A Modelling Tour of the Climate System
This is the computational half of Principles of Planetary Climate (EC2213). The course notes derive everything analytically — the shell model, the Schwarzschild equation, the weighting function, gray radiative equilibrium — and they carry no figures and no numbers. The tour supplies exactly those. Every page states a profile outright, hands it to a real radiation code, and plots what comes back.
There is nothing to install. Every page runs climt itself, in your browser, via Pyodide — the same library a research group would run on a cluster, compiled to WebAssembly. The first cell on each page downloads the climt wheel and takes a few seconds; after that, edit any cell and re-run it.
Two threads run through every page, and they carry equal weight. The first is the science: what the atmosphere actually does, as opposed to what the shell model says it does. The second is using climt — building a state, reading a component’s input and diagnostic contracts, getting units and dimensions right, telling mid_levels from interface_levels, sweeping a parameter. That craft thread is cumulative: page N assumes you have read page N−1 and does not re-explain what it introduced. The science on each page stands alone; the code does not.
The chapter map
| Page | Course chapter | What you compute |
|---|---|---|
| ε is not a number | 5 — The Shell Model | the OLR spectrum your single ε averages over |
| The window, measured | 6, 7 | per-band optical depth; the window’s real width |
| Where photons come from | 7 — Basics of Radiative Transfer | the weighting function, and a radiating level per band |
| Gray equilibrium, tested | 8 — Gray Gas radiative equilibrium | your analytic profile, checked by a radiation code |
| The CO₂ knob | 4 — A deeper look at the zero-dimensional model | the forcing per doubling, and which bands supply it |
| Water vapour, and the limit | 4 and beyond | the feedback, drawn; and the runaway limit |
What this tranche does not cover
Nothing here integrates in time. There are no time-stepping loops, no convection, and no surface fluxes: each page makes a single radiation call, or a sweep of them, on a profile you prescribe. That is deliberate — it keeps the causal chain short enough to see, so that when a number changes you know precisely which line changed it.
Radiative-convective equilibrium, moist convection and surface energy balance build on chapters 10–12 of the notes, and belong to a later tranche.
Going deeper
The Radiative Transfer walkthrough treats the same physics at research grade — line-by-line absorption, the k-distribution, correlated-k, gas overlap and the two-stream solver — and shows how the code you are calling here is built.