[jules' pics] ume

err.... Ume varieties of deep pink, pinkish and yellowish...

[Engakuji, Kita Kamakura.]

--
Posted By Blogger to jules' pics at 3/26/2012 12:30:00 PM

[jules' pics] paperbush

Some more not-ume flowers. This is mitsumata (paper bush). Not only are the flowers pleasant smelling, they are also furry!
Furry buds at Engakji in January last year,
...and the same bush, last weekend
But Egaratenjin has the bigger blooms. Furry buds formed in January this year,
...and big yellow balls last weekend.

--
Posted By Blogger to jules' pics at 3/25/2012 05:38:00 PM

[jules' pics] Living it up...

Usually when I attend a posh Oxbridgy event I try to admire the view, enjoy the food, be room filler, and actively avoid the important guests as much as possible, simply because I am sure they will be better entertained by others. This explains how I have dined on several occassions with the Crown Prince of Japan, but never even told him how much I like his shiny hair, suit and shoes. However, last night it all went wrong and James and I ended up with the Ambassador to Japan on one side and the head of Exeter College (Oxford) on the other. The main difficulty for me is alcohol tolerance. Being almost half Japanese, my tolerance is quite low, but it takes a couple of glasses of wine before the honoured guests rise to their full intellectual height, by which time I am well on the way down. Anyway, while embarrassing myself, I learned a great deal - that blogging is bad because words are valuable and people should be paid (in cash) for them, that no journalism in the UK happens outside of London, that the cut in the high end income tax rate is a very good thing, that suggesting that people might be considerate towards each other (in the context of drivers, cyclists and pedestrians) is "banal", and that politicians are wonderful people who only "obfuscate" because normal people can't draw the right conclusions when the hard numbers are laid in front of them. I supose that makes it sound terrible, but it really was a lot of fun, and I have taken these messages to heart, particularly the first one, so will make a suitable contribution towards their valuable time and words when I receive my next shiny begging letters from Oxbridge.

On the way home, 2 of the 3 trainlines between Tokyo and Yokohama were down (due to suicide, as usual), so everyone got on one line, and it was an amazing sardine ride, and we got home very late.

[View is from the Ark Hills Club, Roppongi, Tokyo]

--
Posted By Blogger to jules' pics at 3/23/2012 10:40:00 AM

[jules' pics] springing spring

At last the flowers have come out. Usually the first tree blossoms are ume "plum", but with the delay I think we are going to get a lot of different things blooming close together. These are not your normal ume... ...but are oubai "yellow plum". They have a nice jasmine like smell, much stronger than ume. [at Zuisenji, Kamakura]
There is more to come...
Anyone else think the new D800 looks groovy? Pity about the paycut... or perhaps I use it as a sympathy bargaining point?

--
Posted By Blogger to jules' pics at 3/22/2012 01:53:00 PM

HadCRUT4: 1998 and all that

So the long-awaited HadCRUT4 paper is now published, and the UMKO web site has a press release, though the full data set does not seem to be available yet. I'm sure that won't be long.

As was exclusively revealed on this blog some time ago, according to this updated analysis, 1998 is no longer regarded as the warmest year on record, having been overtaken first by 2005 and then again in 2010. Of course, this was not really controversial in scientific circles, as the small cold bias of HadCRUT3 (due to data voids in the most rapidly warming areas) had been well documented. Competing analyses (NCDC and GISTEMP) that smooth over the gaps already showed these results.

Some of the major differences between the data sets are illustrated by this figure from the paper:


I've added two green ovals to each map to highlight some differences - firstly, filling in the big gap in Siberia, where it was clear in HadCRUT3 that there had been lots of warming even though there were gaps in the coverage. The addition of more observations has merely confirmed what everyone knew, and the previous data has not been changed in any significant way. Moreover, the additional data in the relatively cool area of the south Indian ocean shows that they didn't just try to collect data in the hottest regions. So whatever desperate attempts the sceptics make to discredit this update, they simply don't have a leg to stand on.

The problem really is in the presentation of the data average as representing "global mean temperature" in the first place, when it doesn't, as the data are not missing at random. One good way to deal with missing data in this sort of situation is to fill the gaps with some sort of smoothing (there are a wide range of options of varying sophistication) to generate a global field before averaging, as NCDC and GISTEMP have always done. However, this doesn't matter in the context of the sort of detailed model-data comparisons that underpin detection and attribution, since they generally work at the level of the gridded data and voids are ignored. The mid-century changes to ocean data may however have a modest impact here.

David Whitehouse (or someone impersonating him) has been quick off the mark to bluster about how he would have really won the bet anyway. Unfortunately his comments are highly misleading.

