The Numerical Modelling of the Sedimentation of Polar Stratospheric Cloud Particles
Abstract The denitrification of the polar night stratosphere is crucially controlling the amount of stratospheric ozone depletion over both the Arctic and Antarctic. It is accomplished through the formation and sedimentation of polar stratospheric cloud (PSC) particles. The latter process, though co...
| Published in: | Berichte der Bunsengesellschaft für physikalische Chemie |
|---|---|
| Main Authors: | , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
Wiley
1992
|
| Subjects: | |
| Online Access: | http://dx.doi.org/10.1002/bbpc.19920960323 https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1002%2Fbbpc.19920960323 https://onlinelibrary.wiley.com/doi/pdf/10.1002/bbpc.19920960323 |
| Summary: | Abstract The denitrification of the polar night stratosphere is crucially controlling the amount of stratospheric ozone depletion over both the Arctic and Antarctic. It is accomplished through the formation and sedimentation of polar stratospheric cloud (PSC) particles. The latter process, though conceptionally simple, is not easily incorporated into numerical PSC models. We present results of computational experiments demonstrating advantages and pitfalls of various numerical methods. The velocity profiles which are assumed are similar to real particle velocity profiles, but simple enough to allow an analytic solution of the advection equation. The proposed schemes are upwind biased finite volume schemes, possessing many desired properties, in particular mass conservation and positive definiteness. Numerical diffusion presents a major problem, particularly since PSCs are observed to exhibit pronounced layering which requires the modelling of steep gradients in the particle number density. On the other hand, artificial structures can be introduced if semi‐Lagrangian advection over several grid boxes is allowed. Procedures obviating these difficulties are described, in particular polynomial fitting, and the computational efficiency of the schemes is investigated. We conclude that polynomial fitting is a suitable method for treating PSC Type‐I particles, whereas we recommend a novel semi‐Lagrangian method, designed so that artificial structures can be avoided, for PSC Type‐II particles. |
|---|