Moon madness

Here in the UK the media seems to be getting slightly obsessed with the Moon this week following the announcement by Sussex police that they are going to deploy more police officers on the streets of Brighton during full moon (via DaveP). Apparently, Inspector Andy Parr compared crime statistics with the Moon's position to discover the alleged link which is attributed to ambient light levels. With the council of the Swedish town of Övertorneå claiming that (Update: 11/6/2007)Robert of Apparent Brightness claiming that - in the case of the Swedish town of Övertorneå - darkness prevents crime (via Apparent Brightness) this may not appear too crazy. However - and this is the important bit - a previous 15 month study in Florida has shown no evidence for a link between crime and the full moon but did show a link with higher ambient temperatures.

So have the Sussex police done a long enough study and tested alternative hypotheses? In fact, do their data even support the conclusion that is being reported? The only data I've seen were on the BBC video clip (see BBC website) so I've tried to reconstruct the plot here. I have kept as true to the original as I can although I had to create a label for the x-axis, complete the text labels for the data series using my own judgement and guess the very last data point as the BBC clip doesn't show those too well.

Dont have nightmares
This is a reconstruction of a plot created by Sussex police as shown on a BBC news video clip on 6th June this year. The year in question may not have been 2006-7 (that is my guess from the context of the clip). Where were you during this year? If you know the whereabouts of the real data, please report it below. And remember, don't have nightmares. CREDIT: Reconstruction by Stuart
What I noticed, while making this plot, is that the "Full Moon" data are quite strange and I'm not really sure I understand why they are not regular around week 13 or week 47. You should also notice that there is a distinct lack of correlation between the full moon and the incidences of crime.

From a statistical point of view the numbers in a particular week are quite low, so the statistical uncertainty in a particular week is quite large (say 25% of the value). If you overlay these uncertainties on the plot you realise that the data are very "noisy"; the only things my eye can pick out are the large increases in weeks 23 and week 31 (did particular events, football matches or summer festivals happen then?) and a possible long term variation. However even this long-term variation - which has a lower reduced chi-squared (6.9) than a 4 week cycle (11.7) - may be due to an increase in or change in policing during the second half of the year. Either way, I don't think a link with the lunar cycle is very valid and certainly not enough to make a change in policy over. I wonder what the folks over at the AstroStat Slog would make of this.

What this report highlights is the popular view that there is a link between the Moon and human behaviour despite there being no evidence for it.

On the other side of the coin, some people aren't too convinced by the real gravitational effects of the Moon; in the BBC clip one of the members of the public said "you hear all sorts of things from everywhere that somehow there's a link - something to do with gravity and magnetic fields - and I don't know if I believe that but I'll keep an open mind". Did he not learn at school about the differential gravitational pull of the Moon shifting large amounts of water every day? The Moon certainly does have an effect on the oceans and people in Brighton should realise that.

This wasn't the only story about lunar influence this week though. Via Cimateer I see that yesterday's Times are reporting that the Moon has a warming effect on the Earth. The article by Paul Simons claims that weather patterns can be linked to the position of the Moon. I've already mentioned that the tides are caused by the Moon, so it is reasonable to assume that tides are generated in our atmosphere too (although solar heating is the dominant driver) and that will affect the weather. However, the article builds up to a study that says "the daily range rises slightly towards full moon, and falls to its lowest at new moon, a difference of up to 0.3C (0.5F)". The article goes on to state that "one explanation is that moonlight can warm the Earth slightly at full moon, although only at night, of course". I can see how reflected sunlight from the Moon could (ever so slightly) warm the Earth but does Mr Simons really think this would only happen at night? The Moon can be visible during the day too!

Anyway, the Moon isn't very reflective (although it is more reflective at infra-red wavelengths) and a full moon is around 465,000 times less bright than a sunny day. So, I'm not convinced that it can do enough to raise/lower the temperatures on Earth by as much as 0.3C. My last complaint about the article is that it is phrased such that people may think that this is relevant to the issue of climate change. The article's title is the main culprit - it could equally have been titled "New moon has a cooling effect on Earth" - and that may be the fault of a copy editor rather than the reporter. So to clarify, this isn't relevant to climate change, because the Moon has a monthly cycle so any effects average out to nothing.

Tags: | |
Posted in astro blog by Stuart on Sunday 10th Jun 2007 (14:54 BST) | Permalink
[an error occurred while processing this directive]
[an error occurred while processing this directive]