Firstly, we never clearly specified HadCRUT3. I did try to get David to confirm exact details via email, but he refused, and I believe the precise phrase used by Tim Harford was "the Hadley Centre analysis". Maybe his behaviour should have been a red flag that he would resort to rewriting history whenever possible, but I thought it was unlikely to be ambiguous. Of course I didn't know the HC were planning to change their analysis, even though it is inevitable that these sort of changes do take place over time (hence the "3" in HadCRUT3). For how a bet of this type is more clearly specified, see here for example: "in the data set HadCRUTv. or successor data set. Successor data set is the data set used by the Hadley Centre to compare hottest years to the media at that time." Incidentally, Gabi has not won that bet yet, as it specifically refers to the setting of a new record, and 2010 was not.

Secondly, he claims that because 2010 is no hotter than 2005 in the new analysis, he wins anyway. This is more than a bit ridiculous, as the entire bet was predicated on the usually long interval after 1998 in which the record had (apparently) not been broken.

(Of course it will only be a matter of weeks before we start seeing sceptical arguments along the lines of "no global warming since 2005").

A very British coup

There's a short news article in Nature this week - but it's free to read, so please reward their generosity by having a look. The author presumably went over to Hawaii to cover the big CMIP5 workshop that we briefly popped in to. However, the first quote goes to Gavin who steers things neatly in the direction of his workshop that was held the previous week. And Masa Kageyama gets her say too. Poor Jerry Meehl and Karl Taylor are relegated to relative afterthoughts at the bottom of the page. Yes, there is a petabyte of data to look at, and several colleagues are tearing their hair out trying to access it. Apparently the system will be all peachy in a few more months, by which time they will have got it already :-)

To be fair, IMO probably the most exciting thing about the CMIP5 experiments has been the official inclusion of paleoclimate simulations - these have been done before of course, but only as a separate and rather lower-key project. Though this opinion is obviously not shared by many of the modelling groups who haven't actually done the paleo simulations in time for the IPCC report deadlines.

No good deed goes unpunished

As a follow-up to the parenthetical comment I made in a previous post, I thought I would consider the question of IPCC citations more thoroughly. I should start off with a disclaimer, that I consider this sort of bibliometric analysis in general to be a rather limited and blunt tool. For example, I've criticised the h-index in the past, since it rewards co-authorship rather than actual contribution, but on the other hand, it cannot be denied that someone with an h-index of 30 (say) has surely contributed to a large volume of useful research, whereas someone with an h-index of 3 almost certainly has not (yet).

So, I counted up the number of co-authored papers that are cited in the current (first-order) IPCC AR5 drafts. I hope no IPCC authors think this is too naughty, given the semi-private status of these documents, but it's not like I am really giving much away. Also, my totals may be a little bit off, due to the difficulty of identifying "et al" for long author lists, but then again, someone who is nth out of many probably didn't make a huge contribution anyway. I didn't attempt to count the actual citations themselves, because it would have been far too much work, but simply counted up the number of papers that were cited. While of course not all citations represent work of equal importance, it would probably be fair to say that papers which are uncited have not had a great deal of impact, and papers that have not even been written, even less so. Of course, contributing to the IPCC is not necessarily top priority for everyone, but it's a fairly obvious benchmark to indicate the science that people actually care about.

Author	# Papers cited / 1st author
Hargreaves	16 / 3
IPCC LAs
Abe-Ouchi	17 / 1
Kitoh	14 / 5
Emori	9 / 1
Kimoto	8 / 0
Nakajima	6 / 1
Takemura	5 / 2
Kondo	4 / 1
Aoki	3 / 2
RIGC TLs
TL1	9 / 0
TL2	5 / 4
TL3	3 / 0
TL4	2 / 2
TL5	2 / 0
TL6	2 / 0
...	0 or 1

According to my counting, jules has co-authored 16 cited papers (I missed one before) and actually wrote 3 of them herself. Below her, I've listed equivalent values for the Japanese IPCC Lead Authors. These include several of the most productive and eminent active climate scientists in the country. (There is a another whole cadre of emeritus and super-senior types above them, but they don't actually write much, and it's mostly this "younger" generation - where "younger" is 40s-50s - who are leading the research.) Anyone working in climate science is likely to know of most of the top few of this group, but the bottom ones...well, even after 10 years here I had to look them up. To be fair, these guys are mostly genuinely young(er than me :-)) and are presumably considered up-and-coming rather than leading figures.

