Recollections from our early years in the Met Office

Glenn Shutts

Met Office, UK

I’m sure I’ve never told you this before Tim but the first time I became aware of your presence on the scene was at what I imagine was a Dynamical Specialist Group meeting of the Royal Meteorological Society. You were talking about wave theory and more specifically, about following wave groups, ray tracing and the like. Having spent some time studying Gerald Whitham’s papers on wave theory, I was mightily impressed to see your mathematical presentation include the term C_g • ∇C_g as part of the description of the rate of change of the group velocity following the group velocity! I could see that we were obviously on the same wavelength as it were .... I suspect you had joined Raymond Hide’s group Met O. 21 by this stage.

I had finally run out of NERC research assistant funding and in the summer of 1982, had to apply for a job at the Met Office – something I’d hoped to avoid. I got a rather begrudging acceptance from the Met Office admin department in a letter that began “I’m sorry to inform you... blah blah”. It went on to offer me a job at a rather low grade. Anyway, those early days in the Office were pretty soul-destroying: it felt as if I’d joined the Department for Health and Social Security after being in a Nobel prize winning department at Imperial College. One day, while I was in the foyer staring at ‘the charts’, Raymond Hide came up to me (I already knew him fairly well from Met Office visits) and he told me not to be depressed!! Not long after that, you were posted to Met O 13 and I finally had someone I could talk to about research ideas. In fact, these chats often happened on walks to and from Bracknell railway station, sometimes accompanied by the Director of Research at the time – Andrew Gilchrist.

And then the big problem was presented to us. I don’t know who actually did the forecast model runs on the Cyber 205 demonstrating it (James Murphy, Richard Swinbank, Doug Mansfield?) All I recall was seeing what happened when you ran a weather forecast for 30 days. The surface pressure charts looked as if they would verify well against 200 mb geopotential height charts! Strong wintertime Atlantic westerlies blasting across the frozen Siberian wastes as if the continental land mass had been removed from the model. The zonal-mean of the surface zonal wind was way too strong. Following John Green’s teachings at Imperial College, I was already aware that the mid-latitude surface zonal drag is almost entirely balanced by the convergence of the depth-averaged poleward eddy momentum flux. The surface drag of course has a large parametrized component to it so was that wrong? Or was the model incorrectly representing the poleward eddy flux of momentum? Richard Swinbank’s timely FGGE analysis of the global momentum budget provided a clear answer – the surface drag was about half of what it should be.

I have another related recollection which you may think I’ve imagined. It was JAR time (Job Appraisal Review) at the Met Office and you had one with Andrew Gilchrist. When you emerged from that meeting you said something along the lines of “Andrew thinks it’s gravity wave drag”. Over the years I’ve often wondered whether, in his capacity as director of research, he’d visited the Canadian Met Service and was shown their early work on gravity wave drag? Or maybe he was thinking back even further to John Sawyer’s work in the early 1960s(?) which referred to the need for wave drag. None of this really matters that much and I think we were far better placed to make a case for gravity wave drag than the Canadians. It’s true though that neither of us knew of their work at the time. Your presentational skills and international connections made sure our work would take centre stage too. That’s not to say I don’t have some sympathy for McFarlane and co. Actually, their paper is usually cited alongside ours, as it should be. It does make you think though about how ideas appear to spring up independently – tapping into the zeitgeist. Maybe not so independent as people think!

You eventually escaped from the Met Office (as it was “making you physically ill”). I do remember you being in hospital for a spell and can imagine that this must have been very concerning for you at the time as your career was taking off in a big way. Thankfully you were soon fit again and went on to play a major part in ECMWF’s move towards ensemble prediction methods.

Pardon me for another recollection but before you left the Met Office, you, me and Andrew Gilchrist were walking to Bracknell station talking about ensemble methods and I said rather flippantly, “you could create an ensemble of forecasts by randomly adjusting the values of key parameters in parametrization schemes”. I don’t think Andrew was very impressed and maybe rightly so. Of course, I’m not claiming to have originated the idea – it’s a pretty obvious thing to do. Other times though one can get accused of stealing ideas. Someone once wrote to me implying that I had used their global random spherical harmonic spectrum technique (without crediting them) in the early ECMWF stochastic backscatter scheme. I had to point to an internal working paper that was written in the M.O. during the 1990s where – in response to a request from Paul Mason – I’d devised a simple method for randomly perturbing global wind fields using spherical harmonic-based stream-function forcing patterns. That code was given to Judith Berner years later! Nothing is new under the sun.

There’s a lot of funny moments that loom large in my memory. Probably about 12 years ago or so, you had received a scathing paper review and you showed me some of the reviewers’ comments. One thing stood out though – the reviewer was very fond of ‘dictionary pie’. He/she had used a word that neither of us had ever heard of but sure enough, it was a real word (can’t remember what it was). Anyway, you had the ingenious idea of doing a Google search on that word alongside other words relevant to that scientific discipline e.g., climate simulation. And out popped a name from the search! A timely reminder of how difficult it is to be an anonymous referee.

Another not-so-funny moment from the early days of Met O 13 was when you were talking to me sat at your desk fiddling with a staple gun and accidently stapled your thumb! Ow Ow! I shouldn’t laugh though. While chatting to David Parker seated at his desk once (around the same time), I decided to sit down on the table behind me. Unbeknown to me it was a glass table for drawing with background illumination... Crash! The glass top broke into the wooden desk and a shocked David said “check you are OK in the backside department”, followed by “phone Office Services”. When I explained to the person in charge in Office Services what had happened, she replied angrily “You bloody idiot!”. Welcome to the Civil Service.

Enough of these trivial memories! Just want to say how thankful I am for your support and friendship – both at the Met Office and ECMWF. There were very few scientists I could relate to in those days and I think your easy-going informal attitude was a very helpful antidote to the institutional Met Office environment. I suspect a factor in this was our common interest in 1960s (70s?) music (rock, pop and blues) – even if we followed different branches of the genre. It was interesting to later discover that ECMWF had so many similarly-minded people, one of whom I already knew... Martin Miller.

Enjoy your 70th birthday and thanks for a great symposium!