Ci2026Hackathon: ERA5 and AIMIP Weather Data for Cloud Cover Prediction

ERA5 reanalysis and AIMIP project data form the official dataset for the Climate Informatics 2026 hackathon. The challenge aims to learn the mapping from common weather variables to total cloud cover. Data is provided for two distinct regions and is stored in zarr or netCDF formats.

Use Cases

• Training machine learning models to predict total cloud cover based on commonly predicted weather variables.
• Benchmarking model performance on two distinct geographical regions (region 1 and region 2).
• Analyzing relationships between reanalysis weather fields and cloud formation patterns.

Strengths

• Data originates from authoritative sources: the ERA5 reanalysis product and the first AIMIP project.
• Explicitly licensed under the CC-BY 4.0 license, facilitating open use and sharing.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset scale are unknown, which may limit suitability assessment.

Provenance

Source

ERA5 reanalysis product and the first AIMIP project.

Collection Method

Reanalysis data and model output, likely gathered from climate data archives.

Time Range

null

Freshness

Last updated 2026-04-25 09:50:27; freshness should be verified.

Geography

Covers two distinct regions (region 1 and region 2); specific locations are not detailed.