ERA5-Land - the land-sea mask is now available/Invariant parameters

The land-sea mask and geopotential used to produce the ERA5-Land dataset is now available from:

ERA5-Land data documentation

Please see the section Parameter listings

This is interpolated to a regular lat/lon grid of 0.1°x0.1°, and is in grib2 format.

To download auxiliary land invariant parameters, please use ERA5 (see table1 of ERA5 documentation).



C3S User Support

Auxiliary land invariant parameters such as soil and vegetation types are not yet available in ERA5-Land catalogue but the information is available in the ERA5 catalogue  ("soil type" ) and the ERA5-Land IFS documentation. Particularly, Figure 11.26 shows the type of soil corresponding to each value of the field.

The two reanalysis use different IFS versions. Therefore, the users have to pay attention to the different horizontal resolution when combine the information coming from the two datasets. Table and figure below show the dataset horizontal resolutions:

Dataset Horizontal resolution
ERA5 0.25°x 0.25°; 31 km grid spacing

0.1°x 0.1°; 9 km grid spacing

Soil temperatures in ERA-Interim (left), ERA5(centre) and ERA5-Land (right). Figure courtesy of ECMWF (Newsletter Number 159 - Spring 2019).

Hi Michela!

How to access these parameters from the CDS API? Trying a similar command as for ERA5 yields a "no data" error. Here is what I tried:

import cdsapi

c = cdsapi.Client()

'format': 'netcdf',
'variable': [
'land_sea_mask', 'orography'
'year': '1981',
'month': '01',
'day': '01',
'time': '01:00',

Thanks for your help!


the land sea mask and orography files are downloadable from the link in the first table of Parameter listings section in the ERA5-Land documentation.

Kind regards


Thanks! But these are in GRIB format. Is it possible to download them as netcdf format from the CDS API for consistency with the other data variables?

EDIT: I know how to convert the files (How to convert GRIB to netCDF), but it would be very nice to have the data from CDS as well to avoid the many steps I had to do to get to the data.

Hello, can you please specify which link, I'm not able to find it

Land sea mask is the third row on the first table. You can download it from the link in the second column.

The land-sea mask provided with ERA5-Land contains fractional values of land coverage

overlapping both seawater and freshwater domains as a single field.

Is there a way (or a specific mask) to extract only the overlap between land and seawater domains?

This would help to identify the coastal runoff gridcells to extract the effective river discharges, 

especially in those deltaic areas (e.g., Amazons) where ocean and freshwater regions are connected

and it is impossible to split these domains using only the provided land-sea mask.



How does the land-sea mask affect the generation of ERA5-Land 2m temperature data?  It seems that some cells along the coastline that are covering land are set to 0 or missing. Is there any detailed documentation on coastline cell related issues?