I also checked on most of the Team Leaders in RIGC - these are the ones whose management positions are effectively protected in a way that jules was not, according to the arbitrary decree of JAMSTEC - and have put a bunch of the best scores I obtained in the table. I don't want to embarrass these guys by identifing them here publicly (though they are all available on this page), I think that at least some of them do good and interesting work and have no real complaint with them holding their positions. The top person on this section has a full-time faculty position at Tokyo University as well as a bit of an army at RIGC, but I don't need to bother with too much special pleading to explain away his score, as despite the huge disparity in resources and position he only scraped to just over half as many papers cited as jules does. I suppose I could have potentially included Abe-san and Emori-san in this section of the list too, due to their RIGC positions - as I have mentioned before, JAMSTEC's short-term contract ideology lead to a highly unsatisfactory absentee management culture that still remains here to a significant degree. I didn't bother to list individually the many TLs who score only 1 or less. Some of these may have the excuse that their main focus just isn't that IPCC-relevant, but that certainly won't wash for all of them, or at least it shouldn't - the clue is in the name, "Research Institute for Global Change"! The fact that the TL results are rather lower than those of the IPCC Lead Authors, seems to broadly support my use of the metric as a rough guide to impact.

Some of our colleagues and collaborators in non-management positions are quite productive, and have several cited papers - for those I looked up, their scores are anything from 1-9 in total papers, with up to 5 as first author for the best of the bunch. Interestingly, the highest values here are all for people outside of RIGC. It's not that RIGC people are particularly stupid or lazy, but for the most part there is an absence of any meaningful mentoring or leadership, such that many people just waste their time going down irrelevant rabbit-holes that no-one outside of their cubicle walls actually cares about. FWIW, my values are 13/4, but as a sort of part-manager in collaboration with jules, that comparison may be a little flattering.

As a reality check, I also looked up a bunch of people who I expected to have made a really significant impact - in no particular order, Gavin Schmidt, Mat Collins, Mike Mann, Hugues Goosse. And sure enough, they all have more (or many more) papers cited, as befits the substantial contributions they have made across a wide range of topics. I haven't actually calculated their numbers as I got bored of counting (and checking duplicates) at about 20! If you haven't heard of Hugues, he was the EGU's young scientist of the year back in 2005 and has remained amazingly productive since then. Mat Collins was Head of something-or-other at the Hadley Centre before moving to a professorship at Exeter Uni. The other two certainly need no introduction. So I don't want anyone to think we are claiming to be really important on the cosmic (or even global) scale. But in the context of what more normal people seem to have achieved, certainly within the small fishpond that is Japan, or the tiny puddle that is RIGC, I think it would be a challenge to argue that our contribution has been so piss-poor as to actually merit demotion and a disbanding of the research group. Nevertheless, this is JAMSTEC, and that is what they have done.

The Narrow Road to the Disaster Zone

People may find this BBC Radio program interesting - I did, anyway. One of the interviewees (Yannick) was someone who went up to Ishinomaki with us for our first week of volunteering, and apparently has stayed there thereafter, though with a different organisation. The narrator also went to some of the fishing village areas that we mostly worked at. Good to hear that they are making progress. I'm currently feeling strangely tempted to go and spend a month or 6 there...

Shame about the silly radiation hype that infected the program - the ~10 µSv per hour that the narrator was panicking about is nothing at all to worry about on a short-term basis, and is in fact of a similar magnitude to the continuous radiation limit for nuclear power workers. 10 µSv per day is typical background. 

I think you've got about a week left to listen.

New Jobs

So as you may have gathered, jules and I have got shiny spanking brand new jobs. From April Fool's Day we will be working in climate change research at an institute called RIGC, located in Yokohama. Conveniently, we will not have to move too far away from our current jobs in climate change research at RIGC in Yokohama. It must be emphasised, however, that any similarity with our current positions is entirely coincidental. Furthermore, our work on predicting climate change is merely a sideshow, a bolt-on optional extra, which is of minor significance to the core strategy of RIGC (founding motto: "Towards the prediction of global change"). It's important to be clear about all this, because under no circumstances should anyone (least of all us) suffer under the illusion that we've been employed for more than 10 years on core research that has been central to the mission of the institute right from its inception. Oh no. That might imply some expectation of continuous future employment, too, and that simply would not do. The core funding has to be reserved for the middle managers and an assorted bag of researchers whose work is sufficiently irrelevant and useless that they can't actually raise any focussed external funding for it. It is important to get the priorities right.

