Alice: People should do more cloud seeding.

Carol: Cloud seeding doesn't work.

Alice: Governments spent a lot of money doing it, and there are a bunch of studies. And you're saying it just doesn't work?

Carol: Why is a cloud opaque? Because it has water droplets in it. If it already has droplets formed, then nucleating droplets does nothing.

Alice: Ah, a common misconception. Cloud seeding is about nucleating ice crystals, not water droplets. Scientists tested the concept before cloud seeding was deployed.

Carol: That was tested in unrealistic conditions: closed chambers with negligible suspended particles. And it's not that clouds are at altitudes with no particles, either - aerosol number density is similar up to around 9 km.

Alice: The fact is, cloud seeding was done on a large scale, it's still done some places, and there were a bunch of studies on its effectiveness.

Carol: Those studies showed positive effects only because of the same kind of publication bias you see for bad psychology studies. The better meta-analyses showed negligible effect sizes. That's why usage mostly stopped, but it was kind of an embarrassing failure with nobody motivated to publicize it, so people weren't very vocal about that.

Alice: No, if that was the case, there would be more articles and stuff about it, and besides, there are still governments doing cloud seeding today. After so much time to study things, governments wouldn't do cloud seeding at all if it just didn't work, which means it must work, so we should do more of it.

Who is it correct to believe here?

New Comment
2 comments, sorted by Click to highlight new comments since: Today at 2:33 PM

There’s two kinds of “work” here.  One is to induce rain that would not have otherwise occurred, and the other is to induce rain that would have naturally occurred a few minutes or a few kilometers away.  Call it “making” versus “moving” rain.  A cursory examination of Wikipedia suggests the recent claimed successes have been at moving rather than making rain.  And the failures at finding any effects have been at making rather than moving.  And of course it would be far more useful to make rain than to move it.

In Alberta in the 1970s, there was a cloud seeding program whose aim was not to produce rain, but rather to reduce hail (which can be very damaging to crops, as well as property).  It was viewed as experimental, with researchers doing surveys about how much hail fell.  I don't know whether it worked or not, but I don't think one can come to any firm conclusion about it on the basis of simple observations such as "clouds are opaque".

But whether it works or not, it seems it's still actively done in Alberta:

New to LessWrong?