Revision as of 18:00, 24 February 2022 by P.petrelli (talk | contribs)

GLEAM (Global Land Evaporation Amsterdam Model) is a set of algorithms that separately estimate the different components of land evaporation (or 'evapotranspiration’): transpiration, bare-soil evaporation, interception loss, open-water evaporation and sublimation. Additionally, GLEAM provides surface and root-zone soil moisture, potential evaporation and evaporative stress conditions.

The latest available version is v3.5, this includes two datasets v3.5a and v3.5b

There are 10 variables included:

  • Actual Evaporation (E) 
  • Soil Evaporation (Eb)  
  • Interception Loss (Ei)
  • Potential Evaporation (Ep) 
  • Snow Sublimation (Es)
  • Transpiration (Et)
  • Open-water Evaporation (Ew)
  • Evaporative Stress (S)
  • Root-zone Soil Moisture (SMroot)
  • Surface Soil Moisture (SMsurf)

These two datasets differ only in their forcing and temporal coverage:

  • GLEAM v3.5a: a global dataset spanning the 41-year period from 1980 (January 1st) to 2020 (December 31st). The dataset is based on satellite and reanalysis data (ERA5 net radiation and air temperature).
  • GLEAM v3.5b: a global dataset spanning the 18-year period from 2003 (January 1st) to 2020 (July 31st). The dataset is based on satellite data.

The key differences between v3.5 and the previous (v3.3) release are:

  • The v3.5 datasets are produced using a new Python version of GLEAM.
  • Both v3.5a and v3.5b use the latest version of MSWEP precipitation (v2.8), ESA-CCI soil moisture (v5.3), and VODCA VOD.
  • The v3.5b dataset uses the latest version of CERES radiation (v4.1) and AIRS temperature (v7.0).
  • The vegetation fractional cover in v3.5b comes from MOD44B instead of the MEaSUREs dataset (still used in v3.5a).
  • The variable potential evaporation is now released including interception loss.

For more information refer to the GLEAM website.

Initially we replicated GLEAM V3 as part of the CABLE benchmarking collection (project wd9). The v3.5 version is hosted in the ua8 project, not in wd9 as previous versions.

Previous versions

The development of V3 started in 2017, since then 3 subversions were made available:  v3.1, v3.2 and v3.3, each one composed of two datasets, usually referred to as "a" and "b".

The GLEAM v3 includes:

  1. A new data assimilation scheme that has been validated for Australia (Martens et al., 2016) and that has been optimised to work at the global scale.
  2. An updated water balance module that describes the infiltration rates as a function of the vertical gradient in soil moisture.
  3. Updated evaporative stress functions that combine the vegetation optical depth and the root-zone soil moisture estimates.

This version is described in detail by Martens et al. (2017, GMD), listed in the acknowledgements section.

The most recent is v3.3, the key differences between the new datasets and v3.2 are:

  1. The v3.3a dataset is now produced using surface radiation and near-surface air temperature from the latest reanalysis of ECMWF, ERA5 (as opposed to ERA-Interim in v3.2a).
  2. Both GLEAM datasets are now produced using dynamic land cover information based on the MEaSUREs Vegetation Continuous Fields dataset (as opposed to a static map based on MODD44B v52 in the v3.2 datasets).
  3. All forcing datasets have been updated to their last versions and extended until the end of 2018. Due to the latency in CERES radiation data, GLEAM v3.3b only runs until September 2018.
  4. 4. Next to the daily-resolution data, both monthly and yearly datasets are available as well.



The latest v3.5 version is available in ua8


where frequency is daily/monthly/yearly.

The GLEAM_v3-5 folder also contain the original documentation as pdf.

The older versions are part of the CABLE benchmarking collection and are part of the wd9 project, for each version # files are in:


where the specific datasets are called "a" or "b" and in case of v1 also "c" .

To access this data currently you need to be part of the cable group, access for the cable group can be requested as for other group on, you might be asked though also to agreed to the cable group terms and conditions.

Access for the wd9 collection will change on the next few months following NCI re-organization of dataset projects.



Whenever GLEAM datasets are used in a scientific publication, the following references should be cited:

  1. Martens, B., Miralles, D.G., Lievens, H., van der Schalie, R., de Jeu, R.A.M., Fernández-Prieto, D., Beck, H.E., Dorigo, W.A., and Verhoest, N.E.C.: GLEAM v3: satellite-based land evaporation and root-zone soil moistureGeoscientific Model Development10, 1903–1925, doi: 10.5194/gmd-10-1903-2017, 2017.
  2. Miralles, D.G., Holmes, T.R.H., de Jeu, R.A.M., Gash, J.H., Meesters, A.G.C.A., Dolman, A.J.: Global land-surface evaporation estimated from satellite-based observationsHydrology and Earth System Sciences15, 453–469, doi: 10.5194/hess-15-453-2011, 2011.

The current v3.5 datasets have been produced by Dr. Akash Koppa and Dr. Dominik Rains.

Scientific use

GLEAM datasets will not be used for commercial purposes.


Any feedback about the datasets and/or website are highly appreciatedand can be sent to