit in the past about some of the drawbacks of working in Japan, so it's important also to give credit where it's due - the facilities here are in many ways second to none. The entire UK doesn't even have anything to compete with this Altix, let alone the ES. And this is all for climate science - I'm not even counting the 10 petaflop Simulator 2 which is likely to target other basic research, perhaps biological. Such a consistently high level of investment in basic research can't help but enhance Japan's image as a tempting destination for foreign scientists looking to expand their horizons...

Friday, September 01, 2006

What is probability?

I happened to come across a somewhat off-hand question "what exactly does it mean to assign probabilities for a single event?" during some random blog-surfing a few days ago. I thought it was widely accepted that such probabilities are essentially Bayesian, that is, subjective expressions of the degree of belief of a person in the proposition in question (eg as Stefan Rahmstorf writes). There are, to be sure, practical difficulties in accessing this belief in a precise and consistent manner (especially if people are prepared to lie), and personal probabilities may change from minute to minute and day to day, but the basic theory seems clear enough and forms the foundation of a large field of research with many practically useful outputs. One thing that is certainly clear (and I believe undisputed) is that the main competing interpretation (frequentism) cannot apply at all in such situations. So if you want to talk in probabilistic terms at all, you've simply got to go outside that framework, and the standard Bayesian angle seems the obvious one.

Anyway, today I finally got the Reply from Allen and Frame to our attempted Comment. [This had been accidentally omitted from the set of reviews that were sent a couple of weeks ago.] I don't intend to publish and fisk it in detail - that would be tedious, lengthy and no-one would care. However, since it was offered for publication, they can hardly complain about me making a couple of comments on it.

One striking sentence in particular jumped out at me:
"We do not think most scientists interpret probabilistic forecasts purely as expressions of degrees of belief."
(And just to clarify, the context makes it clear that this is not indended as a snide comment about the ignorance of "most scientists", but rather as support for A&F taking this same position.)

While they are being admirably clear and frank in acknowledging that they do not actually believe the estimates that they have published, it does rather raise the issue of what they consider the status of their probabilistic estimates to be.

Although I do favour what I understand to be the standard subjective Bayesian viewpoint for non-frequentist probability, I'm not dogmatically going to insist that it is the only possible one - philosophers and mathematicians have argued for centuries over probability, and I don't pretend to have all the answers or to have covered all the bases. Note, however, that Wikipedia only mentions 2 broad categories, Bayesian and Frequentist - any others seem to be rather esoteric philosophical finesses of these two, not major revolutions (excluding imprecise probability which is a whole new can of worms wholly irrelevant to this discussion). Salmon (1966) proposes three criteria for a proposed interpretion of probability:
  1. Admissibility or coherence (must satisfy the Kolmogorov axioms).
  2. Ascertainable (there's a method for calculating it)
  3. Applicable (useful in real life applications)
Obviously, whatever A&F's interpretation is, it fails on admissibility - a point which they have also explicitly acknowledged (indeed they claim it as a feature rather than a bug). Failure on point 2 is therefore a gimme - through being multi-valued (see my previous example on P(x>4) ≠ P(x4>34)) their methods also fail ascertainability, since any answer can be generated by reformulating the question in logically equivalent ways (hmm...I can see a semantic dodge here - does the answer "whatever you want it to be" count as a method for calculating their probability? I'll leave them to decide on that). All that needs to be shown is that their results are not useful and they'll have a 0/3 score :-) Of course this just all means that they think Salmon is wrong too, I guess...but more importantly, it leaves unanswered the question of what their version of probability actually is. What axioms does it satisfy (if any)? What does it mean?

According to their Reply (and indeed the referee who supported them), all this is entirely clear to all climate scientists (except us, I guess) and needs no further clarification. I'd be interested to hear from anyone, climate scientist or not, who can make head or tail of it!

There's a further funny point which I can't resist mentioning. Their Reply makes much of the fact that the D&A stuff (of which Myles Allen is a major contributor) routinely commits the Prosecutor's Fallacy in turning the (frequentist) confidence intervals that classical D&A methods produce, into the (Bayesian) probability intervals that people really want to see. But rather than being embarassed by this, they use it to justify their claim that a uniform prior is in fact the appropriate choice! It really is Emperor's New Clothes stuff.