Saturday, 6 June 2009

The Things That Google Does

Now that Bing's the thing, I have stumbled across some Google capabilities that I didn't know existed. I have been using Google as my search engine of choice for many, many years but I had absolutely no idea it could do things other than search. So, without further ado, here's a list of ways to have fun with Google ...

Its a calculator ...
Type (1+1)/0.5 into the search box, or even 36% of 134, or 2^8.
It Tells You the Weather Forecast ...
Type weather:PO29JY, or weather London.
It Converts Units for You ...
Type 100 parsecs in light years, 1 kg in pounds etc.
It's a Capitalist's Tool ...
Type stock MSFT, or even just a stock code e.g. VOD.
Don't wake up Your Relatives Abroad ...
time adelaide
Speak to Your Relatives Abroad ...
Answer the Ultimate Question ...
Type this into the search box: the answer to life the universe and everything

I'll bet there are many other useful features. Let me know if I've missed something!

Wednesday, 13 May 2009


While you are reading this blog, and in fact while you are doing most things on your computer, its Central Processing Unit (CPU) is doing next to nothing. It's probably using only about 1%-5% of its computational ability. Wouldn't it be great if you could put the CPU's 'idle time' to good use? You won't be surprised to learn that you can - and here's how it works.

Academic research groups around the world frequently need access to very powerful supercomputers to perform very long calculations, or to run short calculations millions of times under slightly different conditions. Supercomputers are relatively few and far between - it's difficult to book time on them, and expensive if you can.

This is where BOINC comes in. From your point of view, it's a program that runs on your computer that allows scientific computation programs to run when your CPU would otherwise be idle. That's about the nub of it: You and about a million others donate your idle CPU time to academic research groups.

There's also a social side to it. You can choose which projects you want to use your idle CPU cycles. There are currently about 50 active projects, ranging from advanced cosmology simulations to computer experiments on protein molecule folding dynamics to give insight into disease such as cancer. All of the active projects have flourishing forums where you can chat to like-minded people (you're all there because you have decided that project is worth supporting). There are also "crunching teams" you can join (e.g. BOINC UK), which themselves have active forums and usually have competitive "crunching drives" on the go - often against other teams in a spirit of, mostly, friendly competition.

Nice idea - how do I get started?

Simply download and install the BOINC software from the link below, then choose your projects.

BOINC - Compute for Science

BOINC always needs new contributors. I've been BOINCing for years - I hope you get into it as well, it's very satisfying being able to provide practical support to projects you believe in.

Some links to further reading ...

BOINC homepage at UC Berkeley
BOINC on Wikipedia

Some of my favourite projects ...


Thursday, 30 April 2009

In a Galaxy Far, Far Away ...

... or "GRB 090423 (And Why it Matters)"

This is a detective story concerning the explosion of a star in a galaxy nearly at the edge of the universe, the tale of how it was discovered, and also a few words on why it is an important discovery.

At 07:55:19 UT on 23rd April 2009, the Swift satellite detected a faint Gamma Ray Burst (GRB) using its Burst Alert Telescope (BAT), determining its direction in the sky to within about 3 arcminutes (a rough estimate by modern astronomical standards, but accurate enough for other telescopes and instruments to do follow-up work). The BAT has a large field of view (2 steradians, approximately the same as a single human eye), and is designed to quickly detect gamma ray bursts over a large area. Within seconds, the Swift satellite slewed around to point directly at the burst, and its on-board X-Ray Telescope (XRT) began observing the field at 07:56:31.8 UT, 72.5 seconds after the BAT trigger. The field of view of the XRT is 23.6 arcminutes square. It can record follow-up data as well as localising the source. The XRT found a faint, and fading, source of X-Rays coming from the same direction as the gamma-ray burst. The plot thickens. Swift also carries a 30cm mirror telescope capable of observing in UV and optical wavelengths (UVOT), which points in exactly the same direction as the XRT. It found, on this occasion, absolutely nothing. The plot thickens further - there's a gamma-ray and X-ray source out there but it's not giving off much (if any) ultraviolet or optical light.

