Improving volcanic ash predictions with the HYSPLIT dispersion model by assimilating MODIS satellite retrievals

摘要: Abstract. Currently, the National Oceanic and Atmospheric Administration (NOAA) Weather Service (NWS) runs HYSPLIT dispersion model with a unit mass release rate to predict transport of volcanic ash. The predictions provide information for Volcanic Ash Advisory Centers (VAAC) issue advisories meteorological watch offices, area control centers, flight others. This research aims quantitative forecasts ash distributions generated by objectively optimally estimating source strengths, vertical distribution, temporal variations using an observation-modeling inversion technique. In this top-down approach, cost functional is defined quantify differences between satellite measurements column-integrated concentrations weighted observation uncertainties. Minimizing adjusting sources provides emission estimates. As example, MODIS (Moderate Resolution Imaging Spectroradiometer) retrievals 2008 Kasatochi clouds are used test inverse system. Because include cloud top height but not bottom height, there different diagnostic choices comparing results observed loadings. Three options presented tested. Although estimates vary significantly options, subsequent all show decent skill when evaluated against unassimilated observations at later times. Among three integrating over layers yields slightly better than from surface up or single layer. Inverse tests also that including ash-free region constrain beneficial current case. addition, extra constraints on terms can be given explicitly enforcing no-ash atmosphere columns above below height. However, in case such helpful modeling. It found simultaneously assimilating times produces hindcasts only most recent observations.