La Palma Lava Mapping

A JupyterLab notebook and associated scripts for the analsyis and semi-automated mapping of the La Palma Cumbre Vieja eruption (September-December 2021), using Sentinel-1 imagery.

The notebook gives an overview of the processing involved, and for further information there is an associated seminar paper available here.

⚠: This is experimental code. Expect it to be slow and fragile (but hopefully not too much).

Main contents

📁 data - Source imagery. See About the data.
📁 data-jpg - JPEG versions of the source imagery, for presentation.
📁 ems-reference - Copernicus Emergency Rapid Mapping reference files.
📄 01_overview.ipynb - The notebook.
📄 .Dockerfile - The Docker image definition.
📄 lplm_*.py - Scripts that contain the bulk of the logic - algorithms, data handling, presentation helpers etc.
📄 readme.md - You're reading it.
📄 start.sh - Script for running locally.

Getting started

The notebook is developed using a Pangeo-based Docker JupyterLab environment.

The dataset is quite small, so is stored alongside the code in this repository.

MyBinder

Now available on MyBinder - this is probably the easiest way to run this.

Running locally

Requires Docker.

Execute the script ./start.sh¹, assuming this is your current working directory. This will build and start a Docker container. It's about 1GB, so will take some time.

The script will mount the current directory as a Docker volume so that you can make changes, and they will show up in the Jupyter environment.

Once ready, look out for the link announcing the URL, which includes the access token. It will look something like:

http://127.0.0.1:8888/lab?token=xxxxxx...

If something doesn't work, it could be due to a dependency update that this code hasn't caught up with - in the Dockerfile, replace pangeo/pangeo-notebook:latest with pangeo/pangeo-notebook:2022.07.13, and please try again.

Running remotely

The Microsoft Planetary Computer is another option for running this notebook... with some tweaks. It is currently free for non-commercial use, and new accounts typically take a couple of hours to approve.

Once you have an account, start a CPU - Python (Pangeo Notebook environment) instance.

Setup Sidecar

The notebook makes use of the Sidecar widget extension, which is not available in Pangeo, out-of-the-box. To enable it:

• Go to the Extensions tab, and enable extensions.
• Open a new terminal, and run pip install sidecar.
• Close any open notebooks, and reload the page

Upload the files from this repository to your environment, and you should now be able to run the notebook.

Example results

Results of running the workflow for 15th November 2021:

About the data

• Copernicus Sentinel-1 GRD:
  • Acquisition mode: Interferometric wide swath (IW)
  • Signal type: Backscatter intensity (linear) - 𝛾0 (radiometrically terrain corrected)
  • Polarisation: VV
  • Orbit: Ascending
  • DEM correction: 30m Mapzen

Credits

• Imagery - Copernicus Sentinel-1, European Space Agency

• Reference maps - Copernicus EMS Rapid Mapping: https://emergency.copernicus.eu/mapping/ems/volcano-eruption-la-palma-spain

• GEOBIA with Python - Open Source Options

Footnotes

1. If on Windows, you can run the script using Git Bash, or similar. ↩