Earth Evolution — Spatio-Temporal Data and Model Analysis

A standalone Honours (4th-year) module. 3 credit points (approximately 25 contact hours), 10 sessions. You build and interrogate real plate-tectonic reconstructions with GPlately and pyGMT, working from a curated set of real, published research notebooks rather than toy examples.

Quick start: conda env create -f environment.yml && conda activate gplately-pygmt && jupyter lab — see Running the notebooks below for details and a mamba alternative.

Every notebook follows the same pattern: a # === USER CONFIGURATION === cell near the top marks the handful of things you're meant to change (a model, a time, a region, a threshold), followed by cells that run correctly without needing to be understood line-by-line on first read. You are never handed a blank cell — every session asks you to run real, working code, then deliberately change something specific in it and see what happens.

Session map

#	Folder	Theme
1	session01_orientation	Orientation — building a paleo-map: the GPlately + pyGMT plumbing, and why "the model" is really "a model"
2	session02_plate_kinematics	Plate kinematics through time — quantifying motion and its consequences at a plate boundary
3	session03_minerals_tectonic_setting	From minerals to tectonic setting — a mineral grain as a tectonic-setting proxy
4	session04_mantle_dynamics	Mantle dynamics and dynamic topography — surface expressions of deep-mantle processes
5	session05_paleogeography_thermochronology	Paleogeography, paleotopography, and thermochronology
6	session06_sedimentary_basins	Sedimentary basins as the subsidence archive
7	session07_paleobiogeography	Paleobiogeography — reconstructed plate positions as the coordinate system for evolutionary biogeography
8	session08_paleoclimate	Paleoclimate and Earth-system evolution
9	session09_resource_exploration	Reconstruction-driven resource exploration
10	session10_capstone	Capstone — independent adaptation of one extension notebook

Each session folder holds that session's notebooks under descriptive filenames (no numbering tied to any external source — see Where these notebooks come from, below). Session 10 has no fixed notebook of its own: it's a independent mini-project adapting one extension notebook from an earlier session (see instructor_notes/).

instructor_notes/ has the full session-by-session timing plan, the exact data each notebook needs and where to get it, and licensing/attribution. data/ holds the real, small (git-trackable) datasets used in-class; two sessions additionally require a companion data archive too large for GitHub — see Data, below, before running those sessions.

Where these notebooks come from

These notebooks are a curated, in-class subset of a much larger tutorial suite, EarthByte/GPlately-pyGMT-tutorials (80 notebooks in total, BSD-3-Clause). This module's ten sessions select 33 of those as core, in-class material; each session's instructor notes point to further "extension" notebooks from the same suite for independent follow-up work, including the Session 10 capstone.

Filenames here are descriptive only, with no number prefix — the original suite numbers its notebooks against a companion paper's figure/table numbering, which is expected to change after that paper's review is complete. Dropping the numbering here means this module's structure doesn't move when that happens. The session map above is the authoritative ordering.

Running the notebooks

Unlike a pure-Python/scikit-learn module, this one depends on GPlates/pyGPlates-based geospatial libraries that are not pip/Colab-installable — they're distributed via conda-forge. There is no "just click Open in Colab" option here.

Recommended: conda/mamba, once, at the start of term:

conda env create -f environment.yml
conda activate gplately-pygmt
jupyter lab

(mamba env create -f environment.yml is a faster drop-in if you have mamba installed.) Every notebook only needs what's in environment.yml — GPlately, pyGMT, JupyterLab, and the standard scientific Python stack. No GPU is required.

Data

Every notebook reads real, published data. Most of it is small enough to git-track directly — it's already in data/ in this repository. Read instructor_notes/ for the exact per-notebook manifest, but in short:

• Most sessions need nothing beyond what's already here. Plate-reconstruction models themselves (rotation files, plate polygons) are fetched and cached automatically on first use via plate_model_manager — expect a one-off download the first time each session runs, not a manual step.
• Sessions 6 and 8 need a companion data archive (rasters and grids too large for GitHub — sediment/crustal-thickness grids, a Zenodo-only rotation file, dynamic-topography model output) from the source suite's Zenodo record. Download and extract this once, before teaching those two sessions — see instructor_notes/ for the exact archive, the expected zenodo_data/ folder layout, and which notebooks in Sessions 4 and 5 also need a subset of it. Every notebook that needs this data checks for it on its first cell and fails immediately, with a clear message, if it isn't there — so a missing download shows up right away, not partway through class.

Attribution

These notebooks and their data are copied and adapted, with permission, from EarthByte/GPlately-pyGMT-tutorials — R.D. Müller and colleagues' own published, BSD-3-Clause-licensed tutorial suite — and remain under that license; the original notice is reproduced in LICENSE-NOTEBOOKS. Newly written material in this repository (this README, the instructor notes, and any session-specific framing text) is released under the MIT License. Full attribution and per-dataset citations are in instructor_notes/.

License

New material in this repository is released under the MIT License. Notebooks and data copied from the source tutorial suite remain under the original BSD-3-Clause license.