A006-0006
Effects of Land Cover Changes on Isoprene Emissions Evaluated Using Spaceborne Formaldehyde Data From OMI Sensor and Multiyear Model Simulations
Abstract:
An accurate representation of land cover (LC) distributions is a key condition for simulating biosphere-atmosphere interactions. Our objective is to test the implementation of different global spaceborne time-dependent plant functional type (PFT) distributions on the simulations of biogenic isoprene emissions over 2001-2018. The tested PFT distributions include (1) the MODIS dataset (Friedl and Sulla-Menashe, 2019), (2) the MODIS dataset modified to account for tree cover changes from Hansen et al. (2013), and (3) the ESA-CCI dataset (Poulter et al. 2015). The comparisons show a large variability in the representation of the tree cover. In order to assess the uncertainty in isoprene emissions and their trends associated to LC changes, we perform global simulations using the MEGAN emission model (Guenther et al. 2012) coupled with the MOHYCAN multi-layer canopy model (Müller et al. 2008) over 2001-2018 using either static or annually-changing LC (from MODIS or MODIS-Hansen). The derived emission datasets are further evaluated using top-down information provided by satellite observations of formaldehyde, a high-yield isoprene oxidation product. To this end, we perform multiyear simulations of atmospheric composition with the IMAGES CTM using each of the different isoprene emission datasets over 2005-2016. We evaluate the resulting HCHO column distributions and their interannual variability against observed HCHO columns from the OMI sounder over densely forested areas.