As a reward for her recent efforts in managing and organising a small but relatively successful research group(*), jules has been demoted and awarded a substantial pay cut. I have merely had a pay freeze, after a rather productive year that would normally have resulted in a moderate rise. But we can consider ourselves relatively lucky. Others were summarily sacked - this seems to be principally "pour encourager les autres", since there is plenty of money and in fact further recruitment is planned over the coming months. There is no hint that the sacked staff were actually substandard - they were even invited for interviews, only to find out with less than a month's notice that they would not get their contracts renewed. It is horribly reminiscent of a previous experience in the UK, when a newly-arrived lab director decided we all needed shaken up, which meant the contract staff would be rigorously evaluated (according to his newly-invented criteria, not the performance measures they had actually been operating under) and preferably fired. Jules happened to be first in the queue and was initially told she did not meet his new standards, at which point all the senior staff revolted, and in fact a reasonably sane system (3 year contract followed by tenure evaluation) was designed as a result. The director went on to lay waste to the institute in other ways before hopping off to his next rung on the career ladder and is still one of NERC's darlings. But I digress.

The repeated short-term contract system is of course the same idea that was tried, and found to fail, in the UK around 15-20 years ago. The official ideology behind it is that you can motivate young (and not-so-young...) scientists and encourage their independence through perpetual job insecurity and threat of the sack. Of course the mere act of writing this idea down exposes quite how ridiculous it is, but the middle management simply wring their hands and say "it's the JAMSTEC rule" and the bureaucrats who designed it are sufficiently insulated from the results that they honestly don't seem to see any problems. It was criticised by the institute's external reviews as long ago as 2001, but JAMSTEC has simply ignored those. Across the EU, the perennial contract system was basically outlawed some time ago, of course (there are some very limited exceptions).

I don't expect too much sympathy - when all the dust has settled, it appears that we have 5 years of solid funding, on perfectly adequate salaries - a position that many people might be quite envious of. Even better, we have a substantial additional chunk of funding via long-standing collaboration outside of the institute, which will pay for a post-doc and all the travel etc we can handle. If history is a guide, we will be allowed to do pretty much whatever we want, so long as it contributes to the understanding of climate change. That's just as well, with the next project focussing on "tipping points", as I already mentioned. It's amusing to see how everyone overseas who we've mentioned this to recoils in horror at the phrase, whereas it has just arrived in Japan as the big buzzword of the moment. 5 years ago at the start of the previous project they claimed they were going to do 30-year predictions, and we told them straight away that was idiotic, too. I wonder if they will ever learn how to design these things rationally? First, I suppose they would have to care, and they clearly do not.

One unfortunate casualty of all this is our planned trip to the EGU in April, which we decided that we couldn't arrange and commit to in time. It's a shame as Michel Crucifix had been kind enough to invite me to speak at his Climate Sensitivity session. Luckily that Hansen guy stepped in as a replacement :-) We'll be over to the UK shortly afterwards anyway, for the PMIP conference and a bit of a holiday. As for the longer term, we will see how things go. Spring had sprung and Kamakura is very pretty, even if the atmosphere is currently a bit poisonous at work.

[* As part of a larger project which is primarily charged with "contributing to the IPCC", 15 of her recent (co-authored) papers are cited across 7 chapters of the first draft of the AR5. To put that into perspective, our glorious project leader, who is protected as one of the special people with a core position, has the grand total of two citations.]

Bayesian estimation of climate sensitivity based on a simple climate model fitted to observations of hemispheric temperatures and global ocean heat content

Mentioned previously, but I've just got a copy of it from the author (we don't subscribe to this slightly obscure journal):

Bayesian estimation of climate sensitivity based on a simple climate model fitted to observations of hemispheric temperatures and global ocean heat content

The punchline is this plot which seems similar (but not identical) to that in the Isaac Newton seminar:

The indicated ranges denoted by triangles and line are 90% and 95%, ie a 5% probability of S exceeding 3.5 and 2.5% chance of exceeding 4.3. At a quick glance, the main improvements over previous work are a more detailed treatment of aerosol forcing (one the authors is an aerosol specialist) and the statistical treatment is certainly a bit more sophisticated than some have used before. Interestingly, this result is based on a uniform prior for S, but still cuts off the tail of high values rather effectively. The high tail returns if they add another (pseudo-)forcing term relating to the aerosol effect on cloud lifetime, which is ignored in the main analysis.

Of course this is just one study and I wouldn't expect the IPCC to base their whole report on it, but neither would I expect them to ignore it.

OK, since the commenters seem interested, here are two sensitivity tests where an additional forcing effect (effect of aerosols on cloud lifetime) is included:


You can see that the long right tail reappears when a reasonably strong effect is added. Note, however, that this is still contingent on a prior for S that is actually uniform on [0,20] (eg 70% probability that S is greater than 6) and a more reasonable starting point would not generate such a result.