节点文献
三峡库区坡地森林理水调洪功能研究与评价
Study and Evaluation on Water and Flood Regulation by Forests on Slope Land in the Three-Gorge Reservoir Area
【作者】 刘益军;
【作者基本信息】 北京林业大学 , 水土保持与荒漠化防治, 2003, 博士
【副题名】以缙云山为例
【摘要】 缙云山是三峡工程库区亚热带常绿阔叶林保存最完整的区域,是研究和评价森林理水调洪功能理想的天然实验室。始建于1999年的重庆缙云山森林生态效益定位监测站,拥有全自动森林气象、水文观测设施,可实时监测森林生态变化过程。本论文是该站2001年6月—2003年7月的观测研究结果,得到有效原始数据8,619,840个,其中大暴雨4日,暴雨8日,大南23日,中雨43日,小雨293日。 大气降水经过森林环境后表现为森林及其土壤对降水的蓄调过程,把森林的这种调节作用称为理水调洪功能。这种功能具有时空性、生长性、可变性、复杂性、耦合性和可预测性,各层次的贡献大小和方式随森林群落的生长发育及降雨性质的变化而变化。 水蚀预测(WEPP)模型是一个较好的研究和评价森林理水调洪功能的虚拟平台,本文为模型提供了一种气候数据转换方法。人工神经网络(ANN)模型通过观测(学习、记忆)森林理水调洪过程中出现的非线性现象,发现规律(训练),具有较好的预测和评价能力,可作为森林理水调洪功能研究与评价的认知模型。 缙云山的树冠在雾及微雨时常常能形成树雨而增加林内降雨,林冠截留量y(mm)与降雨量x(mm)呈乘幂关系: y=0.6047x0.9001 R2=0.8019 由林分郁闭度、林分平均胸径、林分平均高、林龄、林分平均枝下高、林分密度、冠幅、降雨量、雨强、风速、气温、空气湿度作为BP网络的输入量,林冠截留量作为输出量,构成12-6-1三层BP网络的权值依次为:0.988053、-0.0358021、0.406472、-0.0675435、0.225698、0.412862、-0.00698586、-1、0.652695、1、-1、-1、-0.154032;对林冠截留影响的实际权重依次为0.0718、0.0772、0.0837、0.1600、0.0210、0.0008、0.1849、0.2406、0.0428、0.0523、0.0325、0.0325。 毛竹干流在降雨量大于0.3mm就能观测到,并有固定的流路,平均干流率为26.3%,单株最大流量在19.8L/h以上。干流量与降雨量呈4阶非线性多项式关系(R2>0.9);与出竹年度的关系是Weibull(7.5,1991.9,4.8,1.4)。毛竹干流量在冠幅为1m2以下时随冠幅增大而增大,大于1m2以后则减少,到2.5m2以后就基本保持不变。由冠幅、胸径、树高、枝下高、竹龄、降雨量与干流量构成的6-6-1 BP网络权值依次为:1.91161、1.88262、1.99452、-3.83706、-3.37618、-1.98549、-2.1501;实际权重是.088,0.186,0.197,0.170,0.103,0.256。 混交林(马尾松×大头茶)一年中产流达300天以上,最小侵蚀性日降雨量为12.76mm,地表平均径流系数为0.044,林冠最大截持降雨量为6.9mm,林内起始降雨量为0—4.3mm,林内降雨最早始于林外0.3h,延后8h结束。摘要 阔价t林(大头茶林)一年中产流达犯0天以上,最小侵蚀性日降雨量为13.22mm,地表平均径流系数为0.017,林冠最大截持降雨量为7.8,林内起始降雨量为O一5.3mm,林内降雨最早始于林外0.5h,延后IOh结束。 毛竹林一年中产流达220天以上,最小侵蚀性日降雨量为n .4Omm,地表平均径流系数为0.036,林冠最大截持降雨量为4.smm,林内起始降雨量为0一3.3mm,林内降雨最早始于林外o.Zh,延后6h结束。 灌木林和坡耕地的最小侵蚀性日降雨量分别是12.76mm、10.76mm,平均径流系数分别是0.038、0028。 可用地表径流、壤中流、土壤入渗,以及冠层、枯落物、树干截留理水能力和森林及土壤水库对降雨径流的蓄调量等指标从不同的角度来评价森林的理水调洪能力。试验区坡地森林理水调洪功能强弱依次为:阔叶林、混交林、灌木林、毛竹林。
【Abstract】 Jinyun Mountain is an ideal natural laboratory for studying and evaluating on water and flood regulation by forest. It conserves the integrated subtropics evergreen broad-leaved forest types in the three-gorge reservoir area. In 1999, the forest ecology beneficial monitoring station of Chongqing Jinyun Mountain, starts to construct, provided with automatic forest hydrological monitor. It can tell the living process of forest ecological changes. In this dissertation, there are 8,619,840 effective firsthand monitoring data from June, 2001 to July, 2003, with 4 heavy rainstorm days, 8 rainstorm days, 23 heavy rain days, 43 middle intension rain days, and 293 rain days.When rain falls into forest environment, the forest and soil can be expressed by a reservoir, which regulates water. We call the action water and flood regulation by forest. This forest service has timeliness, extensity, and variability, complexity, coupling, and predictability, changes with its growth and rain temperament.The simulation by WEPP model is rather better on runoff changing with rainfall and the growth situation of vegetation in the field. One way is gived for translating climate