By Stuart Grieve, Software Sustainability Institute Fellow

Traditionally, geomorphologists have performed analysis of topographic data in an ad-hoc manner using commercially available GIS packages such as ArcGIS. Although many of the tools within these packages are useful for academic research, they are effectively black boxes, due to the closed source nature of the software and are typically controlled through a GUI. Open source variants such as WhiteboxGAT and QGIS, which provide methodological transparency, exist but do not resolve the challenges in documenting and sharing scientific workflows.

As a result of this, when I received my Fellowship in 2018, my aim was to use the funds to run a workshop at a large geoscience conference on reproducible topographic analysis. Since my PhD, I have been one of the developers of a software package called LSDTopoTools which has been developed from the ground up to facilitate computationally efficient, reproducible topographic analysis with the ethos of sharing workflows at its heart.

LSDTopoTools started life as a project of the Land Surface Dynamics research group in the School of GeoSciences at the University of Edinburgh, but has significantly outgrown the walls of a single institution. The project now encompasses a complete suite of topographic analysis tools, with capabilities to analyse river channels, floodplains, hillslopes and salt marshes at a wide range of scales, from a single field site to global scale data analysis. It has been developed to have minimal dependencies, making it as portable and stable as possible, with the aim that it should be accessible to anyone who wants to use it. This has led to a large number of users, particularly in countries like India, where commercial software licenses are often prohibitively expensive for institutions.

However, despite our best efforts in documentation, many of our potential users do not have the technical expertise to install LSDTopoTools and begin to use it. Indeed there is often a significant 'intimidation factor' for non-specialists in installing and using scientific software. Therefore, the aim of this workshop was to assist new users in installing the software on their machines, and to expose these users to best practice in using the software to perform reproducible research.

We decided that the best venue for a workshop like this was at a domain specific conference; we wanted to ensure that attendees were not limited to people who already knew about software, as may happen at a more technical conference. Consequently, the ideal venue appeared to be the European Geosciences Union (EGU) General Assembly, held each spring in Vienna. Running a workshop at the EGU General Assembly provided many opportunities:

• It is the largest European geoscience conference and the second largest in the world, with over 16,000 attendees and so there is a large potential audience;

• Workshops are advertised alongside oral and poster sessions, so publicity for the workshop is provided;

• Rooms and AV equipment are dealt with centrally, with staff on hand to assist with the logistics of each session.

As the date of the workshop approached, and our planning became more focused, we began to discover a number of challenges and limitations to running a workshop at such a large event:

• The rooms provided were not set up for a workshop, we only had access to fixed rows of chairs facing a screen;

• The internet access was limited to a very slow wifi connection that was not suitable for downloading large datasets or binaries;

• There was no provision for us to register or contact attendees before the event, or to gather formal feedback afterwards;

• The room was too small for the number of people who wished to attend.

We found that we had to design our workshop around these limitations, in a way which would not have been necessary if we were running a standalone workshop or running it as part of a smaller conference. The biggest lesson I have taken from the experience of organising a workshop in this manner is the need for pragmatism and compromise, there are always solutions to logistical challenges, and the effort to overcome them is more than worthwhile.

In a future post, I will talk about the challenges in organising a workshop at a major conference in more detail.

About the European Geosciences Union (EGU)

The European Geosciences Union (EGU) is Europe’s premier geosciences union, dedicated to the pursuit of excellence in the Earth, planetary, and space sciences for the benefit of humanity, worldwide.