The Orionids are Halley's leftovers. Can a comet really break apart?

"Did you make up the bit where the comet falls to pieces, or do they really do that?"

That question turned up this week from someone who had been out in the small hours looking for the Orionids, and it deserves a proper answer, because a whole book of mine hangs off it. Yes, they really do that. The rest follows, with the usual warning that I read the manuals; I do not write them.

What you were looking at

The Orionids peaked on the night of 21 to 22 October. Every streak was a speck of dust, most of them smaller than a grain of sand, shed by Halley's Comet on earlier trips round the Sun and scattered along its old path. Earth drives through that trail every October, and the specks meet the upper air at around 66 kilometres a second. That is fast enough to turn something you could not find on your fingertip into a line of light across Orion. They burn out far above anybody's roof.

So the first thing to know about comets is that they lose bits of themselves all the time. A comet is ice, dust and rock packed loosely together, and each time it swings in close to the Sun the ice turns to gas and drags dust out with it. The tail you see in photographs is the comet coming undone, slowly, in public.

Coming apart properly

Sometimes it is more than dust. In July 1992 a comet passed so close to Jupiter that the planet's pull stretched it into a string of at least twenty one pieces. Astronomers found the string the following spring and called it Shoemaker-Levy 9. In July 1994 the pieces struck Jupiter one after another over six days, leaving dark marks in its clouds that telescopes everywhere could see. Others have simply crumbled near the Sun, the way old render gives up on a wet wall: no bang, just a steady loss of interest in holding together.

Why the forecast keeps moving

This was the part I cared about most. An asteroid is a lump of rock, and once you have watched it for long enough its path is a sum you can do. A comet is awkward, because it pushes itself. Gas venting off the sunlit side works like a small, badly aimed thruster, and the push changes as the thing turns and warms. Astronomers have a dry name for this, non-gravitational forces, and it is one of the reasons comet paths carry wider error bars.

Now break the comet up. Every piece needs its own orbit worked out, from fewer sightings, with its own jets. The numbers start wide and tighten with each observation.

Which is why, when the radio in Storm Hull says most of the fragments will fall into open water and then says the specialists are revising, nobody in that room is hiding anything or getting it wrong. That is roughly how an honest forecast of a broken comet would sound.My engineer, Keira, hears it with both arms inside a cooling duct, and all she says is that she heard a prediction changing. She goes on seating the gasket. I trust people like that more than I trust the announcer.

The part I made up

The comet in the book is invented, and so is its year, and so is what it does to the coast. I say so plainly because a week of meteors is a pleasant thing and should stay one. Halley's itself is far out in the cold, not due back our way until 2061. What I borrowed from the real science was the shape of the uncertainty: a forecast that changes for good reasons, so that the people inside the story cannot blame the forecaster and have to get on with deciding things themselves.

If you missed the peak, the shower carries on into the first week of November. Go out in the dark hours before dawn, take a flask and a chair that leans back, and give your eyes twenty minutes before you decide there is nothing up there. There usually is: old dust, arriving very fast, from something that has been quietly falling apart for a very long time.