Implications of land-cover misclassification for parameter estimates in global land-surface models: An example from the simple biosphere model (SiB2)

One of the primary applications of the global 1-km land-cover DISCover prod uct is to derive biophysical and ecological parameters for a range of land- surface models, including biosphere-atmosphere, biogeochemical, and ecologi cal models. The validation effort reported in this special issue enables a realistic assessment of the implications of misclassification errors for pa rameter estimates within the models. In most land-surface models, cover typ es are aggregated to coarser groupings than the 17 IGBP classes for estimat ing parameters, with aggregation schemes varying with individual models and individual parameters within each model. Misclassification errors are cons equential only when they occur between cover types that are not aggregated by the model. We use examples of two biophysical parameters-leaf area index and surface roughness-as estimated for use in the Simple Biosphere Model ( SiB2) and other modeling applications to quantify the effects of misclassif ication on parameter estimates. SiB2 relies on satellite data as well as la nd-cover information for estimating the biophysical parameters. Consequence s of misclassification are likely to be greater for those models that do no t use satellite data. Mean class accuracy based on those sites for which a majority of interpreters agreed (percentage of validation pixels classified correctly out of total number of validation pixels, averaged over all clas ses), adjusted by area of each cover type in the IGBP DISCover product, is 78.6 when all misclassification errors are included. By excluding misclassi fication errors when they are inconsequential for leaf area index and surfa ce roughness length estimates, mean class accuracies are 90.2 and 87.8, res pectively. The results illustrate that misclassification errors ore most me aningfully viewed in the context of the application of the land-cover infor mation.