A New Plant Functional Type Dataset for Earth System Modeling: Integrating ESA-CCI MRLC, LUH3, and C3/C4 Partitioning for CMIP7

Accurate representation of land use and land cover change (LULCC) is critical for simulating land–atmosphere interactions and biogeochemical feedbacks in Earth system models. Here we present a new, spatially and temporally consistent global Plant Functional Type (PFT) dataset designed for land surface and Earth system modeling, with direct applicability to CMIP7 model simulations. The dataset integrates high-resolution satellite-based land cover information with the latest Land-Use Harmonization dataset ( LUH3, updated from LUH2 – Hurtt et al., 2020), while improving the representation of grassland and cropland functional diversity through explicit C3/C4 partitioning. 

The historical PFT baseline is derived from the ESA Climate Change Initiative multi-resolution land cover (ESA-CCI MRLC) product for 1992–2022. Annual maps of fractional PFT composition are generated at 300 m resolution using a hierarchical framework that combines satellite observations, auxiliary datasets (tree cover, canopy height, surface water, urban extent), and bioclimatic constraints (Harper et al., 2023). Fourteen generic PFTs are represented, capturing sub-grid heterogeneity relevant for land surface and dynamic vegetation models. These maps are aggregated to coarser resolutions (≥0.1°) to support Land Surface Model (LSM) simulations.

To improve functional realism, natural grasslands are partitioned into C3 and C4 types using a synthesis of optimality-based estimates of potential C4 grass distribution (Luo et al., 2024), satellite-derived herbaceous cover, and complementary global datasets. Croplands are further refined using LUH3 to distinguish managed grasslands and crop types. LUH3 historical land-use states and transitions (850–2024) are merged with the satellite-based PFT baseline, allowing land-use changes such as deforestation, crop expansion, shifting cultivation, and wood harvest to be prescribed consistently through time.

The resulting harmonized PFT–land-use dataset provides a robust forcing for transient Earth system simulations from the pre-industrial period to the present. It improves the representation of LULCC processes, grassland physiology, and human land management, and is fully compatible with LSMs such as ORCHIDEE and CMIP7 requirements. This product enables more consistent assessments of land-based climate feedbacks and anthropogenic impacts on the Earth system.