The promise of ensemble forecasts to improve tracking of emissions from the Hawaiian Kilauea Volcano

Air quality ensembles provide information that is important to hazard mitigation efforts that deterministic forecasts cannot provide. Not only do ensembles provide information about the range of likely forecast outcomes, but also about the uncertainty of the forecast. This information is vital when hazardous air quality conditions may call for forced evacuations, as was the case during the 2018 Lower East Rift Zone (LERZ) eruption at Kilauea Volcano, HI.

Vog (volcanic smog) contains the air pollutants sulfur dioxide (SO₂) and sulfate aerosol (SO₄). The Vog Measurement and Prediction (VMAP) program has produced real-time deterministic forecasts of SO₂and SO₄since 2011 and ensemble forecasts for the public since 2017. We assimilate satellite-derived sulfur dioxide concentrations and use a plume model methodology to assess contributions to SO₂ forecast uncertainty that arise from errors in the initial conditions. Among the major contributors we examine are uncertainty in the wind field, initial concentrations, and plume vertical extent and structure. From our assessment, we are able to identify gaps in observing systems that can be filled to realize forecast improvements for future eruptive episodes and increased lead-time for air quality warnings.