Using Satellite Observations to Evaluate the AeroCOM Volcanic Emissions Inventory and the Dispersal of Volcanic SO2 Clouds in MERRA

Abstract:

Simulation of volcanic emissions in climate models requires information that describes the eruption of the emissions into the atmosphere. While the total amount of gases and aerosols released from a volcanic eruption can be readily estimated from satellite observations, information about the source parameters, like injection altitude, eruption time and duration, is often not directly known. The AeroCOM volcanic emissions inventory provides estimates of eruption source parameters and has been used to initialize volcanic emissions in reanalysis projects, like MERRA. The AeroCOM volcanic emission inventory provides an eruption’s daily SO₂ flux and plume top altitude, yet an eruption can be very short lived, lasting only a few hours, and emit clouds at multiple altitudes. Case studies comparing the satellite observed dispersal of volcanic SO₂ clouds to simulations in MERRA have shown mixed results. Some cases show good agreement with observations Okmok (2008), while for other eruptions the observed initial SO₂ mass is half of that in the simulations, Sierra Negra (2005). In other cases, the initial SO₂ amount agrees with the observations but shows very different dispersal rates, Soufriere Hills (2006). 
 In the aviation hazards community, deriving accurate source terms is crucial for monitoring and short-term forecasting (24-h) of volcanic clouds. Back trajectory methods have been developed which use satellite observations and transport models to estimate the injection altitude, eruption time, and eruption duration of observed volcanic clouds. These methods can provide eruption timing estimates on a 2-hour temporal resolution and estimate the altitude and depth of a volcanic cloud. 
To better understand the differences between MERRA simulations and volcanic SO₂ observations, back trajectory methods are used to estimate the source term parameters for a few volcanic eruptions and compared to their corresponding entry in the AeroCOM volcanic emission inventory. The nature of these mixed results is discussed with respect to the source term estimates.