Significance for dust emission modelling of omitting changing aerodynamic roughness

TitleSignificance for dust emission modelling of omitting changing aerodynamic roughness
Publication TypeConference Paper
Year of Publication2018
AuthorsChappell A, Webb N
Conference NameInternational Conference on Aeolian Research
Date Published06/2018
Conference LocationBordeaux, France
ARIS Log Number364796
Abstract

Experiments of dust emission processes have informed the development and evolution of dust emission schemes that underpin global dust emission models (GDEMs) [1, 2]. Despite these improvements, many GDEMs retain only crude approximations of the emission processes because they do not adequately balance upscaling, parameterisation parsimony and process fidelity [3]. The GDEMs focus on static and homogeneous land surface aerodynamic roughness (and soil characteristics, not tackled here). This approach omits the spatio-temporal dynamics of soil surface shear stress due to land use / land management change, CO2-fertilisation and drought amplification. Consequently, observed reduction in dust emission in e.g., the west African Sahel cannot be replicated by GDEMs [4, 5]. The same omission very likely causes modelled wind speed outputs to be much more uncertain than currently recognized. For example, observed wind stilling is absent in modelled wind speed outputs [6]. To establish the significance of these omissions, we improve shear stress dynamics using an albedo-based parameterization of the aerodynamic roughness normalized by wind speed [7]. We use MODIS albedo (MCD43A1 8-daily, 500 m) and GLDAS modelled wind speed data and other global datasets to investigate the spatio-temporal dynamics of global and dust emission. Global drylands show distinct and opposing trends. Modelled winds dominate dust emission inconsistent with trends. Omitting changing aerodynamic roughness in wind data and dust emission will render GDEM outputs ineffective and interpretations unreliable. A solution is to use u_S*/U_f and embed GDEMs in to LSMs coupled to GCMs. Improved modelling of dust emission spatio-temporal dynamics will likely reduce uncertainty in radiative forcing and biogeochemical cycling for climate change projections.