An overview of snow photochemistryevidence, mechanisms and impacts

Three to four parts per billion of ozone per day are being manufactured at the South Pole. Not in Los Angeles. Not at a coal plant. The South Pole, one of the cleanest places on Earth. And the culprit is snow. Sunlit snow is making it. That number is what forced atmospheric chemists to tear up a foundational assumption about the cryosphere. For most of modern atmospheric science, snow and ice were treated as a passive backdrop. A reflective white cap that altered how much sunlight bounced back to space, but otherwise, a permanent sink — a place where atmospheric impurities went to be buried. That view began to crack in the late 1990s. Honrath and colleagues measured upward fluxes of nitrogen oxides above snowpacks at Summit, Greenland. Sumner and Shepson found formaldehyde being produced photochemically in snow. Mauldin and colleagues reported that oxidant concentrations over parts of the Antarctic plateau, averaged over a full day, were comparable to values in the tropical troposphere. These weren't marginal signals. Snow was actively making and releasing reactive gases, and those gases were reshaping the chemistry of the air directly above. The conceptual shift, articulated comprehensively by Grannas and more than thirty co-authors in their landmark review, is this: the snowpack is a photochemical reactor. Not a metaphor. An actual reactor. To understand how it works, you have to start very small — at the scale of a single snow grain. Snow looks uniform. It isn't. At the microscale, impurities in snow occupy three distinct physical environments, and which one they're in determines almost everything about how reactive they are. The first is the quasi-liquid layer, or QLL — a thin, structurally disordered film on the outer surface of each grain that behaves like a liquid even well below freezing. Evidence suggests it can persist down to minus seventy degrees Celsius. The second environment is liquid-like brine pockets or veins at grain boundaries, where solutes get excluded from the ice crystal during freezing and concentrated into small pockets of very high ionic strength. The third is incorporation into the ice lattice itself, as a solid solution, essentially locked in place. The difference in mobility between these environments is staggering. Diffusion coefficients in crystalline ice for species like nitric acid or formaldehyde at around minus fifteen degrees Celsius run from ten to the negative eleventh to ten to the negative twelfth square centimeters per second. In a liquid brine or QLL, that number jumps to around ten to the negative fifth. That's six to seven orders of magnitude faster. Takenaka and colleagues showed that freeze-concentration can produce such extreme local concentrations in brine pockets that nitrite oxidation proceeded roughly one hundred thousand times faster than it would at room temperature. The geometry of a snowflake is, in chemical terms, a set of microreactors with wildly different reaction speeds depending on location. Now add light. Ultraviolet radiation, particularly the UV-B and UV-A spectral regions, between 290 and 400 nanometers, doesn't just bounce off snow. It penetrates. Snow is a highly scattering medium, which means incoming photons get redirected in all directions, producing a diffuse light field inside the pack. That scattering actually enhances the probability that trace absorbers in the snow will intercept photons, and it creates near-surface actinic flux — the total omnidirectional photon flux available to drive a reaction — that can exceed what you'd measure above the surface. Light intensity falls off exponentially with depth, and the characteristic depth of illumination, called the e-folding depth, places most photochemical activity within the top centimeters of the snowpack. But those centimeters are chemically active. The most mechanistically understood chemistry in snow involves nitrogen. Nitrate ions in snow absorb ultraviolet light and photolyze. The dominant product is nitrogen dioxide, and the quantum yield — the fraction of absorbed photons that actually produce reactive products — is about one percent. That sounds small, but when you integrate over a polar summer day with high ultraviolet albedo, it’s enough to sustain substantial nitrogen oxides fluxes into the boundary layer. Honrath and colleagues found at Summit that upward fluxes of nitrogen oxides and nitrous acid, known as HONO, exceeded the downward fluxes of nitric acid during summer — the snow was a net exporter of reactive nitrogen. HONO formation from snow is highly pH-sensitive; because the acid dissociation constant for HONO sits at a pH of 2.8, formation requires acidic conditions, and suppression above pH 5 is nearly complete. Isotopic studies complicate the picture. McCabe and colleagues provided isotopic evidence at the South Pole for nitrate with a stratospheric signature in winter aerosol and surface snow, followed by photochemical recycling that produces tropospheric isotopic signatures. Modeling suggests photolysis accounts for up to about forty percent — usually less — of observed nitrate losses in Antarctic