Prediction of topsoil stoniness using soil type information and airborne gamma-ray data
The stoniness of topsoil can have a significant impact on the cost-effectiveness and quality of work in mechanized forest operations. The operations and their models should be selected on a stand-specific basis, considering the physical properties of the soil, including stoniness, in order to achieve maximum efficiency and to minimize the damage caused by heavy forest machinery. The aim of this study was to examine whether the stoniness of the topsoil can be predicted using the gamma-ray values available from low-flying geophysical data and soil type information. Stoniness was measured at several sites with varying soil types, which were then divided into stoniness index classes (SIC) for further analysis by ordinal regression analysis using gamma-ray and soil type data. The SIC classification resulted in 52% accurate prediction performance and 79% acceptable prediction performance (-/+ 1 class from the correct class), with Kappa values of 0.55 and 0.72, respectively. The SIC prediction results were promising and showed the potential of gamma-ray and soil type data for the estimation of topsoil stoniness.