“Open Science, Knowledge Sharing and Reproducibility as Drivers for the Adoption of FOSS4G in Environmental Research”
2019-08-29, 16:00–16:20, Coralle Room
EnviDat is the institutional data portal and publication data repository of the Swiss Federal Research Institute WSL. EnviDat actively implements the FAIR (Findability, Accessibility, Interoperability and Reusability) principles and provides a range of services in the area of research data management that were extensively described in Iosifescu et al. (2018a). We continuously improve various aspects of open science support in EnviDat, including implementation of Jupyter Notebooks as documented in Iosifescu et al. (2018b), thus improving the current situation defined by a “replication crisis”.
In Iosifescu et al. (2015) we have presented several reasons for the increasing adoption of FOSS4G (free and open source software for geoinformatics) in academic research: not only the total cost of ownership, but also the growing stability and maturity of the recent open source software packages, the faster bug-fixing turnover, the increasing availability of professional support, and the flexibility to change and repurpose the open software to tackle new research challenges, among others. These reasons are still valid today, and consequently EnviDat trusts PostgreSQL/PostGIS and Apache Solr with the management of its spatial meta(data).
In this contribution, we discuss two novel drivers for the adoption of FOSS(4G) in environmental research: open science and reproducibility. Independent research replication at peer-review is facilitated by the immediate availability of the free and open source software, the absence of software licensing issues and the openness of the code even for older versions of a software. Moreover, researchers producing their own FOSS code can expect a wider distribution of the produced software. In EnviDat, open science support is supported by the combined publication of bundles of datasets and software as for example Fraefel (2018) or purely FOSS4G software as for example Bont (2018). While Fraefel shows how scientists can complement data publication with additional analysis workflows, the work of Bont demonstrates the opening of a methodology for optimizing the geometric layout cable roads and makes it available for everyone as a plugin for QGIS.
In conclusion, supporting reproducibility of research in a portal such as EnviDat is a complex issue that can be simplified by the adoption of FOSS(4G). We would like to stress that reproducibility in science also consists of the transparency of methods and the precise documentation of all steps needed to achieve the published results. In these processes, open source software can play a key role.