1. Zdenka Stojanovska, Faculty of Medical Sciences, Goce Delčev University Stip, Macedonia
2. Zora S. Zunić, Vinca Institute of Nuclear Sciences, University of Belgrade, Belgrade, Serbia
3. Zoran Ćurguz, Saobraćajni fakultet Doboj, Republic of Srpska, Bosnia and Herzegovina
4. Predrag Kolarž, Institute of Physics, University of Belgrade, Serbia
5. Ivan Boev, Faculty of Natural and Technical Sciences, Goce Delcev University, Stip, Republic of Macedonia, Macedonia
6. Blazo Boev, Faculty of Natural and Technical Sciences, Goce Delcev University, Stip, Republic of Macedonia, Macedonia
The first systematic research of indoor toron concentrations in Republic of North Macedonia began in 2008/2009 when the National Survey was conducted in 300 dwellings. Furthermore, a few years later, in 2012, a smaller campaign was carried on, in which indoor thoron measurements were done in 30 schools. In both studies, the same methodology was used; it means that the indoor thoron concentrations were measured with Raduet detectors deployed in ground floor room at a distance of more than 50 cm from any wall surface with an exposure period of 3 months. The main conclusions from both studies were that indoor thoron concentrations are subject to temporal and spatial variations.
Considering this, we decided to developed models that could make predictions of annual thoron concentrations due to seasonal measurements performed in buildings built with different building materials from different parts of the country. Since the log-transformed values of the measured data were with normal distribution moreover the same data grouped by geotectonic units and type of building materials were with homogeneity of variance, so the criteria for multiple linear regression was satisfied.
For the models which presented thoron seasonal variations, the Pearson coefficient of determination (R2) was in the range from 27% for summer measurements in buildings with basement up to 54% for spring measurements. The model for thoron spatial expressed through geotectonic units and building materials had a value of R2=21%. Although statistically significant, indoor thoron seasonal and spatial models are in general with poorer performance compared to the same linear models for indoor radon in Northern Macedonia.