snow, with physical processes like wind-driven loss accounting for much of the rest. All that photolyzed nitrate drives the oxidant story. Hydrogen peroxide in snow is a major photochemical source of hydroxyl radicals, OH. Measured midday OH formation on surface snow grains at Summit, Greenland, ran at around three hundred nanomoles per liter per hour, with peak interstitial air OH concentrations around three million molecules per cubic centimeter. At the South Pole, campaign-averaged OH sat at two million molecules per cubic centimeter — and the explanation is the snowpack. Elevated nitrogen oxides from nitrate photolysis combine with HOx precursors emitted from snow — hydrogen peroxide, formaldehyde, and HONO — to produce an intensely oxidizing boundary layer. Ozone production of three to four parts per billion per day follows from that chemistry: ozone forms when peroxy radicals, both organic and hydroperoxy, react with nitric oxide. High nitrogen oxides, high peroxy radicals — the South Pole boundary layer, under the right stability conditions, runs like a photochemical smog chamber. Halogens add a third dimension. In coastal and sea-ice-influenced environments, bromide and iodide in snow get oxidized to reactive halogen gases. Direct measurements at Alert, Canada, found molecular bromine and bromine chloride at up to twenty-seven and thirty-five parts per trillion, respectively, just above the snow. The CHABLIS campaign at Halley, Antarctica, measured iodine monoxide and bromine oxide peaking near fifteen parts per trillion in spring. These halogens are powerful sinks for hydroperoxy radicals and alter the partitioning of the entire HOx-NOx system. Including halogen chemistry in models improved agreement with measured HO2 at Halley, though OH remained overpredicted — pointing to missing sinks, possibly unmeasured organic compounds. That brings us to organics — and a stark contrast with the nitrogen story. Grannas and colleagues summarize that sunlit snowpacks emit formaldehyde, acetaldehyde, acetone, alkenes, alkyl halides, and organic acids, with proposed sources including direct and indirect photo-oxidation of natural organic matter — humic-like substances, fulvic acids, lignin, and fatty acids. Earlier work from the same group on an ice core found several thousand unique molecular species in the mass range of three hundred to eight hundred Daltons. Robles and Anastasio found that unknown organic species account for roughly half of the sunlight absorbed by filtered snow meltwater, with hydrogen peroxide and nitrate accounting for the rest. So organics in snow are both abundant and photochemically active — and yet the system is poorly constrained. Formaldehyde fluxes from Summit span nearly an order of magnitude across studies. South Pole data suggest photochemical production contributes no more than twenty percent of formaldehyde fluxes, with physical partitioning at ice-air interfaces dominating the rest. Microbes in snow — bacteria and snow algae — can also emit volatile organic compounds, adding another layer of complexity that current models don't capture. The contrast is instructive. For nitrogen, the community has a coherent mechanism: nitrate photolysis, identifiable products, measurable fluxes, isotopic tracers. For organics, there is a rich catalogue of what comes out of snow but a far thinner account of why. Zoom back out to the planetary scale. Carver and colleagues modeled global nitrogen oxides emissions from snow as less than one percent of the total global nitrogen oxides budget — a small number. But in polar regions, where background nitrogen oxides are otherwise near zero, snow-derived nitrogen oxides can be tenfold higher than it would be otherwise, driving ozone production that reaches two to six parts per billion per day near the surface at the South Pole and halogen-driven ozone depletion that extends to one to two kilometers altitude in Arctic spring. These are not trace perturbations to polar boundary layer chemistry. They are the dominant forcing. Ice cores complicate the picture further. Actinic flux penetrates five to twenty-five centimeters into the snowpack, photolyzing deposited nitrate before it can be archived. Up to forty percent of deposited nitrate may be lost this way, with physical losses potentially far larger at central Antarctic sites — factors of one hundred in some cases. That means the atmospheric chemistry record preserved in ice cores has been filtered by snow photochemistry before archiving, and reading that record correctly requires understanding the reactor you're reading it from. As snow cover shrinks globally under a warming climate, the extent and character of this reactor changes. Grannas and colleagues close with a clear assessment: current global models treat snow photochemistry rudimentarily, the full inventory of what snow makes and releases is unresolved, and the feedbacks between snow loss and altered photochemical emissions are unquantified. The field has established that snow is a reactor. The accounting of what it produces and what happens when we take it away is the work still ahead. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.