FAIR and reproducible moisture tracking with WAM2layers V3

Moisture tracking is a technique to characterize the transport of moisture in gridded atmospheric datasets. It enables users to map source and sink region of (extreme) precipitation or evaporation, respectively. With intensifying wet and dry extremes, the interest in moisture tracking has grown substantially over the past decade.

Various models are used for moisture tracking today. They differ in their approach (Lagrangian versus Eulerian), complexity, input data, assumptions, et cetera. The lack of ground truth data makes it challenging to interpret the results, especially where different approaches lead to different moisture tracking results.

To facilitate comparison between models, it is helpful if both models and their input/output data adhere to best practices for reusable and reproducible data and software. In the development of WAM2layers V3, we have made good progress in incorporating such (FAIR) standards. For example, the WAM2layers software is easy to find and install, well documented, has a DOI, and follows semantic versioning standards. Input data for example cases is available as well, also with a DOI, and a download utility is part of the software so one can get started in minutes. Similarly, the output data adheres (as much as possible) to CF-conventions and adds all relevant metadata, to maximize the reusability and reproducibility of the results.

In May of 2024, users and developers of various moisture tracking models will gather in Leiden to analyse the results of the first community moisture tracking intercomparison study. This is a first step towards better understanding of the effects of certain modelling choices and to address the uncertainty among different models. At EMS, we will showcase our WAM2layers contributions to the Lorentz intercomparison study to demonstrate how the incorporation of data and software best practices in WAM2layers contribute to a FAIR and reproducible moisture tracking workflow.