Input data set: Land Transformation (ISIMIP4)

The land transformation dataset provides annually varying land-use states and transitions for ISIMIP4b simulations. It describes the spatial distribution and temporal evolution of land-cover and land-management categories, including transitions between them, as well as associated management practices. The dataset enables impact models to account for historical and future land-use change, such as deforestation, afforestation, agricultural expansion, crop rotations, pasture dynamics, and wood harvest, in a manner that is consistent with the CMIP7 climate simulations. The same prescribed land-use forcing is used for both concentration-driven and emission-driven climate simulations, as land-use change is specified externally and is not simulated interactively by the Earth System Models.

The dataset consists of annually resolved, globally gridded land-use transitions and shifting cultivation following the Land-Use Harmonization version 3 (LUH3) framework required by Earth system and impact models. Historical forcing is provided for the CMIP7 historical period, while scenario datasets represent future land-use development consistent with the selected ScenarioMIP pathways. The data are provided on a regular global latitude-longitude grid in 15 and 30 arcmin resolutions in NetCDF format.

Land transformation forcing is taken from the CMIP7 Input4MIPs Land-Use Harmonization version 3 (LUH3) datasets produced by the University of Maryland. For ISIMIP4b, the following datasets are used:

Historical: UofMD-landState-3-1-2
Very Low scenario: UofMD-landState-vl-3-1
High scenario: UofMD-landState-h-3-1

These datasets are distributed through the Input4MIPs infrastructure and are documented at:

https://input4mips-cvs.readthedocs.io/en/latest/dataset-overviews/land-use/

The historical and scenario datasets are based on the LUH3 framework and provide harmonized land-use states and transitions consistent with the Integrated Assessment Model pathways adopted for CMIP7. The prescribed land-use fields are applied unchanged to both concentration-driven and emission-driven climate simulations.