Utilizing high spatiotemporal resolution soil moisture for dust storm modeling

The resolution and accuracy of initial model input are two fundamental factors for numerical modeling of climate and weather. High quality initial input assimilated in the model has significant impact on dust storm forecasting accuracy, and hence significantly influences the effectiveness of public health services and emergency management. Previous work with the Non-hydrostatic Mesoscale Model (NMM-dust) has been using static input data for parameters like soil moisture content. Since these parameters are changing seasonally or even daily, static input will reduce the model accuracy. This research investigates the sensitivity of the NMM-dust model in response to dynamic inputs of soil moisture, and evaluates the improvement of the model accuracy. The soil moisture data used for this research is generated by the Noah LSM, part of the North American Land Data Assimilation System (NLDAS) modeling suite. Numerical analysis is conducted by comparing simulation results using near-real-time soil moisture data with original model output using static one and MODIS Aqua atmosphere product in Deep Blue band.