The threshold weekend

Okey dokey, I've been at a conference in Holland this week held in honour of the retirement of a prominant professor in my field. The talks were quite varied and very interesting - but more importantly the food was fantastic! The conference centre was situated in the middle of a beautiful forest miles from anywhere, which meant that the food and alcohol excesses were easily burnt off with long walks and bike rides (there were bicycles for free use at any time - as is the Dutch way!).

While I was at the conference, I spoke to a number of people about this viva examiner issue I wrote about in my last post. I have a few ideas which I'll put to my supervisor on Monday. I get the impression that asking for an examiner from abroad is not easy (you have to justify it well), but its looking like there isn't much alternative.

Anyway, this weekend marks a threshold in my timetable. I am going to make the transition to devoting a substantial proportion of my time to thesis writing starting Monday. If I want to finish by mid November, that means I have less than six months - scary! I think I'm going to try and plan a writing timetable this weekend (like a revision timetable I guess). I'll post it up when I get it done.

In order to register for thesis submission at the University of London, I have to write a 300 word "description" of my thesis. I was working on it this week, and have just about finished it. You can find it here.

Viva examiners

It's been a few days since my last post, but I haven't really been doing anything related to writing my thesis recently - just some work on my data and trying to work through some analysis problems (and having my birthday). However, I spoke to my supervisor (Linda) today about who we should choose as my viva examiners. She came straight out and said "I can't really think of anyone"! Great. That's part of her job isn't it? Anyway I suggested a few names from some papers I'd read, but I don't think they're that suitable. She's going to give it some more thought...

Here in the UK, you have to have two examiners for your viva (an aural examination on the content of your thesis which comes after you have submitted the final document). One examiner can be internal, i.e. from your institute/university, but one has to be external. Neither of them can have any connection to your work - you cannot have collaborated with them at any point - but they have to be sufficiently knowledgable in your field to examine you properly. I think I have a pretty good idea about the internal one, but I'm going to leave it up to Linda to think of the external one. I hope to get the relevant forms in by the end of the month, so she has until then!

paper progress

Horray, today I finally got to sit down and have a long chat with my supervisor about my current project! I think this makes it the first time she's made it through the paper I've been writing start to end. She had a few major comments, mostly about the structure of the prose, but she thinks most of it is just about there.

However, just as I sat down to start making the changes, admin took over. Even though I've unofficially been given this post-doc job at UCL starting January (see this post), I still have to fill out the application form in full. I considered doing a half-hearted job, but then thought I really ought to do it properly - afterall it's not completely set in concrete yet. So that meant spending the whole afternoon working on the application and not on the paper. Plus we had a seminar this afternoon, which meant more time out. Hopefully will get more time tomorrow to get stuck in.

My first publication

Today marks a fairly momentus event in my research career - my first paper was accepted for publication in the Monthly Notices of the Royal Astronomical Society. I am second author of six. It won't be in actual print for a few weeks, so you can find it here on the astro-ph pre-print archive if you are so interested.

This paper reports on the work that I first started doing in the very first week of my PhD, about two and a half years ago! Briefly, we use spectroscopic observations taken by the Hubble Space Telescope of the central regions in a nearby, well studied starburst galaxy to derive the properties of a number of massive, young star clusters. One of the main results is that we find a compact ionized region around one of the clusters, that at it's predicted age, should not be there. We attribute this to it being in an unusual, high density environment associated with the spectacular starburst activity.

On a different note, I went to a UCL Graduate School lunchtime forum today, based around the title "How to get into academia". Although it was fairly broadly targetted, the four speakers had some interesting things to say. There were quite a few statistics bandied about (like as an undergrad you have a 1 in 10000 chance of becoming a lecturer!) and they mentioned a few tips about the application/interviewing process for academic jobs. One of the things that was said more than once was "publish!" - as much as you can, whenever you can. I thought this was quite fitting considering the subject of the first two paragraphs.

But what about the science?

Ok, I have about four projects that I've been working on over the last two and a half years that will form the bulk of my thesis. For one of these in particular, it has been a long but interesting journey from telescope to reduced data to results. However, it's one thing being able to measure stuff and derive results (which in the end really boils down to following some preset prescription), but another thing entirely to understand what it all means. This is where the real science lies.

I guess every researcher comes across this problem fairly regularly - especially inexperienced PhD students. Maybe making the step from cranking some data reduction, data analysis or coding handle, to understanding what your results mean, marks the transition from student to scientist. Whatever it is, I haven't got there! Just about all four of my main projects are at the stage now where I need to start understanding what their results mean before I can get any further. For this I need my supervisor. Unfortunately she has been quite busy recently, and as a result very slow at responding to my requests for help.

