北京山区树冠和枯落物结构对幼林水文防蚀功能动态影响

发布时间：2018-06-16 08:57

  本文选题：人工降雨 + 水文功能　； 参考：《北京林业大学》2016年博士论文

【摘要】：森林覆盖对降水输入和输出以及坡面土壤侵蚀有重要影响,但前人研究多集中于计算或预测截留量、径流量和侵蚀量等,对上述过程缺少准确细致地刻划,特别是对降雨过程中的幼林垂直结构与其水文和防蚀功能之间的动态响应关系,及相关模型构建模拟等缺少研究。为揭示上述问题,本文选取北京山区的典型树种(4—5年生侧柏、油松、元宝枫和栓皮栎)为研究对象,在五种人工模拟降雨雨强下(5.7—150mmh-,从单株和坡面径流小区(面积4.5×2.0m2和4.5×1.0m2)两个尺度开展了树冠截留、枯落物截留,以及二者的复合截留研究,同时分析了树冠和枯落物覆盖下坡面径流、截留、入渗及土壤侵蚀过程,结果表明：(1)树冠截留和枯落物截留均是动态过程,包括快速湿润、稳定饱和、以及滞后滴雨三阶段。树冠截留的平均最大截留量Cmax和最小截留量Cmin分别为0.66和0.40 mm,而枯落物的平均Cmax和Cmin分别为2.25和1.45 mm,这意味着降雨结束后分别有近40%和35%的截留降水从树冠和枯落物层滴落至地表。降雨强度对树冠Cmax和Cmm影响不甚明显,但显著影响枯落物Gmax(p0.05)。而树冠结构包括叶片特征(叶面积、叶面积指数LAI、叶片数、叶片生物量、最小叶片密度、相邻叶片距离等)及枝干特征(枝干面积、枝干生物量、枝干数、最小枝干密度、枝干长)均对截留过程及Cmax和Cmin有显著影响(p0.05)。针叶树冠Gmax是阔叶的1.9倍,而阔叶枯落物cmax和Cmin则高于针叶,前者分别是后者的1.5和1.6倍。单独的截留研究表明枯落物Cmax和Cmin分别是幼树树冠的3.4和3.6倍,而复合结构截留试验下枯落物截留能力更强,其Cmax和Gmin分别是树冠的6.4和5.8倍。而树冠和枯落物联合Cmax和Gmin占降雨量的百分比则分别为18.6%和9.5%。(2)研究分别构建了树冠和枯落物截留过程模型,即基于LAI和累计降水量Pc的雨中树冠动态截留模型和雨后滴雨模型PRDGl=1/kIn(D1+I)+cLAId,以及基于枯落物单位面积质量M和累计降水量Pc的雨中枯落物动态截留模型和雨后滴雨模型PRD：Cl=1/kIn(D1+I)+aMb,模型模拟截留效果良好,MRE介于10.7%—23.3%之间。(3)研究发现树冠和枯落物覆盖下可以有效削减地表径流,促进径流入渗。与裸地小区相比,有林小区和枯落物覆盖小区均可以推迟坡面出现积水时间和产流时间5—10分钟,产流速率分别是裸地的73%和85%,入渗速率分别是裸地的1.5和2.3倍,总产流量分别是裸地的82%和63%,总入渗量分别是裸地的1.5和1.6倍。降雨强度显著影响裸地小区、以及有林和枯落物覆盖小区径流过程,当雨强从5.7增至75.6mmh-1时,各小区总径流量增加33—83倍。对有林小区,油松1.0m×1.0m小区因有最大的叶面积指数LAI和覆盖度而总产流量最少。对枯落物覆盖小区,其总产流量Qg可能受控于质量M和雨强RI,三者之间存在关系Qg=-7.24M+0.77RI(R=0.97)。(4)将CIDR模型基于结构参数从单株推广至坡面尺度后,发现树冠和枯落物覆盖下坡面小区内发生的截留、径流、入渗均是动态过程,动态水量平衡的演化规律可归纳为在5.7和11.7 mm降雨下,入渗过程是各小区内的主导过程；而在49.8和75.6mm降雨下,各小区内的水文过程则是由径流主导；25.2mm降水下则是由前30—45分钟的入渗主导过渡至后15—30分钟的入渗主导。就总水量平衡而言,有林小区总入渗量占降水量的47.0%,而总径流量占比为49.6%,总截留量仅为3.4%；而在枯落物覆盖小区中总径流量则为45.3%,总入渗量44.1%,总截留量为10.6%。(5)研究构建并验证了径流和入渗过程模型。在有林小区,提出了基于LAI和累计降水量Pc的径流过程模型RDP:Qt=(-0.09 LAI+0.55)Pc(0.04LAI+1.13),以小雨入渗过程模型LIP:It=(-0.11LAI+1.04)Pc(0.11LAI+0.74和大雨入渗过程模型HIP：It=(0.44LAI+1.25)PcLRI(0.17LAI-0.02)。类似地,在枯落物小区基于枯落物类型及其质量M,以及累计降水量Pc,构建并验证了径流过程模型：Qt=(-0.39M+0.61)Pc(0.15M+1.08)(栓皮栎),Qt=(-0.43 M+0.64)Pc(0.17M+1.05(油松),以及入渗过程模型IDP:It=(1.44 M2-1.60 M+1.08)PcM(-1.05M+1.34)+0.40(栓皮栎),It=(-0.96 M2+1.94M)pcM(-2.37M+2.72)(油松),模型可较为精确地模拟径流和入渗过程,MRE介于13—27%之间。(6)树冠和枯落物覆盖下的坡面侵蚀产沙过程主要包括产沙速率和含沙量等的快速增加阶段和之后的稳定波动阶段,有林和枯落物小区可以有效控制土壤侵蚀,其产沙速率分别为裸地的63.3%和16.4%,总产沙量分别为裸地的58.8%和16.8%。降雨强度显著影响产沙过程,当雨强从5.7增至75.6 mm h-1时,总产沙量增加近30—77倍,而树冠覆盖度C同样对产沙有一定影响,产沙率SLR与之存在如下关系：SLR=e-0.02C,而总产沙量S则可能受控于覆盖度C和降雨强度RI:S=-0.15 C+0.31 RI.类似地,枯落物质量M增加可减少总产沙量约80%,其与雨强对总产沙量的联合影响可概括为：S=-0.46 M+0.02尺,。综合看来,油松1.0 m×1.0 m有林小区和栓皮栎枯落物覆盖小区防蚀功能最为显著。另外,分别建立并验证了侵蚀过程模型,模型可较为精确地模拟产沙过程,有林地覆盖坡面St=(0.04LAI2-0.21 LAI+0.60)QrLAI016LA1+081),枯落物覆盖坡面：栓皮栎和裸地：St=(-0.18 M+0.18)Qt(-0.34M+0.90),油松：St=(-0.02M+0.05)Qt(b=-0.70M+1.32),模型整体模拟效果较好,MRE介于28%-35%之间。WEPP模型(坡面版)可以较好地模拟坡面径流产沙过程,在模拟径流过程(累计径流量)的基础上,模型对总径流量的预测效果(CE0.85)要好于侵蚀产沙量(CE0.50)。上述结论量化归纳了树冠和枯落物结构对其水文和防蚀功能的动态影响过程,通过含有结构参数的动态过程模型可初步实现对上述功能的定向调控,也可为防护林营造过程中的树种选择、空间配置以及结构优化等提供一定的理论和数据支持。
