Abstract
Sap flow velocity was measured by the heat-pulse method in a forested catchment (1.16 km2 in area), and transpiration from the catchment was estimated by three-dimensionally simulating actual sap flow blocked by sensors of definite size. The catchment is mostly forested by 56 % Abies sachalinensis (Sakhalin fir) and 44 % mixture with broad-leaved trees. Radial distributions of sap flow in the sapwood part for representative fir and broad-leaved trees were obtained by the heat-pulse method, and then transpiration Tr1 (L day-1) of each tree was calculated by integrating the sapwood area. The Tr1 values averaged for non-rainfall days proved to be highly correlated to trees’ breast high diameters DBH (cm) with R2=0.93 for Sakhalin fir and R2= 0.71 for broadleaved trees. The catchment transpiration averaged over non-rainfall days was 0.72 mm day-1 for the heat-pulse method and the catchment evapotranspiration, 2.03 mm day-1 for the Penman-Monteith method. The possible underestimate of transpiration from the heat-pulse method is probably due to the ignorance of sensors’ size originally supposed in the heat-pulse method. Actual sap flow blocked by sensors of 2.8 mm in diameter was three-dimensionally simulated. Giving the cut size of 2.8 mm by the sensors in the simulation, the catchment transpiration was estimated at 1.76 mm day-1. This value is likely reasonable to the evapotranspiration at 2.03 mm day-1 from the Penman-Monteith method, since the evapotranspiration includes the evapotranspiration from forest floor covered by striped bamboo (Sasa senanensis).
