PIKART | userguide

About PIKART User Guide Data Access Examples Publications Terms of Use

supported by

PIKART Workflow

All steps that PIKART follows to identify, track, post-process and classify atmospheric rivers are displayed in the flow chart below. Click on the boxes to learn more about each step.

All steps of the algorithm and all parameter choices are described in detail in

S. M. Vallejo-Bernal et al., "PIKART: A Comprehensive Global Catalog of Atmospheric Rivers", Journal of Geophysical Research: Atmospheres, 2025, pp. . doi:10.1029/2024JD041869

Interested in the code?
Python code for detecting ARs and creating the PIKART catalog at Zenodo.

You have a good understanding of the method and want to use the PIKART catalog?
Access it via THREDDS data server.

Any feedback or suggestions?
We are happy to hear from you! Please raise an issue and we will revisit your suggestions.

Eulerian and Lagrangian catalog

The PIKART catalog provides a Lagrangian and a Eulerian version.

Eulerian PIKART — (Left) Eulerian version (right) Lagrangian version of the PIKART catalog. In the Eulerian version, AR footprints are colored by their rank (0 - white if unranked, 1-green, 2-blue, 3-yellow, 4-orange, 5-red for ranked ARs). In the Lagrangian version, AR tracks are colored by their track ID. The slider allows to compare both versions of the catalog.

Lagrangian PIKART — (Left) Eulerian version (right) Lagrangian version of the PIKART catalog. In the Eulerian version, AR footprints are colored by their rank (0 - white if unranked, 1-green, 2-blue, 3-yellow, 4-orange, 5-red for ranked ARs). In the Lagrangian version, AR tracks are colored by their track ID. The slider allows to compare both versions of the catalog.

In the Lagrangian version of the PIKART catalog, each AR track is regarded as an individual entity and assigned a track ID. We as observers travel along with each AR. We can define AR trajectories as continuous sets of AR shapes that have been assigned the same identity. Consequently, this version takes an “AR track-centric” perspective, listing ARs in a typical catalog-like fashion.

Some AR properties can only be recorded from this perspective: for instance, only if we continuously consider individual AR identities, we can check whether they penetrate inland. The Lagrangian perspective is well-suited to study AR transport properties and lifecycle characteristics. In this version of the catalog, the following AR features are included:

In the Eulerian version of the PIKART catalog, AR identities are not considered. We take a geographically localized perspective: as an observer, we reside at a specific location and just record whether this location is affected by “AR conditions”. As long as these conditions persist continuously, we record this as an “AR event”. This version takes a “location-centric” perspective, listing ARs in a four-dimensional field/matrix.

AR ranks are provided within the Eulerian framework, in agreement with the broadly used AR strength scale. The Eulerian perspective is often more appropriate for the study of local AR impacts and hazards.