Source apportionment methodologies for radioactivity, volcano, and forest fires

In several cases, classical data assimilation of the concentration fields provides insufficient information and incomplete analysis. Particularly difficult are the cases with incomplete coverage of observations in space (e.g., missing/sparse profile information) or in time (e.g., infrequent satellite overpasses), with buoyant and vertically extended sources, etc. A way to reduce the impact of these gaps is to extend the data assimilation towards the source apportionment task. Efficiency of this approach was demonstrated for volcanoes by the recent ESA-VAST and SMASH projects. Particular attention will be paid to relation of the quality and abundance of the observational information and the quality of the source apportionment, analysis products, and following forecasts.