The current FLDAS approach is to use a land cover dataset that conforms with a vegetation classification scheme available in the land surface model. The model's default vegetation parameter table is directly used in the FLDAS simulation.

The FLDAS Noah and Noah-MP models use the IGBP modified 20-category land cover dataset, which includes 16 IGBP vegetation classes and 3 classes of Tundra. The vegetation parameter table for the models looks like: https://ral.ucar.edu/sites/default/files/public/product-tool/unified-no… . This table contains three set of parameters for different classification schemes, and the second set, "MODIFIED_IGBP_MODIS_NOAH", is relevant to FLDAS Noah and Noah-MP model simulations.

The 14-class University of Maryland land cover dataset is used in the FLDAS-Forecast Catchment model simulations. This dataset is developed based on Advanced Very High Resolution Radiometer imageries between 1981 and 1994, as described at: http://app.earth-observer.org/data/basemaps/images/global/LandCover_512….