Following the 2010 Eyjafjallajokull eruption the London VAAC implemented an operational dispersion modeling approach that has a “default” ash particle size-distribution between 0.1 and 100 microns. Five percent of the total erupted mass is taken to be within this size range in the distal part of the plume (i.e. after near-source fallout and aggregation effects have taken place). These assumptions appear to give a reasonable fit for the plume, given the other uncertainties involved. However, both the particle size and distal fine ash mass fraction can be changed during an event to better represent the actual eruption. In this presentation we will show the sensitivity of the model to such changes. We will describe how a change in the mass fraction was operationally implemented during the May 2011 Grimsvotn eruption and discuss some of the implications for transparency of process.

One of the challenges of validating such model sensitivity tests is finding comparable observations. Satellite retrieval techniques are advancing and remotely sensed information on ash mass in the total column, plume height and detected particle size is now available to VAACs in areas of the world where appropriate sensors exist. However these products are not directly comparable with the ash concentration charts that are now operationally produced for the European and North Atlantic region. We will show the challenges of comparing volcanic ash concentration charts designed for use by the aviation industry with satellite retrievals of total mass and the advantages of also using modeled total column information for sensitivity comparisons.