Identification of areas of interest for radiological impact assessments in an evolving landscape context.

Regulations concerning potential future health impacts of the final disposal of radioactive wastes in geological disposal facilities are written in terms of annual dose to individuals who are representative of small groups living in the landscape in the vicinity of the repository site. As disposal programmes in Sweden and Finland have progressed towards licensing and construction, so too has detail describing the state and evolution of surface biosphere and the landscape around the proposed disposal sites increased. Simple and generic biosphere dose assessment models in early iterations have grown in complexity with increasing site-specific detail that aims to capture the radiologically significant features of the landscape into which future releases of radionuclides might credibly occur. Current dose assessment models used in support of license applications for disposal programmes in Sweden and Finland are highly complex and their application consequently lacks transparency. An alternative simpler approach to characterising landscape objects for dose assessment models would be beneficial in that it would offer an additional line of reasoning and would add clarity, thereby supporting the decision-making process of the regulatory authorities. In the context of coastal Fennoscandia, landscape change is relatively rapid and dramatic with post-glacial landrise transforming areas of the coastal seabed into terrestrial ecosystems over a period of a few thousand years, global sea level rise notwithstanding. The locations of the geosphere-biosphere interfaces for deep geologic disposal can be estimated with some precision but the nature of the receiving ecosystems at the time of the release is less certain. The approach described here provides a statistical quantification of key morphological characteristics of areas in the landscape where doses could arise, so as to better express uncertainties in dose modelling. The proposed method assumes that the variation in the morphology of potential release locations can be described by the variation in landscape objects seen in the landscape on a wider scale, providing a statistical description of the possible landscape objects, so allowing a more comprehensive range of potential future evolutions to be addressed. Our understanding of the evolution of the landscape, based on the kinds of terrain and ecosystem development models used by POSIVA in Finland and SKB in Sweden, suggests that objects identified in present-day maps can be used as analogues for a statistical characterisation of objects in the future landscape; objects identified in the observed topography and bathymetry can therefore serve as the basis for the statistical description of landscape dose objects over the period during which doses are likely to arise. Using digital elevation models around a disposal site in Finland, we show that the statistical descriptions of landscape dose objects at three times over a period of 10 kyear of the evolved landscape are sufficiently similar to establish the suitability of the approach. The aim of this statistical analysis is to supplement current methods for defining radiological assessment models so as to provide additional numerical support to both the simpler and more complex methods employed by implementors and regulators. The method has been developed in the context of the Swedish and Finnish regulatory review process and is referred to in the IAEA's revised BIOMASS methodology. We briefly address how the method might be applied in other landscape contexts.