Data sharing – a prerequisite to meet the research challenges of a high-dimensional world
Sharing data via established data repositories is increasingly demanded by research networks, funding agencies, and scientific journals. Funding initiatives like the European Open Science Cloud (EOCS) or the German National Research Data Infrastructure (NFDI) join forces to establish powerful platforms and to provide the necessary services. There are more than enough good arguments how science could benefit from such an open data strategy. However, an open data world is still considered more parasitism rather than symbiosis by many scientists. Is data sharing more than a hype?
Such complaints disregard the urgent need to rethink the way we try to increase our understanding of environmental systems and to develop sustainable management schemes. The last decades have seen a wealth of experimental and monitoring studies. We are now aware of a myriad of single processes. Ironically, though, environmental systems seem to become the more complex, the less understandable and the less manageable the more we know.
Mathematical models are a way to integrate this knowledge and to make it operational. However, empirical studies rarely consider more than one or two single processes. Thus models are poorly informed about interactions in real-world systems. Instead, ad-hoc assumptions have to be made. In addition, model developers struggle with the publication bias of positive findings, leading to a systematic overestimation of effect size in the models.
The next level of environmental science needs to address these issues in a systematic way. The presentation will outline some possible pathways and will provide some examples. Not the least, it aims at illustrating some of the exciting new research options that new data and methodological approaches allow.