Researchers of the HUN-REN Centre for Energy Research have been developing atmospheric dispersion and dose calculation tools to assess radiological consequences of atmospheric releases affecting the public. These modelling tools require meteorological data to estimate atmospheric dispersion and radiological exposure in the vicinity of a nuclear installation under both normal operation and accidental conditions. In recent years, the integration of artificial intelligence (AI) and machine learning techniques has opened new possibilities for improving the accuracy, adaptability, and automation of such modelling tools.
In the last couple of years, the renewal of the meteorological measurement system of the KFKI Campus has been an important focus, including the installation of a new SODAR meteorological instrument. This system is able to measure meteorological parameters at various heights, including at ground level and at elevated heights of potential release points. By measuring meteorological parameters on a wider scale, the atmospheric dispersion calculations can be further refined. These high-resolution datasets also provide basis for developing and validating AI-supported and data-driven dispersion models.
The PhD student will join the research group and carry out the following tasks:
- Study of the literature published regarding the existing techniques of automatic dispersion modelling and dose calculation systems including recent AI-based approaches.
- Investigation of the special requirements for meteorological measurements for dispersion modelling and public dose assessment.
- Examination of the installation and operation aspects of SODAR measurement system, calibration of the system, interpretation of results, development of automatic sampling and documentation of data, including preparation of measurement data suitable for machine learning applications.
- Development of methods for automated and AI-assisted atmospheric dispersion and dose calculation, including real-time prediction, uncertainty analysis.
- Demonstration of the applicability of the developed system, conducting dose calculations based on real measurement data for the population living in the vicinity of the KFKI Campus, and comparison of conventional and AI-enhanced approaches.
The PhD candidate will have the opportunity to participate in international conferences in the field of radiological measurements and atmospheric dispersion modelling, as well as AI-related interdisciplinary venues. Professional support will be provided to the PhD candidate by the Radiation Protection Department of the HUN-REN Centre for Energy Research.
The applicant should have a proper level of knowledge and experience in the field of radiation protection, dosimetry and programming. Basic familiarity with artificial intelligence, machine learning, or data science methods may be beneficial. The applicant should be able to work independently, have new ideas and sufficient knowledge of English to be able to conduct review of international literature in the field. Knowledge and experience in programming is an advantage, especially in scientific computing and AI-related frameworks (e.g. Python-based systems).