Mapping of ESA's Climate Change Initiative land cover data to plant functional types for use in the CLASSIC land model

Plant functional types (PFTs) are used to represent vegetation distributionin land surface models (LSMs). Previous studies have shown large differencesin the geographical distribution of PFTs currently used in various LSMs,which may arise from the differences in the underlying land cover productsbut also the methods used to map or reclassify land cover data to the PFTsthat a given LSM represents. There are large uncertainties associated withexisting PFT mapping methods since they are largely based on expert judgementand therefore are subjective. In this study, we propose a new approach toinform the mapping or the cross-walking process using analyses fromsub-pixel fractional error matrices, which allows for a quantitativeassessment of the fractional composition of the land cover categories in adataset. We use the Climate Change Initiative (CCI) land cover productproduced by the European Space Agency (ESA). Previous work has shown thatcompared to fine-resolution maps over Canada, the ESA-CCI product providesan improved land cover distribution compared to that from the GLC2000dataset currently used in the CLASSIC (Canadian Land Surface SchemeIncluding Biogeochemical Cycles) model. A tree cover fraction dataset and afine-resolution land cover map over Canada are used to compute the sub-pixelfractional composition of the land cover classes in ESA-CCI, which is thenused to create a cross-walking table for mapping the ESA-CCI land covercategories to nine PFTs represented in the CLASSIC model. There are largedifferences between the new PFT distributions and those currently used inthe model. Offline simulations performed with the CLASSIC model using theESA-CCI-based PFTs show improved winter albedo compared to that based on theGLC2000 dataset. This emphasizes the importance of accurate representationof vegetation distribution for realistic simulation of surface albedo inLSMs. Results in this study suggest that the sub-pixel fractionalcomposition analyses are an effective way to reduce uncertainties in the PFTmapping process and therefore, to some extent, objectify the otherwisesubjective process.