For more than 150 years researchers have tangled with competing explanations for tiny pits preserved in ancient sediments. Some have interpreted those impressions to be so-called fossil raindrops—rainfall literally embossed in the geologic record—but others have argued for air bubbles rising through sedimentary deposits as a more likely mechanism.

A new study may shed some light on the kind of imprint raindrops leave in fine-grained sediments and other granular materials. Hiroaki Katsuragi, an assistant professor in the department of electrical and materials science at Kyushu University in Japan, catalogued a variety of droplet impacts and the craters they leave in a series of experiments described in the May 28 issue of Physical Review Letters. Among the surprising outcomes of his research: low-speed water drops make deeper craters than medium-speed drops do.

Armed with a high-speed camera shooting 210 frames per second, Katsuragi let drops of water fall onto a surface of loose silicon carbide grains. He experimented with a variety of grain sizes, from four to 50 microns, and controlled the final impact speed by varying the free-fall height from one centimeter to 48 centimeters. (A micron is one millionth of a meter.)

What he found was that the resulting craters fell into four broad categories—sink craters, ring craters, flat craters and bump craters. Perhaps the most curious of all is the bump crater, formed only with the largest granules and the fastest drops. After impact, a central peak remains within the crater, sometimes rising above the level of the pre-impact surface. But even though the fastest drops tend to leave a raised central structure, they make the deepest craters, as measured by the point of maximum depression.

Somewhat slower droplets produce ring craters: the drop forms a broad impact crater and then sinks slowly into the granules, creating a smaller, bowl-shaped structure, bounded by a raised ring, inside the crater—a sort of crater within a crater. Flat craters are a special type of ring crater, only observed with small grains, wherein the ring structure encloses not a subcrater but a sort of level plateau. And sink craters result from slow-impacting droplets that compress the solid layer and then sink slowly into the newly formed basin. Katsuragi calls these "silent, deep ones," because although the low-speed drop has little kinetic energy, its impact leaves a deep scar.

Katsuragi says he was surprised to find that the slowest drops were so adept at creating deep craters, even more so than their medium-speed counterparts. "And I have not completely understood the reason for this effect yet," he says.

He adds that the width of a crater, which roughly scales with droplet speed, is a much simpler thing to predict than its depth. "The complex nature of the depth of the drop-granular impact comes from the competition among impact inertia, capillary force, dissipation, and mixing of fluid and grains," he says. "There are so many parameters that affect the crater depths." The diameter of the crater, in contrast, is essentially determined by how much the drop deforms at impact, making it a much simpler length scale to characterize, he says.

Katsuragi, who in the new study characterized his work to date as a "first step," adds that he is carrying out more research with varied experimental inputs, using different-size water drops and different types of grains. The idea is to build a catalogue of impact parameters and resulting crater shapes, which could help resolve the origin of the disputed fossil raindrops. It could even aid planetary scientists in unraveling the hydrologic history of other planets in the solar system, Katsuragi says, by connecting past precipitation to predicted geologic features.