A cheap and cheerful, open source dashboard for the global effort to monitor seismic noise variations through the COVID-19 pandemic

Abstract:

An international team of seismologists ("lockdown seismology science team") published a study of the impact of the global COVID-19 pandemic on human induced seismic noise [1]. The study showed rapid reductions is noise associated with lockdowns that spread across the globe and the amplitude picking up again as lockdowns eased. This was made possible though open-source jupyter notebooks on GitHub so that anyone could run identical processing scripts on their own data.

The Australian Seismometers in Schools network [2] took this one step further. As this network is used for education and outreach as well as research, we wanted a simple dashboard of the seismic noise signal for the public to follow. We made a repository on GitHub with all the necessary scripts bundled to download and process the data and generate the graphs. The scripts are bundled in the form of GitHub actions which run daily to generate updated graphs and also cache the day's processing in the repository itself. We only cache the processed results because the raw data are too bulky for GitHub's storage limits and this does mean we need to discard a window of processed data where there may be reliability issues with uploading. However, after the first run, actions are completed in a few minutes.

This means that a github repository acts as a self-updating store for the processed output of a single site that is reliably available online for everyone to view (see Figure and [3]). We compiled the results for a small number of the most instructive sites into a web page where the embedded images are links to the latest github images. Hey Presto ! A dashboard with explanatory text with no dedicated server required.

We offer this as a way to level the playing field for high schools or remote research groups - to fork the repository, add actions and configure the setup scripts takes a few minutes and can be completed on a smartphone. The concepts work across any small-scale data-processing task that needs to be run regularly and can be expressed in the form of scripts or jupyter notebooks but do rely on data that are updated in real time.

[1] Lecocq, Thomas, et al. Science, July 23, 2020. https://doi.org/10/gg5txj.

[2] Balfour, N. J., M. Salmon, and M. Sambridge. “The Australian Seismometers in Schools Network ... ” https://doi.org/10.1785/0220140025.

[3] Web Page: self-updating-repositories