Calibration of a wind erosion and dust emission model using continental-scale geospatial soil and vegetation datasets
Abstract
Aeolian processes are involved in the evolution and persistence of dryland ecosystems. The aeolian transport of sediments in dryland ecosystems is dependent on the presence of surface roughness elements such as vegetation. A previously developed and validated wind erosion model (Okin, 2008) based on drag partition (i.e., total wind shear stress apportioned to between soil surface and roughness elements) accounts for the distribution of roughness elements and the erodible gaps. A study by Li et al., 2013 calibrated the wind erosion model developed by Okin (2008) using field measurements of vegetation and soil parameters across several dust producing locations of the western United States to estimate a set of best fit model parameters. Here, we build and expand the Li et al., 2013 study and use gridded soil texture products and a recently developed machine learning algorithm to generate high resolution (30 meters) estimates of distribution of vegetation heights and erodible gaps across space and time to calibrate Okin, 2008 wind erosion model and to predict gridded horizontal sediment flux at several long-term wind erosion monitoring sites in the western United States. Using a cross validation approach, we compared modeled horizontal aeolian flux with observed aeolian flux to generate an optimum set of model parameters. We found that the average relative error of modeled horizontal aeolian flux generated with spatial data was 1.5 and the values of the model parameters were comparable to the published estimates in the original study by Li et al., 2013. With the advancement in computing capacity and the availability of high-resolution remote sensing imagery, the calibrated wind erosion model can be applied to model aeolian flux and dust emission across space and time in the western United States.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.A35E1671B