data into WEPP. WEPP model acts as very efficient fictitious interfaces to study and evaluate on water and flood regulation by forest.When artifical neural net (ANN) model views ( study, recollection ) the nonlinearity appearance of water and flood regulation by forest, it remembers regular pattern (training ),and forecasts the forest circle function dissectionally.When there is fog or little rain, the crown can catch fog or little rain together, so in forest environment the rainfall by fog drip is more than the open air at Jinyun Mountain.The relationship between canopy interception storage Y (mm) and rainfall X (mm) at Jinyun Mountain isIf the inputs of BP model are crown closure, DBH, H, age, under branches height, density, crown breadth, rainfall, rain intensity, wind velocity, air temperature, humidity, and the output is canopy interception storage, the rights in order are: 0.988053, -0.0358021,0.406472, -0.0675435, 0.225698, 0.412862, -0.00698586, -1, 0.652695, 1, -1, -1, -0.154032,but the real weights in order are: 0.0718,0.0772,0.0837, 0.1600,0.0210,0.0008, 0.1849,0.2406,0.0428,0.0523,0.0325,0.0325.The stemflow of Mao bamboo can be observed when the rainfall is over 0.3mm, and it almost flows along the same route. The rate of stemflow to rainfall is 26.3%, and the largest stemflow is over 19.8L/h in a tree. The relationship between stemflow and rainfall is 4 stairs multiple nonlinearity (R2>0.9), and for the age of it is Weibull (7.5,1991.9,4.8,1.4). If the inputs of BP model are DBH, H, under branch height, age, and the output is stemflow, the rights in order are: 1.91161,1.88262,1.99452, -3.83706, -3.37618, -1.98549, -2.1501, and the real weights in order are:0.088,0.186,0.197, 0.17,0.103,0.256.Mixed forest (Pinus massoniana Lamb, x Gordonia acuminata H. T.Chang) reservoir throws away its water for 300 days in a year. In a day, the minimal erosion rainfall is 12.76mm, interception is 6.9mm,and the average rate of surface flow is 0.044.The trees rain when rainfall is 0-4.3mm,and start 0.3h earlier, 8.0h later than the open air.Broad leaved forest (Gordonia acuminata H. T.Chang) reservoir throws away its water for 320 days in a year. In a day, the minimal erosion rainfall is 13.22mm, interception is 7.8mm,and average rate of surface flow is 0.017. The trees rain when rainfall is 0 -5.3mm,and start 0.5h earlier, end 10.Oh later than the open air.Mao bamboo forest reservoir throws away its water for 220 days in a year. In a day, the minimal erosion rainfall is 11.40mm, interception is 4.8mm, and average rate of surface flow is 0.036. The trees rain when rainfall is 0-3.3mm,and start 0.2h earlier, end 6.0h later than the open air.In a day, minimal erosion rainfall of brush and slope field is 12.76mm and 10.76mm, and the average rate of surface flow is 0.038 and 0.028.The indexs of evaluation on water and flood regulatio
【Key words】 the three-gorge reservoir area; forest environment; water and flood regulation; hydrological process; Chongqing Jinyun Mountain;