Python and Open-Source Software for Developing Countries: A Catalyst for Change

One barrier to conducting science in developing nations is the lack of essential computing resources. Poor network connections and power outages are common culprits that limit progress. Moreover, except in unique circumstances, budgetary constraints make using commercial software almost entirely out of the reach of students, scientists, and faculty, alike. For these reasons, it is difficult to keep up with the pace of developed countries in terms of computing experience and skills. Here, we present an argument for teaching with free, open-source software when mentoring scientists, especially in developing countries. We do so by drawing upon observations from the Coastal Ocean and Environment Summer School in Ghana, Africa (coessing.org). During 2017-2019, the school transitioned from one of using commercial software to completely using a free, open-source language: Python. During the virtual school in 2020, some cloud computing was used via Google Colab to allow participants to complete computational labs and research projects without the need to download or install software on their local computers. Remote processing and remote storage (i.e. cloud computing) have emerged as promising solutions for groups seeking access to powerful computing capabilities without a need for maintaining such services. Where a small business or research group previously relied upon supercomputing resources and dedicated staff to maintain such resources, this same group can now purchase computing time and data storage at low monthly costs on remote servers using cloud computing technology. By combining the use of freely-available data and remote processing, students were able to produce publication-quality plots and execute code from their own laptops--skills they could take with them after the school. The excitement amongst the school attendees was astounding. There are numerous examples of participants going back to their home institutions, teaching others Python, and using Python to significantly further their own research. Though this case study focuses on scientists in developing nations, the arguments apply equally to underprivileged communities in developed nations, as well. Open-source software is an essential step towards broadening participation in the geosciences.