Improving dust emission in WRF-Chem GOCART scheme using a high-resolution erodibility dataset

Mineral dust is a major atmospheric aerosol, affecting climate, air quality, and human health through radiative and microphysical processes. The Iberian Peninsula is frequently impacted by dust intrusions from North Africa, leading to episodic exceedances of PM₁₀ concentrations that challenge operational air quality forecasts. Accurate simulation of dust emission and transport remains difficult due to uncertainties in soil erodibility, land surface characteristics, and meteorological drivers.

In this study, we assess the impact of two newly developed high-resolution soil erodibility datasets on regional dust simulations using WRF-Chem with the GOCART scheme. The first dataset, EROD, improves dust source representation by integrating fine-resolution topography (GMTED2010), achieving 0.0625° (≈5 km) resolution globally and 1 km locally for the Iberian Peninsula. The second dataset, SOILHD, further refines dust source characterization by incorporating local-scale soil composition (sand, silt, clay fractions) and removing areas erroneously classified as bare soil, reaching 1 km resolution globally. These datasets aim to capture the spatial heterogeneity of dust sources, which is critical in semi-arid regions with sparse vegetation and variable soil properties.

We conduct WRF-Chem simulations for five periods between 2022 and 2025, representing a range of dust episodes with local and long-range transport. Model performance is evaluated against PM₁₀ measurements from the SINQLAIR network across coastal and inland stations in the Region of Murcia. Results indicate that the high-resolution datasets substantially improve the spatial and temporal representation of dust emissions. Inland and low-anthropogenic-influence stations show better agreement with observed PM₁₀ peaks in both magnitude and timing compared to simulations using standard coarse-resolution erodibility fields. At coastal and industrially influenced sites, improvements are more limited due to missing anthropogenic emissions and additional aerosol components, but statistical metrics such as correlation, Mean Bias Error (MBE), and Root Mean Square Error (RMSE) still indicate significant enhancement.

Overall, the results demonstrate that high-resolution, type–aware soil erodibility datasets significantly enhance the skill of dust simulations in WRF-Chem, reducing biases and capturing observed variability more accurately. These findings underscore the importance of detailed soil and topographic information for regional dust modeling and highlight the potential benefits of incorporating such datasets into operational dust forecasting systems.