This study proposed a quantitative model for analyzing the mass balance of large wood and its export on an annual scale at watershed scale. The study sites were five dam reservoir watersheds of the Kitakami River catchment located in the north-eastern part of Japan. From analyses on 20 years patterns of actual large wood export to target dam reservoirs, 1) the annual 24-hour maximum rainfall would be a key factor for the annual large wood recruitment, and 2) large wood export characteristics on an annual scale would be based on two relationships were found. Based on above findings, the model was consisted of two frameworks: the rainfall-induced analytical shallow landslide model for the large wood recruitment, and the double storage functions as with lumped hydrological method at a watershed scale for the large wood entrainment.

The model was used to re-analyze 20 years patterns of large wood export observed at the dam reservoir watersheds. The Nash-Sutcliffe values were 0.7 or more, which is indicated high reproductivity of simulation, for large wood export estimated using this model for four among five dam reservoir watersheds. Namely, our results could demonstrate that the landslides are a significant procedure of large wood export, and the characteristics of large wood export can be defined by two relationships, which are the direct export of large wood caused by an increase in large wood recruitment with the extreme rainfall event, and the baseflow of large wood, which is mainly old large wood recruitment stored in the stream.