Global CMIP6-based datasets of vegetation-responsive and feedback-deflated potential evapotranspiration

Conventional potential evapotranspiration (PET) formulations often neglect evolving vegetation structure, physiological responses to rising atmospheric CO₂ concentration and vapor pressure deficit (VPD), and feedback-driven amplification of atmospheric aridity. To address these limitations, we present a global gridded dataset of vegetation-responsive PET (PET_veg) and its feedback-deflated counterpart, wet-environment PET (WET_veg), derived consistently from 12 Earth system models in the Coupled Model Intercomparison Project Phase 6 (CMIP6). PET_veg is computed using a physically and physiologically based two-source Penman–Monteith formulation that explicitly accounts for leaf area index, atmospheric CO₂, and VPD through dynamic aerodynamic and surface resistance terms. WET_veg is obtained by applying the complementary evaporation principle to reconstruct the atmospheric state that would prevail over a saturated surface, thereby removing land–atmosphere feedback amplification of air temperature and VPD. The dataset spans the historical period (1850–2014) and four future Shared Socioeconomic Pathways (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5) for 2015–2100, provided on a common 1° × 1° grid at monthly resolution. Together, PET_veg and WET_veg offer internally consistent yet physically distinct representations of atmospheric evaporative demand, enabling transparent evaluation of PET formulation uncertainty in hydrological, ecological, and climate-impact studies.