#L09 Evaluating CMIP5 and CMIP6 land use forcings for ACME v1

Abstract

Land-use activities have influenced the Earth’s climate system by significantly altering biogeochemical and biogeophysical properties at local to planetary scales. Yet the precise magnitude and character of land-use effects on climate remains uncertain, and this uncertainty in turn limits the accuracy of future projections. To address this issue, we have developed a new set of global gridded land-use forcing datasets to link historical land-use data and future projections in a standard format required by climate models. For CMIP6, land-use has become a required forcing and use of the land-use harmonization dataset is designated as an entry card for participating in CMIP6 experiments. The second generation land-use harmonization (LUH2) product expands on the land-use forcing dataset generated for CMIP5 and includes new land-use states (and all transitions between them), updated inputs, and new management layers. 

As a first step towards understanding the impacts of different land-cover datasets on the Earth climate system, we compared the land-cover information and surface datasets generated from two versions of the land-use harmonization products (LUH1 and LUH2). This process involved converting the latest version of the land-use harmonization dataset (LUH2) into the land-use harmonization (LUH1) dataset format by mapping the thematically and spatially more resolved LUH2 land use states and transitions to the LUH1 land use states (primary, secondary, crop, pasture and urban). The dataset in LUH1 format was processed through a land use translator to produce a surface dataset. We assessed the two land-use datasets across a suite of diagnostics, including global comparisons at different time-periods for the following metrics: net and gross transitions, land use area for different land use states, secondary age profile, increase in secondary land due to human impacts, wood clearing estimates, and potential biomass and forest area estimates. To assess the land-cover and surface datasets produced from the land-use datasets, we compared the changes in cropland and forest plant functional types between the datasets. Implications of differences in the two datasets on carbon cycle simulations will also be presented.