Between about 15–25 km over the winter polar vortex, the background stratosphere carries a permanent haze of liquid sulfate aerosol. As the parcel cools, two solid/liquid phases can nucleate on top of it — nitric-acid trihydrate (NAT, HNO₃·3H₂O) and, at still lower temperature, water-ice:
Ice frost point (simplified Magnus-type fit):
e_s,ice(T) = 6.1115 · exp[22.452·(T−273.15) / (T−0.6)] [hPa]
T_ice solves e_s,ice(T_ice) = p(H₂O)
NAT onset (illustrative — real onset follows the
Hanson–Mauersberger 1988 equilibrium curve):
T_NAT ≈ T_ice + ΔT(HNO₃), ΔT rising with log[HNO₃]
Once a particle's local temperature drops through T_NAT it grows as a NAT crystal; through T_ice, as ice. Both are far larger than the liquid background aerosol, so gravity pulls them down through the layer — sedimentation. If they fall out of the parcel before the air warms back up, the HNO₃ or H₂O they carried leaves with them: this is denitrification and dehydration, and it is largely irreversible for that air mass that winter.
The particle surfaces also matter chemically: reservoir species are converted to reactive chlorine on PSC (and cold sulfate aerosol) surfaces, e.g.
ClONO₂(g) + HCl(particle) → Cl₂(g) + HNO₃(particle)
Cl₂ + sunlight → 2 Cl· (rapid catalytic O₃ loss follows)
Less denitrification removes HNO₃ that would otherwise deactivate chlorine back into ClONO₂/HNO₃ reservoirs — so denitrified, PSC-processed air stays primed for ozone loss into spring. The model tracks this as an activation fraction that grows while PSC surface area is available and relaxes back down once the parcel warms.
- Temperature — the primary driver; slide it below T_NAT or T_ice to nucleate particles.
- HNO₃ / H₂O sliders — set the reservoir abundance, which shifts both thresholds (more vapor → particles form at a warmer temperature).
- Speed / Play-Pause — runs sedimentation and chlorine-activation kinetics forward in simulated time.