The Rain Is Not the Perfumer

Imagine the first drops darkening a dry patch of earth. Before the ground is properly wet, the scene seems to have acquired another dimension: an earthy smell, almost as though the rain has made something that was not there a moment ago.

The useful question is not just what that smell is called. It is how something in the ground gets into the air.

Petrichor names the earthy rain-associated smell. Geosmin names a particular soil-smelling compound. Those are not interchangeable answers, and neither word explains the journey. Naming geosmin is a little like identifying a song without explaining how it reached the radio.

I went into this expecting a chemistry story. The more surprising part turned out to be a tiny piece of plumbing inside a falling drop.

The earth has already been busy

Geosmin is one of the volatile compounds made by soil bacteria called Streptomyces. Another is 2-methylisoborneol, usually shortened to 2-MIB. These compounds have an earthy odor; they are not an exhaustive recipe for every smell associated with rain. The biology also gives a reason to resist the idea that rain must manufacture the smell on contact.

In a 2020 study, Paul Becher and colleagues linked production of these odorants to the bacteria's reproductive cycle. Genes for the compounds' synthesis were controlled by sporulation-specific regulators. In the system they studied, emitting the smell was tied to making spores. [1]

The creatures paying attention were springtails: small soil arthropods, not human admirers of an approaching shower. The researchers found springtails attracted to odors from Streptomyces colonies. The laboratory springtail Folsomia candida responded to geosmin and 2-MIB; the study reports both attraction and electrical responses from its antennae. Spores could travel on the animals' bodies and through their guts, emerging in fecal pellets. [1, 2]

That is a rather good reversal of perspective. What humans encounter as earthy atmosphere can, in this bacterial–arthropod relationship, help bring a spore carrier to dinner.

It does not explain every source of geosmin, every soil community or our own response to rain. I used the paper's accessible abstract and the researchers' full explanatory account for this biological example, not its subscription-only methods. But the important distinction survives that limit: making an odorant and making it available in the air are different events.

A drop is not just a hammer

In 2015, Young Soo Joung and Cullen Buie published high-speed-imaging experiments on water drops striking soils and other porous materials. Their result was more particular than a miniature splash. Their stated soil conditions were dry, wettable and firm enough to keep a porous structure during impact. On a suitable surface, the drop became a temporary bubble machine. [3]

First, impact flattened the drop across the surface. Its spreading could outrun the water's absorption into the pores, trapping small pockets of air underneath it. As water entered the porous material, displaced air could feed those bubbles. Meanwhile, the layer of water above them became thinner.

When a bubble met the top of the drop, it burst. A small water jet rose and broke into still smaller droplets, some becoming suspended in the air. The paper's Figure 1 follows trapped bubbles and jets after a drop hit clay loam at two metres per second; Figure 2 lays out the sequence of bubble formation, growth and bursting. [3]

The soil is not simply being struck harder until something flies off. The mechanism depends on timing: spreading water, absorbing pores, trapped air and a thinning liquid roof.

The researchers also asked a crucial second question. Were those airborne droplets merely water, or could they carry material from the surface?

They loaded porous chromatography plates with a fluorescent dye, Rhodamine B, let them dry, and struck them with clean water drops. Glass collectors above the plates acquired fluorescent spots. Air flowing through a duct moved the deposition pattern downstream. Material from the porous surface had traveled with the ejected droplets. [3, material-delivery methods]

Here the lovely explanation needs its boundary kept intact: the tracer was dye, not geosmin. This was evidence for a delivery route, not a chemical measurement of petrichor and not a test of what a human nose detected. The paper offers a mechanism relevant to rain smell; it does not measure the entire journey from a bacterium to someone's perception.

Why harder is not automatically better

If rain were only a hammer, heavier impact ought to make the explanation more convincing. But the experiments found an intermediate region of impact and wetting conditions in which this aerosol mechanism occurred. Too little or too much could miss that region. The authors observed it under tested conditions consistent with light and moderate rainfall, but not their heavy-rain conditions or on sand. [3, Figure 4 and Discussion]

That is not a license to declare that drizzle always smells stronger than a downpour. A real storm changes many things together, including the wetness of the ground and what remains airborne after subsequent drops have passed through it.

A repeated-impact test makes the pore-air requirement unusually concrete. With drops spaced ten seconds apart, aerosol generation stopped after the second impact. With five minutes between drops, it occurred each time. The authors interpreted the contrast in terms of pores filling with water and later drying. Those intervals describe their experiment, not a universal recharge schedule for the smell of a trail. [3, “Effect of porous structures on aerosol generation”]

The laboratory ground matters too. The soil samples had been prepared, wetted and dried for three days at 50°C before use. Controlled porous plates made the physics easier to examine. A demonstration on those surfaces is powerful precisely because it separates variables; it is not a guarantee that every loose, muddy or leaf-covered patch behaves identically.

The distance still to travel

There are at least four questions inside the apparently simple question “Why does rain smell?”

What makes the odorant? What releases it? What lets it persist and travel? What finally becomes available to a nose?

The bacterial study illuminates one biological answer to the first. The raindrop experiments illuminate a route through the second. Joung and Buie explicitly leave environmental fate open: later rain can remove aerosols, and droplets can settle. Their observations do not tell us how much airborne geosmin reaches a person, or how much of a particular rain smell this route explains. An aerosol droplet and an odor molecule are not the same thing: the tracer experiment did not measure gas-phase geosmin or its transfer from droplets into vapor available to a nose. [3]

The next question I would want to test is where those stories meet: under matched soil-moisture and impact conditions, how do measured airborne odorants and human odor judgments change together? A visible jet is not yet that answer.

Still, the changed picture is worth keeping. The earth may have made the relevant compounds before the first drop arrived. Rain can alter their route out. Inside an ordinary impact, bubbles rise through a roof that is already coming down.

The rain need not be the perfumer to change the air.

Sources

[1] Becher, P. G., et al. (2020). “Developmentally regulated volatiles geosmin and 2-methylisoborneol attract a soil arthropod to Streptomyces bacteria promoting spore dispersal.” Nature Microbiology 5, 821–829. Accessible abstract. Biological claims above are limited to the abstract-level findings; full experimental methods were not inspected.

[2] Becher, P. G., and Flärdh, K. (2021). “The scent of earth – a connection between sporulating soil bacteria and springtails.” Global Soil Biodiversity Initiative. The researchers' account of the same study, not independent corroboration.

[3] Joung, Y. S., and Buie, C. R. (2015). “Aerosol generation by raindrop impact on soil.” Nature Communications 6, 6083. Main text, methods and figure captions consulted, especially Figures 1–2, Figure 4, the repeated-impact experiment, soil preparation and Rhodamine B delivery test. No new fieldwork, sniff test or experimental replication is claimed here.