This is an independent altitude-resolved companion to the 3D box model, not a flattened copy of it. Instead of one uniform temperature with random per-particle jitter, this model imposes a coherent mountain lee-wave temperature field — the real, well-documented driver of "lenticular" polar stratospheric clouds observed downwind of the Antarctic Peninsula and Scandinavian mountains:
T(x,z,t) = T_mean + A(z)·sin(2πx/λ), λ = 20 km
A(z) = A₀·exp[−((z−19 km)/3 km)²] (wave amplitude peaks mid-layer)
Pressure — and therefore both phase thresholds — now varies with altitude via the barometric law p(z) = 50 hPa·exp[−(z−20 km)/6.5 km], instead of the single fixed pressure the 3D box used:
Ice frost point (WMO Magnus-ice fit):
e_s,ice(T) = 6.1115·exp[22.452·Tc/(Tc+272.55)] [hPa], Tc in °C
T_ice(z) solves e_s,ice(T_ice) = p(z)·(H₂O ppmv)·10⁻⁶
NAT onset: T_NAT(z) = T_ice(z) + ΔT(HNO₃) (illustrative offset, as in the 3D model)
Once a parcel's local temperature (mean + wave) drops through a threshold, its particle radius grows by the classical parabolic diffusional-growth law dr/dt = G·(S−1)/r — solved analytically each frame as r(t)=√(r₀²+2G(S−1)Δt), so the model stays stable at any frame rate. The grown radius then sets a real Stokes–Cunningham terminal fall velocity (sedimentation):
v_sed = (2/9)·(ρ_particle−ρ_air)·g·r² / μ_air · C_c(r,T,p)
μ_air(T): Sutherland's law C_c: Cunningham slip correction (kinetic mean free path)
This is a genuine upgrade over a fixed fall constant: small NAT crystals near 1 µm fall at fractions of a mm/s, while larger ice crystals near 10 µm fall roughly 1 km/day — matching values reported in the PSC sedimentation literature. Particles that reach the bottom of the 14–24 km layer are permanently removed from the HNO₃/H₂O budget (denitrification/dehydration). Heterogeneous chlorine activation is driven by the particle surface area actually present (∝r², not just a particle count), so a few large crystals activate chlorine faster than many small ones at the same number fraction.
- Mean temperature — sets the column-average cooling; combined with the wave, the coldest phase of each cycle can nucleate PSCs well before the mean crosses a threshold.
- Lee-wave amplitude — 0 K recovers a spatially uniform column (like a still atmosphere); realistic mountain waves reach 3–8 K and are the leading real-world cause of the earliest, most localized PSCs each winter.
- HNO₃ / H₂O sliders — set reservoir abundance, shifting both altitude-dependent thresholds.
- Cross-section (top) — one horizontal wavelength × the 14–24 km layer, particles colored/sized by phase and grown radius.
- Time series (bottom) — live traces of remaining HNO₃/H₂O and activated ClOx as the column evolves, impossible to read directly off the 3D scene.