Caravan: Global community large-sample hydrology dataset (LEAP/Pangeo zarr)
Caravan standardizes and aggregates multiple regional CAMELS-family datasets (US, GB, AUS, BR, CL, HYSETS, LamaH-CE and extensions) into a single standardized format of ERA5-Land meteorological forcing, discharge, and HydroATLAS catchment attributes for thousands of catchments worldwide. Published here as an analysis-ready, cloud-optimized Zarr store via the LEAP-Pangeo catalog.
catalog.leap.columbia.edu
At a glance
- Region
- Global
- Coverage
- 6,830+ catchments globally
- Access
- Free
- Category
- Academic
- Last updated
- 2024
What's inside
- streamflow
- meteorological forcing
- catchment attributes
- time series
Terms explained
What you can do with it
Typical work that data of this type supports.
- 01
Design flows
Estimate 1-in-50 and 1-in-100 year floods for bridges, culverts and spillways.
Because it has: streamflow
- 02
Low flows
Set environmental flows and check what is left in the river in a dry year.
- 03
Model calibration
Calibrate and test catchment models against observed flow.
Where AI adds value
Ideas for machine learning and AI agents built on this kind of data.
- 01
Flow forecasting
Train LSTM-style models on past records to forecast river levels and flow days ahead.
- 02
Ungauged basins
Combine with catchment data to estimate flow where there is no gauge.
- 03
Automatic QC
Flag gauge errors, shifts and gaps before they reach a report.
Related water data
- Caravan dataset collection: 4TU.ResearchDataGlobal | Free
- Global streamflow indices time series dataset for large-sample hydrologyGlobal | Free
- BfG GRDC (Global Runoff Data Centre)Global | Registration Required
- CAMELS (US): Catchment Attributes and Meteorology for Large-sample StudiesNorth America | Free
- CAMELS-BR: 897 catchments in Brazil (Sextant/Ifremer record)South America | Free
- CAMELS-CH: 331 catchments in hydrologic Switzerland (DORA record)Europe | Free
Find more water data
Search and compare 1,900+ water data sources by region, type and access.