A phenological model of pollen emissions for climate models

To understand the climatic relevance of pollen in the atmosphere, a prognostic model of pollen emissions for interactive input into climate models is needed. We use modern surface pollen count data to develop a model that can simulate pollen count over the seasonal cycle based on geography, vegetation type and meteorological parameters such as temperature, solar radiation, and wind speed. Analysis of pollen data from 2003-2010 suggests that latitude is a strong driver for the start day of year, with event duration controlled by temperature. This provides evidence that observed data can be used to develop a pollen emissions model that predicts pollen on interannual climate timescales. These emissions will be provided as input data to climate models and allow assessment of the direct and indirect effects of pollen on the atmosphere for present-day and potentially paleoclimate studies.