Abstract
Colluvial soil or sediment from cut slopes often stops traffic on forest roads and it is troublesome to remove it too often. The objectives of this study are to quantify such soil dropped by rainfall or by freezing and fusion and to forecast seasonal and annual variation of it based on the results of measurement. In this study, we made three research plots on cut slopes whose slopes were 63.4, 55.0 and 48.0 degrees and observed the volume of colluvial soil or sediment. As a result, the soil volume of 48.0-degree slope decreased year by year while the others did not. In the fourth year, the soil volume of 48.0-degree slope reached 127kg while that of 63.4-degree slope reached 476kg, which was about 3.8 times as much as that of 48.0-degree slope. We used the analysis of variance to analyze relationships between the soil volume and the independent variables such as frequency of freezing and fusion, amount of rainfall and slope angle. As a result, all the independent variables were found to be significant at the 1% level. It should be noted that the contributing portion of frequency of freezing and fusion was 25.8% and was by far the highest among the three factors. Time series data on the volume of colluvial soil was separated into three portions such as TC (Trend and Cyclical component), S (Seasonal variation) and I (Irregular variation) using the EPA method (Economic Planning Agency method). The model to forecast the future volume of colluvial soil on a cut slope was subsequently made using TC and S components. The results showed that the model forecast seasonal and annual variation of such soil with high accuracy.
