The dynamics of rainfall interception by a seasonal temperate rainforest

Net canopy interception (I-net) during rainfall in an old-growth Douglas-fir-westem hemlock ecosystem was 22.8 and 25.0% of the gross rainfall (P-G) for 1999 and 2000, respectively. The average direct throughfall proportion (p) and canopy storage capacity (S) derived from high-temporal resolution throughfall measurements were 0.36 and 3.3 mm, respectively. Derived values of S were very sensitive to the estimated evaporation during canopy wetting (I-w). Evaporation during wetting was typically small due to low vapor pressure deficits that usually occur at the start of an event, therefore I-w is best estimated using the Penman method during canopy wetting, rather than assuming a constant evaporation rate over an entire event. S varied seasonally, from an average of 3.0 mm in the spring and fall, to 4.1 mm in the summer, coincident with canopy phenology changes. Interception losses during large storms that saturated the canopy accounted for 8 1% of I-net. Canopy drying after events comprised 47% of I-net, evaporation during rainfall comprised 33%, and evaporation during wetting accounted for 1%. Interception associated with small storms insufficient to saturate the canopy accounted for 19% of I-net. The Gash analytical model accurately estimated both Inet and the individual components of I-net in this system when applied on an event basis, and when the Penman method was used to compute evaporation during rainfall. The Gash model performed poorly when applied on a daily basis, due to a rainfall regime characterized by long-duration events, which violated the assumption of one rain event per day. (C) 2004 Elsevier B.V. All rights reserved.