Open-source and open data: combining both worlds for optimised decision making in geological subsurface models

Open data and open-source code are influencing each other: the availability of open data sparks new developments for data analysis and processing. Open-source codes on the other hand have the potential to show the value of open data. This symbiotic effect is well visible in the successful recent developments in the field of machine learning, which was strongly influenced by open data sets and benchmark tests, for example in the famous Kaggle competitions. We outline here the evolving landscape of open-source software developments for (subsurface) geoscience applications. Our overview includes codes and software packages for processing of typical geological and geophysical data sets (borehole data, seismic data, wireline logs, geological maps, outcrop and laboratory data etc.), as well as packages for data processing, up to full 3-D geological modeling and geophysical inversion approaches. The long-term maintenance of these packages is often a challenge, especially when they are developed in research projects. But a combination with open geological data has the potential to lead to transparent and reproducible decision processes, which are relevant in many cases where geological subsurface investigations are used for public decisions such as evaluating possible nuclear waste repository sites or for geothermal energy exploration.