31573.52
No really. At least according to Duncan Forgan at the Institute for Astronomy at the University of Edinburgh.
The Drake equation famously calculates the number of advanced civilisations that should populate our galaxy right now. The result is hugely sensitive to the assumptions you make about factors such as the number of planets that orbit a host star that are potentially habitable, how many of these actually develop life and what fraction of that goes onto become intelligent etc.
Disagreement (ie general ignorance) over these numbers leads to estimates of the number intelligent civilisations in our galaxy that range from 10^-5 to 10^6. In other words, your best bet is to pick a number, double it….
So Forgan has attempted to inject a little more precision into the calculation. His idea is to actually simulate many times over, the number of civilisations that may have appeared in a galaxy like ours using reasonable, modern estimates for the values in the Drake equation.
With these statistics you can calculate an average value and a standard deviation for the number of advanced civilisations in our galaxy.
Better still, it allows you to compare the results of different models of civilisation creation.
Horgan has clearly had some fun comparing three models:
i. panspermia: if life forms on one planet, it can spread to others in a system
ii. the rare-life hypothesis: Earth-like planets are rare but life progresses pretty well on them when they occur
iii. the tortoise and hare hypothesis: Earth-like plants are common but the steps towards civilisation are hard
And the results are:
i. panspermia predicts 37964.97 advanced civilisations in our galaxy with a standard deviation of 20.
ii. the rare life hypothesis predicts 361.2 advanced civilisations with an SD of 2
iii. the tortoise and hare hypothesis predicts 31573.52 with an SD of 20.
Those are fantastically precise numbers. But before you start broadcasting to your newfound friends with a flashlight, it’s worth considering their accuracy.
The results of simulations like this are no better than than the assumptions you make in developing them. And these, of course, are based on our manifestly imperfect but rapidly improving knowledge of the heavens.
The real question is whether we’ll ever have good enough data to plug in to a model like this to give us a decent answer, without actually discovering another intelligent civilisation. And the answer to that is almost certainly not.
Ref: http://arxiv.org/abs/0810.2222: A Numerical Testbed for Hypotheses of Extraterrestrial Life and Intelligence