Assessing the Influence of Topography and Environmental Factors on Gamma-Ray Air Dose Rates Under Canopies

Researchers have measured aerial gamma-rays using remote measurement devices to estimate the radionuclide concentrations in soils. However, an issue arises when measuring air dose rates in forested areas, where canopies exist between the remote sensing device and the ground. When researchers do the reverse conversion from soil contamination levels to the air dose rates, the conversion formula is derived by assuming a flat ground surface (Jacob et al., 1994; Saito and Jacob, 1998); however, topography in forested areas is often not flat. This study examines the relationships among the air dose rates under the canopies, topography, and the soil contamination levels on forested hillslopes in Fukushima, Japan. Gamma-ray air dose rates were measured using a hand-held scintillator during the summer of 2018 (two days), and in the winter of 2019 using KURAMA, a portable scintillator carried in a backpack (one day). This study employed numerical methods, including semivariogram and R's CAR (Companion to Applied Regression) package, to find aerial gamma-ray activities' spatial structure and the optimal combinations of topographic predictors. The survey routes and soil sampling points did not coincide completely. Thus, spatial and temporal reorganizations and re-definitions of the spatial boundary to incorporate the air (boundaryless mass), topography (three-dimensional structure), and soil samples (point measurements) were needed to overcome data analysis challenges. Some preliminary results show that the median air dose rate on a summer day with stronger winds (wind speed 7.1 m/s) was closer to the one on a winter day than to the median on the other summer day with calm air. Distance dependency (semivariogram range) on the windy summer day and the winter day was 30 to 60 m. Distance dependency on a summer day with calm air was much longer, > 500 m. The aerial gamma-ray levels were not in a linear relationship with the elevation. Meanwhile, the areas under evergreen trees showed higher air dose rates (3-8 %) than the areas under deciduous trees in absolute measurements. However, the differences were not statistically significant. The combination of slope degrees, hillslope aspects, and curvature or upslope distance best described the air dose rates, depending on the survey routes, although their predictabilities (R²) were low, 0.35 at the most. The air dose rates, which were estimated from soil samples' effective relaxation mass depths, did not correlate with the actual air dose measurements, but this result is under further investigation. This study's complete results will provide additional consideration points to the gamma-ray air dose vs. soil contamination assessment in the forested areas.