Meteorology¶
working pillars 1, 8
Atmospheric physics, the models that forecast it, and the observations that constrain them.
An active growth area — these pages are marked working rather than solid because
they will keep expanding, not because they are incomplete as recaps.
Pages¶
| Page | Covers |
|---|---|
| Atmospheric thermodynamics | \(\theta\) and \(\theta_e\), moisture variables, Clausius-Clapeyron, CAPE and CIN |
| Atmospheric dynamics | Rossby number, geostrophy, thermal wind, PV, QG omega, Rossby waves |
| Atmospheric boundary layer | Monin-Obukhov similarity, log profile, diurnal cycle, low-level jet |
| Radiation and clouds | Two-stream transfer, emission height, cloud radiative effect, microphysics |
| Numerical weather prediction | Dynamical cores, physics suites, the convective grey zone, ensembles |
| Data assimilation | 3D/4D-Var, EnKF, localization and inflation, observation error |
| Observations and datasets | Radar geometry, satellite radiances, reanalysis caveats, GRIB and NetCDF |
| Fire weather | Fuel moisture, Rothermel, plume-dominated vs wind-driven, coupled modelling |
Reading order¶
Thermodynamics and dynamics are the physics; boundary layer and radiation and clouds are the processes models must parameterize; NWP assembles them; data assimilation and observations supply and constrain the state. Fire weather is the applied end.
A theme worth naming¶
Meteorology is unusual among the sections here in that the observation is often the least certain element. In fluids or structures you compare a model to a measurement and suspect the model. Here:
- Reanalysis is a model, so verifying against it is partly model-to-model comparison.
- Radar and satellite measure proxies — backscattered power, radiance — and everything geophysical is a retrieval.
- Representativeness error usually dominates instrument error, so a point observation and a grid cell are not the same quantity even when both are perfect.
That inverts the usual V&V posture and it shapes how data assimilation is set up: the observation error covariance is doing at least as much work as the model.
Connections¶
- Fluids — the atmosphere is a rotating, stratified fluid.
- Foundations › UQ — ensembles and uncertainty.
- CV › Scientific imagery — satellite observation.
- Toolchains › Weather models — running WRF and MPAS.
Sources¶
Per page. See the course archive — the METEO course folders are indexed there.