The forest industry in Finland faces challenges in maintaining a year-round supply of fresh roundwood due to low bearing capacity in forest terrain and roads during certain seasons. The accessibility of forest terrain depends on the physical properties of the soil, which in turn impacts the cost-effectiveness and quality of mechanized forest operations. However, the resolution of currently available trafficability data is insufficient for accurate stand-specific operational planning. Similarly, only a limited number of forest roads can support heavy-duty vehicles year-round, and manual assessments of road conditions are time-consuming. Thus, improved methods utilizing readily available data for route planning and targeted road maintenance are required. The objective of this thesis is to explore methodologies for predicting terrain mobility and forest road bearing capacity, particularly using gamma-ray spectrometry.
The thesis comprises three sub-studies using soil and forest road data alongside gamma-ray datasets from Eastern Finland. The first and third studies focus on predicting small-scale soil property changes (stoniness, soil depth, peat depth) using airborne and ground-based gamma-ray measurements. The third study also examined the correlation between airborne and ground-based gamma-ray datasets. Ordinal regression and linear discriminant analysis were employed as analytical methods in these studies. The second study focuses on predicting forest road bearing capacity using easily measurable road and terrain properties, applying linear mixed-effects models.
The third study revealed a weak correlation between airborne and ground-based gamma-ray datasets. The first and third studies showed promising results in predicting soil properties using gamma-ray datasets, with up to 70% accuracy in stoniness and soil depth predictions. The second study showed moderate success in predicting forest road bearing capacity using easily measurable field data, achieving accuracy of 34% and RMSE% of 36%. Overall, the findings highlight the potential of gamma-ray spectrometry in enhancing the prediction of soil properties and forest road bearing capacity.