Tim Collinson wrote:
 > Could you have an exocomet setting off a chain reaction in an
asteroid belt inside the
 > orbit of an inhabited planet such that the dust/fragments dimmed the
light of the star
 > sufficiently to cause cooling?

Yes, but it could take a lot of work; to get the observed dimming (up to
20%) of KIC 8462852 for example you're talking a cloud comparable in
diameter to the star (figure 9, p.10 of the paper).

So this has implications for terraforming or aggressive use.

The angular diameter of the sun from Earth is about 32 arcminutes or
about half a degree.

So at a million kilometres from Earth we need a cloud about 9300km in
diameter to shade the sun (l = r.theta).

For the sake of rough quantification, using Earthly examples:
Last glacial maximum/ice age equals depressing average surface
temperature by 5 Kelvin
Nuclear winter, 5-10 Kelvin
Freezing all life, more than 30 Kelvin

Last glacial maximum = a few percent dimming of surface irradiance
Nuclear winter = almost 50% dimming
Freezing all life 90+% dimming

Bigger hotter stars will be harder to block.
More opaque dust clouds will need more mass, in the absence of 'none
more black' superdark materials.
The closer to the star, the bigger the cloud required.

You may be able to get away with less depending on the inhabited planet
you want to disturb, for smaller values of disturbance.

Some things to think about:
How optically deep/thick is the dust cloud you can generate?
The thicker the cloud, the deeper and faster the cooling effect.

How persistent is the dust cloud?
It will need to have a half life of years, possibly decades or centuries
for sustained cooling of the target (e.g. terraforming Venus).

What is the normal stellar flux at the top of the atmosphere?

How much of the flux makes it to the surface (optical depth of the
atmosphere)?

How much greenhouse effect is present in the planet's atmosphere (heat
trapping)?

How much heat capacity does the planet's oceans have (bigger oceans slow
the cooling rate)?

Rob O'Connor