<p>In a science-based site selection process (StandAV), the Federal Republic of Germany searches for the site with the best possible safety for a repository of high-level waste (HLW) over a period of one million years. For this purpose, the geological subsurface of the German federal territory must be investigated and evaluated.</p><p>Challenges include the large area under investigation, that encompasses the entire federal territory of Germany with its large variability in geology, as well as the verification period which must be met to ensure the best possible long-term safety. Furthermore, an enormous amount of heterogenous geodata will have to be processed.</p><p>The application of AI-based methods in geosciences promises high potentials when dealing with large heterogenous data sets and cost- and time-consuming model calculations of complex and coupled processes. Accordingly, research on the application of AI has increased significantly in the geosciences over the last few years.</p><p>In our recent study &#8220;The use of artificial intelligence (AI) in the site selection process for a deep geological repository&#8221;, we succeeded developing an interdisciplinary assessment tool to evaluate the applicability of AI methods in geosciences in general and especially regarding their use for geoscientific issues in the StandAV. Here, we focus on potentials and challenges of applying AI in geosciences with respect to geological key activities in the StandAV. Thus, we emphasize on limitations that may arise from the use of AI regarding key activities in StandAV and propose necessary conditions for its applicability in the future.</p><p>Our results show that AI methods are superior to conventional methods, especially when it comes to data management and dealing with large geological data sets and model calculations of complex long-term and coupled geological processes. However, AI methods are generally only transferable to the geoscientific issues in the StandAV with methodological and subject-specific adaptations. Nevertheless, sufficient data, both in quality and quantity, is a prerequisite for the use of AI. Our study also shows that AI should only be used in a supportive way to tackle geological issues in the key activities and must not have any decision-making power when used in the StandAV.</p><p>High demands must be placed on the traceability of the applied AI methods. AI methods that do not meet the transparency requirements of the StandAV bear considerable risks of jeopardizing the trust of the population in the participation process. This could increase the general suspicion and scepticism towards AI in the public perception. Therefore, we strongly recommend to always evaluate and validate iteratively all methods and providing results to the public when applied to the key activities of the StandAV.</p><p>Title of study: &#8220;The use of artificial intelligence (AI) in the site selection process for a deep geological repository&#8221;, <br>a project on behalf of the Federal Office for the Safety of Nuclear Waste Management (BASE-FKZ 4721E03210)</p>
The risk of radioactive contamination in the biosphere surrounding the Asse salt mine has been assessed to determine the possible radioactive exposure to humans if the mine collapses. Geological conditions and anthropogenic activities have made the mine instable and allow salt-saturated ground water to seep in. This uncontrolled brine inflow significantly increases the risk of the mine collapsing. If the mine collapses, the brine will be pressed into groundwater, where the radionuclides can migrate into the biosphere and cause radioactive exposure. The key issue discussed in this paper is estimating the short- and long-term radiation burden for humans under several possible scenarios of radionuclide release. Only a radioecological model able to quantify and estimate processes taking place can generate usable results. This work develops the radioecological model describing both radionuclide migration and the resulting radiological exposure along several exposition pathways. Development of the model took into account the sorption processes, solubility limits and special aspects of decay chain migration. The radiological exposure was estimated under non-equilibrated conditions for the case of short-time expositions. At the end of this paper, the model’s background, the results of the computations and a comparison of several scenarios will be presented.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.