Comparison of radioactivity data measured in PM10 aerosol samples at two elevated stations in northern Italy during the Fukushima event.

The follow-up of Fukushima radioactive plume resulting from the 11th March 2011 devastating tsunami is discussed for two Italian stations in the northern Apennines: Mt. Cimone (Modena) and Montecuccolino (Bologna). Radioactivity data collected at both stations are described, including comparison between local natural background of airborne particulate and artificial radioactivity referable to the arrival of the radioactive plume and its persistence and evolution. Analysis of back-trajectories was used to confirm the arrival of artificial radionuclides following atmospheric transport and processing. The Fukushima plume was first detected on 3rd April 2011 when high volume sampling revealed the presence of the artificial radionuclides (131)I, (137)Cs and (134)Cs. The highest activity concentrations of these nuclides were detected on 5th April 2011 at the Montecuccolino site. Fukushima radioactivity data at the two stations were usually comparable, suggesting a good vertical mixing of the plume; discrepancies were occasional and attributed to different occurrence of wet removal, typically characterized by a scattered spatial pattern. To understand the relevance to the local population of the extra dose due to the Fukushima plume, atmospheric activities of the related artificial nuclides were compared to those of the main natural radionuclides in ambient particulate, and found to be lower by over one order of magnitude. Radiation doses referable to Fukushima, maximized for a whole year occurrence at the highest activity level observed at our stations in the weeks affected by the Japanese plume, were estimated at 1.1 µSv/year.

Solving radiative transfer problems in highly heterogeneous media via domain decomposition and convergence acceleration techniques.

This paper deals with finding accurate solutions for photon transport problems in highly heterogeneous media fastly, efficiently and with modest memory resources. We propose an extended version of the analytical discrete ordinates method, coupled with domain decomposition-derived algorithms and non-linear convergence acceleration techniques. Numerical performances are evaluated using a challenging case study available in the literature. A study of accuracy versus computational time and memory requirements is reported for transport calculations that are relevant for remote sensing applications.