Global Modeling

Global modeling can provide (spatially, temporally, and spectrally) complete and consistent datasets for all aerosol properties. Concerns exist, however, as to the accuracy of the underlying assumptions (e.g., emissions, transport, water uptake) and parameterizations (e.g., aerosol processing, interactions with clouds). With rather general constraints (e.g., column and component-integrated data from remote sensing), there is significant diversity in aerosol global modeling, especially at modeling substeps (Textor et al. 2006). To counteract this diversity and to establish characteristic particle properties from global modeling, aerosol simulations of more than twenty different models were considered. All of these models employed advanced aerosol modules, which distinguished between aerosol components of dust, sulfate, sea salt, organic carbon, and black carbon (Kinne et al. 2006). Simulated monthly averages were regridded to a common 1°x1° latitude/longitude horizontal resolution. The local median value of monthly averages suggested by all models at any grid point was picked; thereafter, these median values were combined to define monthly fields from global modeling for AOD, ro0, and AnP. These median fields have the advantage that extreme behavior (e.g., outliers) of individual models is suppressed. In addition, these median model fields tend to score better when evaluated than individual models (Schulz et al. 2006).

0 0

Post a comment