From the Impact of Radon to Forecasting: the Different Shades of Environmental Radioactivity
ISEF · 2026 Physics and Astronomy
Overview
Every day, without realizing it, we live immersed in environmental radioactivity, a natural and generally harmless phenomenon. In the past, its strong correlation with meteorological variables has been verified, and the possibility of estimating its value using these variables has been demonstrated. In this project, this physical phenomenon is investigated in greater depth by measuring different combinations of alpha, beta, and gamma radiation, as well as radon gas concentration. Thanks to these measurements, performed mainly with instruments designed and built by the finalist, it was possible to verify phenomena already known in the literature and, above all, to achieve some original observations. The weak correlation between the different types of radiation and radon concentration showed that, on the balcony where the instruments were installed, several different phenomena contribute to environmental radioactivity. In fact, it was demonstrated that, contrary to what is often reported, the presence of radon - favored by winds coming from the Apennines behind Genoa - is not the dominant cause of radioactivity. It was also shown that, by adding information on wind and rain, meteorological variables allow for an accurate estimation of both different types of radiation and radon concentration. Above all, a method for forecasting future radioactivity has been introduced for the first time in the scientific literature. An LSTM neural network, receiving input values for radioactivity over the last six days, together with meteorological variables (which are essential for achieving good accuracy), can forecast the average radioactivity over the next two days.
Competition history
- ISEF 2026
Resources
Related projects
ISEF · 2025
Environmental Radioactivity and Meteorological Variables: Statistical Relationships and Machine Learning Regression
ISEF · 2022
Polaris II: Gamma Ray Visualization System and Risk Mitigation Through Real-Time Radionuclide Atmospheric Dispersion Modeling
ISEF · 2017
Radiochemistry and AI: Development of an Analytical Capability Using Machine Learning Algorithms for Early Detection of Radionuclides in the Environment
ISEF · 2018
Affordable Web-Enabled Continuous Radon Detector
Closest projects by meaning, across every fair and year in the corpus.
Source: Regeneron International Science and Engineering Fair