Skip to main content

GLDAS Vegetation Class/Mask

GLDAS currently uses the land surface datasets from the Moderate Resolution Imaging Spectroradiometer (MODIS) as our "standard".  Some land surface models are however, tied to the Advanced Very High Resolution Radiometer (AVHRR) based datasets for the vegetation parameters.  The GLDAS version 2 products (GLDAS2) use the data described on this page. 


GLDAS Land/Sea Mask Dataset

GLDAS adopted the MOD44W MODIS Water Mask as the standard land/sea mask. The mask data [Carroll et al, 2009] is produced at University of Maryland and described here. Instead of deriving the mask from a land cover data as done in GLDAS1, the standard mask is "imposed" to land cover data and other surface parameter maps in GLDAS2. In order to create the land/sea masks on this page, the percent coverage of 0.01 degree water pixels in each GLDAS grid box was used such that grid boxes made up of more than 50% water were assigned a "water" value of 0 in the GLDAS land/sea mask, while all other grid boxes received a "land" value of 1.


GLDAS standard Land/Sea Mask


GLDAS Land Cover Dataset

GLDAS runs its one dimensional land surface models on vegetation-based tiles to simulate variability below the scale of the model grid squares. A tile is not tied to a specific location within the grid square. Instead, each tile represents the area covered by a given vegetation type.  A high resolution land cover dataset at 1 km resolution is the basis for designating tile space. In GLDAS2, different land cover classification schemes are used depending on the land surface model.

The land cover in GLDAS2/Noah is based on a 30 arc second MODIS vegetation data that uses modified IGBP classification scheme (see details). The predominant vegetation type in each 0.25 degree grid square is shown below.



The land cover in GLDAS2/Catchment is based on a 30 arc second USGS Global Land Cover Characteristics Data data that is mapped onto the Mosaic classification scheme.  The catchment model uses hydrologic catchment tiles and each tile has the primary and secondary vegetation types.  The predominant vegetation type in each 1 degree grid square is shown below.


In the 0.25 degree, Daily Catchment simulations (GLDAS-2.0 and GLDAS-2.2), vegetation tiling is not applied and the UMD land cover classification scheme [Hansen et al., 2000] is used.  The vegetation type in each 0.25 degree grid square is shown below. 


The land cover in GLDAS2/VIC is also based on a 1 km AVHRR land cover data of UMD classification scheme [Hansen et al., 2000].  The predominant vegetation type in each 1 degree grid square is shown below.


Vegetation Types for the Land cover schemes in the GLDAS2 products (key for above figures)

Model Noah Catchment VIC, Catchment (daily)

Modified MODIS IGBP 

Vegetation Type

Mosaic Vegetation Type UMD Vegetation Type
0 missing value missing value missing value
1 Evergreen Needleleaf Forest  Broadleaf Evergreen Trees Evergreen Needleleaf Forest
2 Evergreen Broadleaf Forest Broadleaf Deciduous Trees Evergreen Broadleaf Forest
3 Deciduous Needleleaf Forest  Needleleaf Trees Deciduous Needleleaf Forest 
4 Deciduous Broadleaf Forest Grassland Deciduous Broadleaf Forest
5 Mixed Forest  Broadleaf Shrubs, Bare Soil, Desert Soil Mixed Cover
6 Closed Shrublands  Dwarf Trees Woodland
7 Open Shrublands    Wooded Grasslands
8 Woody Savannas   Closed Shrublands 
9 Savannas   Open Shrublands 
10 Grassland   Grasslands
11 Permanent Wetland   Cropland
12 Cropland   Barren
13 Urban and Built-Up   Urban and Built-Up
14 Cropland/Natural Vegetation Mosaic    
15 Snow and Ice     
16 Barren or Sparsely Vegetated     
17 Ocean     
18 Wooded Tundra     
19 Mixed Tundra     
20 Bare Ground Tundra     
  • GLDAS2/Noah Dominant Vegetation Type Data (NetCDF): 0.25 degree, 1 degree
  • GLDAS2.0 & GLDAS2.1/Catchment Dominant Vegetation Type Data (NetCDF): 1 degree
  • GLDAS2.0 & GLDAS2.2/Catchment Dominant Vegetation Type Data (NetCDF): 0.25 degree
  • GLDAS2/VIC Dominant Vegetation Type Data (NetCDF): 1 degree