FAIR Research Data Management for Complex Land-Atmosphere Observations & Modelling: The LAFI–GLAFO Approach

The FAIR management, documentation, and publication of heterogeneous datasets remain key challenges in land–atmosphere (L–A) interaction research, particularly for data derived from complex three-dimensional observational systems and high-resolution modelling. A new generation of Global Energy and Water Exchanges (GEWEX) Land–Atmosphere Feedback Observatories (GLAFOs) is expected to routinely generate such data. The GLAFO prototype at the University of Hohenheim, Stuttgart, is already operational, producing advanced multi-sensor observations and high-resolution model outputs within the DFG Research Unit 5639, Land–Atmosphere Feedback Initiative (LAFI). Here, we present the research data management approach developed within LAFI to ensure FAIR-compliant handling of these complex datasets, enabling effective collaboration and accelerating scientific discovery.

Addressing LAFI’s scientific aim of closing key knowledge gaps in land–atmosphere (L-A) feedbacks that limit the accuracy of weather and climate simulations is critically dependent on robust research data management. In particular, effective data standardization, interoperability and reliable data access is required to support seamless collaboration across LAFI’s highly interdisciplinary and international community, spanning atmospheric, soil and agricultural sciences, hydrology, bio-geophysics, and neuro-informatics.

To address these requirements, ongoing activities focus on the standardization of diverse datasets to enable straightforward inter-comparison. In line with FAIR principles, LAFI datasets are being converted into Climate and Forecast (CF) convention–compliant NetCDF files and stored on a secure server hosted by the University of Hohenheim. Initially, all standardized data are freely accessible to LAFI researchers, with plans for broader public access via an API service and/or web portal. Associated data conversion and processing workflows, including Python scripts and documentation, are managed through GitLab.

In parallel, the research data management team collaborates with international initiatives such as obs4MIPs to enable the use of LAFI observations for climate model evaluation, including the development of protocols for advanced instrumentation such as Doppler, Raman, and water vapor differential absorption lidars. Beyond documenting best practices, current efforts emphasize the development of training and tutorial materials to support knowledge transfer to the wider community. These activities are aligned with broader initiatives within the German National Research Data Infrastructure (NFDI), including the NFDI4Earth service portfolio, to support FAIR-compliant dissemination across Earth system sciences.

We will present insights from ongoing research data management activities, discuss key challenges encountered, outline potential solutions, and share ideas for leveraging the potential of large-scale AI tools and generative AI. These experiences are intended to contribute constructively to improved Earth system understanding and modelling, broader discussions on research data management and shared challenges, and the development of harmonized guidance for effective scientific data stewardship through the European Open Science Council.