A collection of publicly-accessible Earth science datasets and how to access them remotely

---

Here, we will collect some information about ocean, ice, and atmosphere datasets and how to access them remotely. The basic idea is to be able to get started with any of these datasets quickly and avoid having to locate them every time.

Each entry should contain a short description of the dataset and a Jupyter notebook with an example of how to access the data remotely using Python.

Where possible, data are accessed using xarray.open_dataset(), which lazily loads the data without acually entering it into memory. In some cases we also use custom APIs such as copernicusmarine / earthaccess / cdsapi which may require a (free) user sign-up.

In addition, I may add some recipies for working with large gridded datasets.

DATASETS

OCEAN MODELS AND REANALYSES

• Copernicus global ocean reanalysis: [ Notebook ]

• TOPAZ 4b Arctic and North Atlantic model [ Notebook ]

• ECCO?

• ORA S5?

• Barents 2.5 ice-ocean model: [ Notebook ]

ATMOSPHERE REANALYSIS

• ERA5 atmospheric reanalysis: [ Notebook ]
• CARRA - Not available remotely.

SEA ICE

SEA ICE CONCENTRATION

• Bremen AMSR2 - Not available remotely
• NSIDC
• Met/Copernicus

SEA ICE DRIFT

• NSIDC - Not available remotely (monthly available here)
• Met/Copernicus

SEA ICE THICKNESS

• SMOS L3 SIT sea ice thickness: [ Notebook ]

SEA ICE AGE

• NSIDC

OCEAN SURFACE FROM REMOTE SENSING

SEA SURFACE TEMPERATURE

• NOAA OI SST V2 sea surface temperature: [ Notebook ]

SEA SURFACE HEIGHT

SEA SURFACE SALINITY

OCEAN COLOUR

OTHER

• Runoff
• Climatologies
• Waves
• Climate model forecasts
• Bathymetry

APIs

RECIPIES FOR WORKING WITH GRIDDED GEOSPATIAL DATA (TBW)

Want to incude some recipies for common methods like:

• Extracting spatial and temporal subsets.
• Regridding from one lat/lon grid to another.
• Interpolating onto a single-depth surface from a model grid with variable z-coordinates.
• Parallelizing tasks with dask.
• Computing gradient and curl.