Having said that, there have been some indications that she's been trying to get herself up to speed on my results this last week. She's asked me for a couple of paper references that might indicate that she's doing some relevant background reading. Maybe we'll get around to talking science one of these weeks...

What exactly goes in a thesis?

I've been thinking about what exactly goes in a thesis - I mean more about the style of content than the actual science i.e. in what kind of level of detail do I write? Do I state about everything I ever did, even though it didn't lead anywhere? Do I list every step, or do I summarise like you would for a paper? In order to try and get some kind of feeling about this, I've borrowed the thesis of someone who completed a couple of years ago who did their work on a subject similar to mine. I shall have a peruse over the next few days.

Meanwhile, I've been working on a follow-up paper to some work we submitted for publication last month. I was second author on that paper (what will be my first publication), but my supervisor is happy for me to be first author on this work. The situation is fairly simple - we have a good quality set of observations taken by the Hubble Space Telescope that lent itself quite nicely to two separate (but related) studies. The first paper is concerned with looking at one specific part of the dataset, and this second paper is aimed at looking at the set of observations as a whole. I'm happy writing the paper now, as I know that the words I'm committing to 'paper' will find themselves in my thesis in the end. In effect, I have started writing!

The beginning days

Ok, I've said a few words about what the thesis is going to be about - now onto the more practical side of things. When I say I've just started writing my thesis, I mean of course actually putting words to paper. This is my third year of my PhD studies and I've done quite a bit of research towards my thesis, but no actual words yet.

I'm going to be writing in a program called LaTeX. For those who don't know what this is, it's a document preparation system for high-quality typesetting. Its strengths are its ability to typeset large documents with figures, tables, references, cross-references and complex mathematical formulae in a pre-defined style. The pre-defined style comes from what is known as a style-file, which contains all the 'code' that describes the format of the document (font, layout, title styles, etc.). LaTeX is not a word processor and as such does not show you what the document will look like as you type (sometimes hard to get your head around if you're used to MS Word). To get a typeset document from your LaTeX file, the words and commands have to be compiled, and depending on your system, will usually result in a pdf document being created.

Figures are included in the document using a "include figure" command and can be in eps, pdf or jpeg format. However the exact placement of each figure is left to LaTeX to decide. This can be very frustrating sometimes if you want one particular figure close to the text that describes it, but usually work out ok, and definitely is less hassle than doing it in Word.

Although the text what I really need to get started on, you can't help being a little sidetracked on the style aspect of it all. Infuriatingly, UCL does not have a standardised LaTeX style-file for theses, and as such we have to 'borrow' the style-file of some respected American university. Fortunately, there was someone in the department a few years ago who did this, and created a customised style-file that seems to have formed the basis of many a PhD here in the Astronomy Dept. for the last few years. Over the summer last year, UCL went through a major re-branding excersise, and a new logo was one of the results. They're super-fussy about its usage, so the thesis title page will have to be redesigned.

My work is cut out.

Thesis intro

In this blog I'm going to try and emphasise the actual writing aspect of my thesis, rather than the scientific content, however there isn't any harm in a small introduction. Galaxies in the universe (in relative terms) are not that far apart. Considering this, it's not hard to imagine them passing close by each other, or even colliding, on their path through the cosmos. When this fairly frequent event occurs, the gravitational interaction can cause the gas and dust content of both or either galaxy to become agitated. Individual gas clouds start colliding and the conditions for star formation become very prevalent. The result is a burst of star formation, which is termed a 'starburst'. Starburst galaxies are characterised by intense infrared emission, a population of very young stars, and sometimes by signatures of high-speed outflows.

Similar to the solar wind, very massive stars drive a strong wind, but many orders of magnitude more powerful. With the presence of many massive stars, such as you find in a starburst, the collective power of all these winds can be pretty dramatic. Bubbles and shells form in the ambient gas, which if the conditions are right, can burst and create large-scale galaxy-wide outflows called superwinds.

The subject of my thesis is the study of the interaction between the generation of new stars formed in the burst and the existing gas in the galaxy (called the interstellar medium; ISM). This encompasses the study of the young star clusters that drive the outflows, the dynamics and ionization state of the gas close to the star clusters, and the overall characteristics (morphology, dynamics) of the superwind itself.


I have concentrated my work on two nearby examples of starburst galaxies, M82 and NGC 1569. The links should take you to spectacular Hubble Space Telescope (HST) images of the galaxies, and should help to convince that studying them in detail is a pretty cool business. I am using data from HST, Gemini (on Hawaii) and WIYN (Kitt Peak, Arizona) observatories to further our understanding of galactic winds. Here is an image that I put together combining HST and ground-based images to emphasise the incredibly complex nature of the M82 superwind.