PlateTectonicTools

Python tools for plate tectonic research.

This repository contains the Python package ptt (short for Plate Tectonic Tools) which provides a collection of common plate tectonic functionality that researchers can use in their workflows. It is primarily built on top of the pyGPlates Python library.

There are also some Jupyter notebooks to demonstrate usage of the ptt package:

• Calculating average seafloor spreading rates of mid-ocean ridges
• Calculating subducting rate of paleo rasters along subduction zones
• Predicting CO2 from age and bottom water temperature.

Note: This repository is only in its initial stages. More functionality and examples will be provided in future.

Documentation

There are some demonstrations of ptt that may be installed from the package itself by running:

import ptt
ptt.install_documentation(path="PTT-Notebooks")

Installation

Dependencies

The following Python packages are required:

• numpy
• scipy
• pygplates

Optional dependencies for running the Notebooks:

• matplotlib
• cartopy

Installing using pip

You can install ptt using the pip package manager with either version of Python:

python2 -m pip install PlateTectonicTools
python3 -m pip install PlateTectonicTools

To install the latest version from GitHub, use:

pip3 install --no-cache-dir --upgrade git+https://github.com/EarthByte/PlateTectonicTools
pip install --no-cache-dir --upgrade git+https://github.com/EarthByte/PlateTectonicTools

Installing using Docker

Coming soon...

API Documentation

continent_contours.py

The current contouring algorithm implements a landmass flood-fill to find all contours for a particular continuous landmass followed by the Marching Squares algorithm to create the contours for that landmass. This is done on the 3D globe and solves the following problems:

• A single very large landmass can have more than one contour (polygon).
• In some cases (particularly for 0-130Ma) the polygons are so large that the inside/outside region of each contour polygon becomes swapped.
• Clipping at the dateline (with a 2D approach) causes perimeters to be larger than they should be.

Essentially you can create a ContinentContouring object using rotation files, some continent/craton features (polygons), a contour resolution to determine how finely tessellated the contour outlines should be, a buffer/gap threshold to expand continents outward, and two area thresholds to separately exclude small continental islands and small oceanic islands. It will then reconstruct the polygons to an age and contour them into continents. This will output the continent contours (as polyline continental-oceanic boundaries) and/or continent masks (a 2D NumPy boolean array of continental crust at each age):

from ptt.continent_contours import ContinentContouring

continent_contouring = ContinentContouring(...)
...
contoured_continents = continent_contouring.get_contoured_continents(age)
continent_mask = continent_contouring.get_continent_mask(age)

The returned contoured_continents is a list of ContouredContinent. And continent_mask is a 2D NumPy boolean array of continental crust (that can be converted to floating-point 0.0 and 1.0 values using continent_mask.astype('float')). The contoured continents can be used to query a continent perimeter, a continent area and whether arbitrary points are contained inside a continent. If you want to query distance to the continent-ocean boundaries you can first retrieve the contours (which are polylines) and then query distance to those.

Documentation for other modules is coming soon...