Dust Forecasting: A Study of U.S. Air Force Numerical Weather Modeling Performance and Skill

Dust and aerosols often adversely impact military activities. One of the most obvious manifestations of these impacts can be seen through the spring and summer in Southwest Asia and in Equatorial Africa when the combination of the summer dry season and strong winds loft dust into the atmosphere and reduce horizontal visibility to one kilometer or less. In these regions, dust storms present the greatest challenge to aviation forecasting, presenting great danger to aircraft and halting operations in many cases. To better understand numerical modeling system ability to simulate atmospheric dust and its impacts, we perform studies using the Global Air Land Weather Effects Model (GALWEM), the Air Force unique instantiation of the United Kingdom Met Office’s Unified Model (UM). We employ both the 2-bin and 6-bin configurations of the dust scheme and compare model output to observations (both satellite and in-situ) and to other models performing dust simulations over the same regions. Our primary period of interest is April-August 2017. However, we also present results from a similar study of GALWEM in April-August 2016 using only the 2-bin dust configuration. We present results of these studies, highlighting strengths and weaknesses of the dust scheme. With an operational baseline capability identified, we present a roadmap to improve the model’s ability to forecast dust, including proposals to the model configuration and physics schemes. We also provide recommendations on other techniques that may be used to combine direct model output with data analytics techniques to help Air Force meteorologists anticipate these mission limiting events.