Gash模型在黄土区人工刺槐林冠降雨截留研究中的应用

Modeling canopy rainfall interception of a replanted Robinia pseudoacacia forest in the Loess Plateau

为了验证Gash林冠截留解析模型在黄土高原人工林中的适用性,基于2009年黄土丘陵沟壑区吕梁市王家沟小流域刺槐林样地降雨观测数据,采用Gash模型对林冠截留进行了模拟。所选刺槐林分为人工纯林,林龄约30a,阴坡,坡度24°,密度为990株/hm2,平均树高10.8 m,平均胸径12.4 cm,郁闭度0.76。根据回归方法确定了Gash模型中的主要参数,包括饱和林冠的平均蒸发速率(E珔)、林冠枝叶部分的持水能力(S)、自由穿透降雨系数(p)、树干持水能力(St)和树干茎流系数(Pt)。结果显示,2009年5月至10月人工刺槐林样地实测降雨量为366.9 mm,穿透降雨量为317.5 mm,树干茎流为10.2 mm,林冠截留量为39.2 mm。模型模拟的林冠截留量为42.4 mm,高于实测值3.2 mm,相对误差为8.2%。敏感性分析表明,S、E珔、St和pt每增加10%,林冠截留量分别增加4.7%,3.1%,1.7%和0.5%;p增加10%,林冠截留量则减少2.6%。说明树干持水能力(St)和树干茎流系数(pt)两个参数对黄土高原人工刺槐林冠截留量的预测值影响程度较小。模拟值与实测值有较好的一致性,显示Gash模型适用于黄土高原人工刺槐林冠的截留计算。

Scientific understanding about the effects of afforestation on the water balance in catchment areas has been significantly enhanced by studies on the interception and evaporation of rainfall. In order to reduce severe soil erosion on the Loess Plateau, a policy of reforestation has changed land use from cropland to planted forests. However, this has been found to lead to a reduction in water yield, partly because of an increase in loss through rainfall interception by the forest canopy, which is of critical importance to the water budgets of forests. The Gash analytical model has been widely used throughout the world for estimating rainfall interception by the forest canopy. In order to verify its applicability to the Loess Plateau, we compared the modeling results of the Gash model with those of field data collected from a monocultnre Robinia pseudoacacia forest stand in 2009. The studied Robinia pseudoacacia were located on the north-facing slope （24°） in the Wangjiagou Watershed, which covers an area of 9.1 km^2, near the city of Lvliang in the north of Shanxi province, China （N 37°33＇, E110°09＇）. The age of the Robinia pseudoacacia forest was 30 years at the time of the study. The mean tree height was 10.8 m, the mean stem diameter at breast height was 12.4 cm, and the planting density was about 990 trees/ hm^2 with the coverage being 0.76. We studied the principal components of rainfall interception loss and estimated forest structure parameters, including the mean evaporation rate （ E ）, the canopy storage capacity at saturation （ S ）, the free throughfall coefficient （ p ）, the rainfall fraction diverted to the trunks （ Pt ）, and the trunk storage capacity （ St ）, by using the intercepts and slopes obtained from regression analyses of the measured interception loss, throughfall, and sternflow versus gross rainfall. The results showed that the total estimated interception loss during the period of observation was 8.2% higher than that calculated on the basis of measurements of the gross rainfall, throughfall, and stemflow. Compared with the previous studies in other ecosystems, the performance of the Gash analytical model in the Loess Plateau region was reasonable. Sensitivity analysis indicated that the canopy interception by the Gash model was affected most by the canopy storage capacity at saturation （ S ）, the mean evaporation rate during rainfall （ E ）, and the free throughfall coefficient （ p ）, while interception was neither sensitive to trunk storage capacity （ St ） nor to the rainfall fraction diverted to the trunks （ Pt ） since these two parameters only had a small effect on the total interception. The good agreement between the estimated and measured values indicated that Gash＇s analytical model is suitable for estimating interception losses in forests on the Loess Plateau of China. This study will contribute new information about the applicability of the original Gash analytical model to a new geographic location. It is predictable that, in the future, the area of planted forests will increase rapidly due to the government policy of converting cropland to forest on the Loess Plateau, and our study will provide forest plantation managers with valuable information for estimating the interception losses and assessing the hydrological impacts of land-use modifications.