Practical Reproducibility in Geography and Geosciences

Abstract

Reproducible research is often perceived as a technological challenge, but it is rooted in the challenge to improve scholarly communication in an age of digitisation. When computers become involved and researchers want to allow other scientists to inspect, understand, evaluate, and build upon their work, they need to create a research compendium that includes the code, data, computing environment, and script-based workflows used. Here, we present the state of the art for approaches to reach this degree of computational reproducibility, addressing literate programming and containerisation, while paying attention to working with geospatial data (digital maps, GIS). We argue that all researchers working with computers should understand these technologies to control their computing environment, and we present the benefits of reproducible workflows in practice. Example research compendia illustrate the presented concepts and are the basis for challenges specific to geography and geosciences. Based on existing surveys and best practices from different scientific domains, we conclude that researchers today can overcome many barriers and achieve a very high degree of reproducibility. If the geography and geosciences and communities adopt reproducibility and the underlying technologies in practice and in policies, they can transform the way researchers conduct and communicate their work towards increased transparency, understandability, openness, trust, productivity, and innovation.

Spatial Interaction Modelling: A Manifesto

“…Reproducible work can facilitate the generation of updates as new data become available. Reproducible open data products can also be used as communication and impact strategy expanding the original purpose of research findings (Nüst and Pebesma 2021). Reproducible code and data can be used to provide educational training and enable practitioners to address policy questions which were not outside the scope of the original research project (Rowe et al 2020).…”

Towards Conducting Reproducible Distributed Experiments in the Geosciences

“…Containerization has in recent years gained popularity in research communities as a way to distribute not only isolated programming projects, but the complete configuration of a system. While indeed containerization can aid reproducibility further (Nüst and Pebesma, 2020), distributing the description of the complete computing environment for an experiment may obscure relevant details from less relevant system configuration and "boilerplate" code to set up the system. For mobile and distributed setups the possibility for containerization may currently also be limited.…”

“…More on the user side, there is the idea to make use of the notebook paradigm (Nüst and Pebesma, 2020) for experiment specification. Observable 4 is a platform for JavaScript notebooks, and we plan to evaluate whether such notebooks can be used to specify and simulate experiments in an interactive, visual environment and to supply them to the stimsrv runtime for distributed execution for the actual experiment run.…”

“…Recently, the issues of reproducibility and repeatability have gained some attention in the research community (Nüst et al, 2019;Kosara and Haroz, 2018;Barba, 2018), partly as a consequence of the so-called replication crisis, referring to the fact that in some fields researchers have failed to replicate the results of well-known prior studies. One challenge that researchers who attempt to reproduce a study face is that it may be hard to distil the exact details about an experimentinput parameters, internal calculations, setup of distributed computing environments etc.from the text describing the experiment, even if at first glance the description seems complete.…”

Ordination analysis in sedimentology, geochemistry and palaeoenvironment—Background, current trends and recommendations

“…An additional concern regarding the use of statistical tools has to do with reproducibility (Nüst & Pebesma, 2020). Over the last few years, there has been a growing concern in the scientific community regarding the reproducibility of scientific results (Cooper, 2018;Nissen et al, 2016).…”