The excellent design of Swift gives us a hierarchical way to observe these bursts: Fast detection and rough position calculation (BAT), followed by X-Ray photometry, spectroscopy, and precise localisation of the source (XRT), and UV/visible photometry using UVOT. Meanwhile, through a carefully choreographed and automated communications system, the operators of ground-based telescopes are alerted to the burst. In the case of GRB 090423, a mere 108 seconds after the burst was first detected by the BAT, the 2m Faulkes Telescope South began observing at visible wavelengths in the direction of the burst. Nothing was found. The plot thickens yet again - any optical source would have to be fainter than magnitude 20. About a minute later, the mighty Palomar 60-inch telescope got much the same result - nothing out there unless it is fainter than about magnitude 21.

About 15 minutes later, the 3.8m UK Infrared Telescope (UKIRT), perched atop Mauna Kea on Hawaii, at a lung-busting altitude of over 4000m, was commanded to stare at the source using its huge infrared eye. Bingo! Source found - faint but clearly detectable and not previously seen. The object seen was a point source (so not a big diffuse cloud of glowing gas and dust) shining faintly upon us at a magnitude of about 18 in the mid infrared range (2.2 micron wavelength).

The plot is now getting positively gloopy: Some point source has started spewing out gamma-rays, x-rays and infrared light - but no visible light that we can detect. We now have some very excited astronomers.

The universe is expanding in all directions, which is to say that the distance between galaxies is getting bigger all the time. The further away an object is from us, the faster it is receeding away from us. A side effect of all this extragalactic fleeing is that the light from far away objects is Doppler-shifted towards the red end of the spectrum (the wavelength of the light is stretched out, becoming longer). In fact, if an object is far enough away from us, then it will be receeding away from us very quickly indeed, and any visible light it shines towards us can be red-shifted right out of the optical range and into the infrared range of wavelengths. Spectroscopic measurements taken by ESO's Very Large Telescope (VLT), now give us a redshift for our mysterious point source of 8.2. This is the furthest redshift yet measured, and it tells us that the object is about 13 billion light years away.

Whilst I'm sure you'll agree that this cosmic detective story has so far been gripping enough, in and of itself, we're actually only just getting to the really interesting part. Because light travels at a finite speed, any light we observe from very far away left its source a long time ago - we are looking at old photons that have travelled a long distance. In the case of our mystery object, we are, today, observing light that was produced 13 billion years ago - only 600 million years after the big bang. Whilst 600 million years seems a long time to us humans, it is actually in the infancy of the universe. If the universe were a human, GRB 090423 erupted forth when the universe was a 4 year old toddler. According to Derek Fox at Pennsylvania State University:
"The burst most likely arose from the explosion of a massive star. We're seeing the demise of a star -- and probably the birth of a black hole, in one of the universe's earliest stellar generations."
That is why GRB 090423 matters - it is tantalising evidence that we may be on the right track with our theories of how stars are born, evolve, and sometimes die a violent gamma-ray bursting death. Even in the very early universe.

Some links to further reading ...

The Swift Gamma Ray Burst Explorer Mission
"Gamma-ray burst." Wikipedia, The Free Encyclopedia
The Electromagnetic Spectrum
GCN Circulars
United Kingdom Infra-Red Telescope (UKIRT)
New Gamma-Ray Burst Smashes Cosmic Distance Record (NASA)
The European Southern Observatory's Very Large Telescope array (VLT)
Faulkes Telescope Project

Yogi Stereogram

Here's an image I processed of the famous "Yogi" rock on Mars. The original images were taken by the Mars Pathfinder lander. Go cross-eyed whilst looking at the centre of the image to get the 3d effect.
This is my first ever stereogram and it's made using exactly the same image for both sides of the "stereo input pair", so it isn't the best stereo image I have seen. I think I'll have some more fun with this later and learn how to make good ones.
Thanks to these guys for the free software. And also to NASA et al for spending truly vast quantities of cash in order to take pictures of far-far away rocks (and other science stuff, of course).

And now for something completely different ...

... A test blog. Not all that different, then.