User-friendly data pipeline for FAIRification of environmental sensordata

In recent years, the requirements for data from earth system sciences have increased massively. Data from observation systems needs to be transferred into larger research data infrastructures, evaluated and flagged via well-defined quality checks, enriched with standardized metadata and finally made available to the public via standard interfaces. And in order to fulfill the FAIR principles, we have to ensure transparency and reproducibility of all these steps. Moreover, the rising demand of near-real-time (NRT) data requires the whole data pipeline to run operationally with minimal manual effort.

However, in many cases, there are still heterogenous data landscapes to be found without centralized control of data, data processing, version control and QA/QC. This is often aggravated by to inconvenient, outdated and isolated tools and software solutions.

Therefore, we develop and implement an adaptable automated pipeline, which combines the assurance of data consistency, QA/QC (Quality Assurance / Quality Control), graphically supported validation and unified persistence and publication of data. User friendliness is achieved by making the system configurable and trackable through lightweight user interfaces over the complete data lifecycle. By only using open-source software solutions and applying community standards for data formats and interfaces, a high level of sustainability and independence can be ensured.

In this presentation, we hence want to demonstrate such an end-to-end data pipeline that finally allows for the FAIRification of typical environmental sensor data.