Evaluating the Frequency, Size, and Duration of Dust Storms in the United States using Satellite and Surface Data
December 31, 1969
Jennifer McGinnis
Hosted by Jeffrey Pierce (Advisor); Emily Fischer (Co-advisor); Steven Miller; Sheryl Magzamen (Environmental and Radiological Health Sciences)
Download Video
Abstract
The United States (US) experiences frequent dust storms that are primarily identified using ground-based monitoring networks. However, current ground monitor networks are unable to capture the full extent of dust exposure due to limited spatial coverage. Leveraging satellite data can help bridge these gaps in surface in situ data. Using the GOES-East geostationary satellite and a longwave-based satellite algorithm, we have identified dust events and evaluated the algorithm against surface-level air quality data. To our knowledge, this is the first work to evaluate a satellite-identified dust product for surface-level dust in the US. We found that these satellite data miss the vast majority of dust events observed by surface monitors, only correctly identifying 9.2% of elevated (>330 µg m-3) coarse particulate matter (PMcoarse; particulate matter with diameters >2.5 µm and <10 µm) concentration events in the contiguous US. Conversely, over 90% of the satellite-observed dust storms occur further than 10 km from monitoring stations. Although there were more frequent dust events in the Southwest US, dust events over the South Central US tend to be larger and last longer, according to satellite data from 2018-2023. In the Southwest, 0.69% of satellite-identified dust events (14 events) reached a maximum area of 10,000 km2 and in the South Central, 6.0% of dust events (53 events) reached 10,000 km2. Additionally, 8.7% (177 events) of dust storms in the Southwest and 16% of (142 events) dust storms in the South Central lasted longer than 5 hours. This analysis implies that large dust events are prevalent in the US and the existing monitoring networks fail to capture a substantial portion of them, highlighting the need for additional data sources and more comprehensive networks.
