Three billion years ago was way, way before humans. It was before mammals. It was before dinosaurs and insects and even plants. It was before Earth had any forms of life more complex than microbes. But it still rained back then, and paleoclimatologists have used fossilized raindrops to figure out what kind of atmosphere our planet used to have.
Of course, raindrops themselves can't fossilize, since they're made of water. What can fossilize are the little craters that they leave behind when they hit the ground.
Conditions have to be just exactly right for this to happen: you need a fresh layer of very fine volcanic ash, you need rainfall that's hard enough to cause craters but not hard enough to wash them away, you need another layer of very fine volcanic ash to settle down on top and then you need the whole thing to be left undisturbed long enough to fossilize. Oh, and then someone needs to dig it up and recognize it for what it is. This combination of circumstances doesn't happen very often, but history is sort of a long time, and once in a while, it happens.
This particular little patch of ex-raindrops, discovered in South Africa, are 2.7 billion years old. While the rain itself is long gone, we can still learn plenty from the leftover craters: specifically, we can use them to figure out about how dense the atmosphere was way back then.
The size of the impact crater left by a raindrop (or anything else) impacting a surface is related to several things: the size of the impactor, its composition and density, its speed, and the characteristics of the impacted surface. If you know three of these things, you can solve for the fourth. In this case, we know that the impactor is a raindrop (composition and density), we know the upper limit on raindrop sizes from lab experiments, and we can duplicate the impacted surface with volcanic ash. The last piece is the speed of the raindrop, which varies with atmospheric pressure.
In this month's issue of Nature, researchers at the University of Washington report that their analysis of fossil raindrops shows that the atmospheric pressure 2.7 billion years ago was most likely about the same as it is today. This is big news because there are some important unanswered questions about Earth's climate that long ago: the sun was significantly younger, and between 70% and 80% cooler than it is now, so paleoclimatologists have been trying to figure out just how the heck the whole planet didn't freeze over. The most likely explanation now seems to be that the atmosphere was kept cozy by a hydrocarbon haze that generated a greenhouse effect, the sort of thing that we're artificially creating again by accident.
It's difficult to picture how utterly alien the Earth was 2.7 billion years ago: the atmosphere was inhospitable, since microbes that could create oxygen hadn't even evolved yet. Days were shorter since the entire planet spun faster, and the moon was much closer, generating huge tidal forces. Odds are, if you swung by for a visit, you wouldn't recognize the place. But it's comforting, somehow, to know that sometimes there were also warm and gentle rains, and it blows my mind that we can use those fossilized droplets to tell us more about the history of the planet we call home.