The effects of non-photosynthetic vegetation on windblown dust emissions

Abstract:

Windblown dust emissions strongly depend on the surface vegetation coverage. Vegetation determines the fraction of land with potential for erosion, alters the threshold friction velocity that governs the initiation of saltation, and affects the surface friction velocity. Many current dust models use green or photosynthetic vegetation (PV) to represent the total vegetation due to the availability of various remote sensing indices for estimating PV coverage. However, the effects of vegetation elements are not fully accounted with the exclusion of brown or non-photosynthetic vegetation (NPV), which is prevalent in arid and semi-arid regions. This study incorporates the NPV into a dust model to examine the impacts of total vegetation (NPV+PV) on dust emission simulations. To estimate the fractions of PV, NPV, and bare soil, a non-linear spectral unmixing technique is performed, which uses 1) the reference spectra of each component calculated from the fractional coverage observations and corresponding satellite data for surface reflectance, and 2) seven bands of the MODIS Nadir Bidirectional Reflectance Distribution Function-Adjusted Reflectance (NBAR) product. The resulting spatially and temporally variable total vegetation coverage is then provided to a windblown dust scheme implemented in the Community Multiscale Air Quality (CMAQv5.3) model. Simulations are performed for Spring 2016 over the conterminous U.S., and the results are evaluated against observations from the IMPROVE and AQS sites. The effect of accounting for the total vegetation coverage on dust emissions are quantified.