Abstract:

This study analyzed the damage to three forest roads in September 2015 and October 2019. Precipitation data from the nearest Automated Meteorological Data Acquisition System (AMeDAS) stations were analyzed as return periods of the annual maximum hourly and daily rainfall intensity, soil water index, and the storage in each tank in an in-line three-layer tank model, between 1979 and 2019. The first and second layers of the tank model related to the landslides at the Kanuma station near the Yokone forest road are approaching a 100-year return period in both 2015 and 2019, and the sediment from the landslides damaged the forest road. The Imaichi station near the Akaihara forest road had a large value of 75 years for the first and second layers of the tank model in 2015, but it was lower than the 50-year return period in 2019. However, a disaster occurred, as the culvert was blocked owing to an increase in the outflow rate of the sediment and logging residues by clear-cutting in the catchment area in 2019. At the Ikari station near the Nishizawa forest road, road collapse occurred in 2015 as the probability of the first and second layers exceeded 300 years. No disaster occurred in 2019, as the probability was about 10 years. Future research will review the 10-year probability of rainfall intensity, which is the standard for the forest road drainage facility design in Japan, considering the increases in runoff rates of the sediment and logging residues after clear-cutting.

Keywords: