Improving Forecasts of Volcanic Clouds: An Analysis of Observations and Emission Source Term Methods

Thumbnail Image


Publication or External Link





Volcanic eruptions can occur with little or no warning and explosively inject dense ash and sulfur dioxide (SO2) clouds high into the atmosphere. I investigated different types of observations and analysis methods used to monitor and quantify volcanic ash and SO2 clouds. I begin with an analysis of the 2010 eruption of Eyjafjallajökull, employing ash cloud transport modeling capabilities I developed for the Goddard Earth Observing System, Version 5 (GEOS-5). The emission source terms describing the initial state of the Eyjafjallajökull ash clouds were estimated using radar observations of the ash cloud’s initial injection altitude. Results of the initial simulations agreed with operational ash forecasts from the time of the eruption and with many other published studies, but showed notable disagreement with satellite observations. The emission source term was estimated using an alternative approach, yielding simulations that better matched satellite observations. I used the result to highlight limitations of radar observations not accounted for in previous studies of the Eyjafjallajökull ash clouds.

UV satellite observations are often used to monitor and quantify volcanic clouds of ash and SO2. I tested the limitations of the OMPS SO2 satellite observations using an Observing System Simulation Experiment (OSSE). The framework used GEOS-5 simulations of the atmospheric composition in the wake of a Pinatubo-like volcanic eruption to generate synthetic top-of-the-atmosphere (TOA) radiances. The TOA radiances served as input to the OMPS SO2 retrieval. In comparing the OMPS retrieval SO2 to the original GEOS-5 SO2, I found that the sulfate aerosols and ash can cause the OMPS SO2 retrieval to underestimate the total SO2 burden. These effects were amplified at increased satellite viewing angles.

I finish my analysis by looking at observations from the satellite-based Cloud-Aerosol Transport System (CATS), where I show that even under the time constraints of an operational forecast, the available CATS observations were able to improve forecasts of volcanic SO2 clouds.