本研究為探討降雨經過樹冠層後的再分佈(redistribution)現象，蒐集臺灣大學實驗林溪頭 營林區第3林班柳杉(Cryptomeria japonica )人工林樣區的幹流量、穿落水量，配合溪頭通量測站的 降雨量，應用水收支法估算樹冠截留量。觀測期間自2012年2月至2014年2月，於樣區每間隔約15 日蒐集幹流量、穿落水量，共蒐集45筆資料。經刪除估算之樹冠截留量小於0的資料後，統計降雨 量為1786.5 mm，穿落水、幹流及樹冠截留量分別為降雨量的88.4%、2.5%及9.1%。試驗地之柳杉 為63年生，立木密度為721株/ha，冠層葉面積指數僅2.66，因此樹冠截留量較低。其次，間隔約15 日的累積降雨量高於100 mm情況下，穿落水、幹流、樹冠截留百分比平均分別為90.1%、3.1%、 6.8%；低於100 mm情況下，則分別為85.1%、1.6%、13.3%，顯示穿落水量及幹流量受降雨量多寡 的影響。另外，穿落水量及樹冠截留量與降雨量呈顯著相關，當累積降雨量高於200 mm時穿落水 百分比趨近於100%；樹冠截留百分比會隨著降雨量增加而降低，當累積降雨量大於150 mm時截留 百分比趨近於0%。

The purpose of this study was investigating the phenomenon of rainfall redistribution through forest canopy. Throughfall and stemflow were measured in a Japanese cedar (Cryptomeria japonica ) plantation located at the third compartment of the Xitou tract of the NTU Experimental Forest. The collected data, in conjunction with the rainfall data from the Xitou flux station, were used to calculate canopy interception by the water budget method. Study period was from February 2012 to February 2014 with throughfall and stemflow measured every 15 days; thus, resulting in a total of 45 sets of data. After removing data with negative canopy interception, the total rainfall amounted to 1786.5 mm with throughfall, stemflow and canopy interception accounting for 88.4%, 2.5% and 9.1%, respectively. The Japanese cedar plantation was 63 years old with stem density of 721 trees/ha and canopy leaf area index of 2.66; therefore, the percentage of canopy interception was expectedly low. Secondly, when total rainfall at every 15-day interval was greater than 100 mm, the average percentage of throughfall, stemflow and canopy interception were 90.1%, 3.1% and 6.8%, respectively. However, when it was less than 100 mm, throughfall, stemflow and canopy interception were 85.1%, 1.6% and 13.3%, respectively. This indicates that throughfall and stemflow were influenced by rainfall. In addition, throughfall and canopy interception were significantly correlated with rainfall. The percentage of throughfall was nearly 100% when accumulated rainfall was greater than 200 mm. The percentage of canopy interception decreased as rainfall increased, whereby it was nearly 0% when accumulated rainfall was greater than 150 mm.