| To detect, analyze and predict the movement of volcanic ash in real time, dispersion models and satellite remote sensing data are important. A combination of both approaches is discussed here to enhance the techniques currently used to quantify volcanic ash emissions, based on case studies of the eruptions of the Kasatochi (Alaska, USA, 2008), Mount Redoubt (Alaska, USA, 2009) and Sarychev Peak (Russia, 2009) volcanoes. Results suggest a quantitative approach determining masses from satellite images can be problematic due to uncertainties in knowledge of input values, most importantly the ground surface temperature required in the mass retrieval. Furthermore, a volcanic ash transport and dispersion model simulation requires its own set of accurate input parameters to forecast an ash cloud's future location. Such input parameters are often difficult to assess, especially in real time volcano monitoring, and default values are often used for simplification. The objective of this dissertation is to find a quantitative comparison technique to apply to satellite and volcanic ash transport and dispersion models that reduces the inherent uncertainty in the results. The binary `Ash - No Ash' approach focusing on spatial extent rather than absolute masses is suggested, where the ash extent in satellite data is quantitatively compared to that in the dispersion model's domain. In this technique, neither satellite data nor dispersion model results are regarded as the truth. The Critical Success Index (CSI) as well as Model and Satellite Excess values (ME and SE, respectively) are introduced as comparison tools. This approach reduces uncertainties in the analysis of airborne volcanic ash and, due to the reduced list of input parameters and assumptions in satellite and model data, the results will be improved. This decreased complexity of the analysis, combined with a reduced error as the defined edge of ash cloud is compared in each method rather than defined threshold or mass loading, will have important implications for real time monitoring of volcanic ash emissions. It allows for simpler, more easily implemented operational monitoring of volcanic ash movements. |