[Abstract]:Forest cover has an important impact on the input and output of precipitation and soil erosion on the slope. However, the previous studies mainly focus on the calculation or prediction of interception, runoff and erosion, and the lack of accurate and meticulous characterization of the process, especially the dynamic response relationship between the vertical structure of the young forest and its hydrological and corrosion protection functions during the rainfall process. In order to reveal the above problems, the typical tree species (4 to 5 year old Platycladus orientalis, Pinus tabulaeformis, Acer truncatum and Quercus variabilis) in Beijing mountain area were selected as the research objects. Under five artificial simulated rainfall intensity (5.7 - 150mmh-), two scales were carried out from single plant and slope runoff plot (area 4.5 * 2.0m2 and 4.5 x 1.0m2). The canopy interception, the litter interception and the compound interception of the two were studied. At the same time, the runoff, interception, infiltration and soil erosion of the canopy and litter cover were analyzed. The results showed that (1) the canopy interception and the litter interception were all dynamic processes, including fast wetting, stable saturation, and lagging drop rain three stages. Tree crown interception The average maximum interception Cmax and the minimum interception Cmin were 0.66 and 0.40 mm respectively, while the average Cmax and Cmin of the litter were 2.25 and 1.45 mm respectively, which meant that nearly 40% and 35% of the precipitation dropped from the canopy and the litter layer to the surface respectively after the end of the rainfall. The rainfall intensity had no significant influence on the crown Cmax and Cmm, but the effect of rainfall was significant. The canopy structure including leaf area, leaf area index (leaf area index LAI, leaf number, leaf biomass, minimum leaf density, adjacent leaf distance, etc.) and branch dry characteristics (branches area, branch dry biomass, branch dry density, minimum branch dry density, stem length) have significant influence on interception and Cmax and Cmin (P0.05). The crown of Gmax. The Gmax is 1.9 times that of broad-leaved, while the broadleaf litter Cmax and Cmin are higher than that of the needles, and the former is 1.5 and 1.6 times the latter respectively. The single interception study shows that the litter Cmax and Cmin are 3.4 and 3.6 times the crown of the young tree respectively, and the litter interception ability is stronger under the compound structure interception test, and the Cmax and Gmin are 6.4 and 5.8 times the tree crown respectively. The percentage of Cmax and Gmin in the combination of crown and litter is 18.6% and 9.5%. (2) respectively. The model of canopy and litter interception is constructed respectively, that is, the canopy dynamic interception model based on LAI and cumulative precipitation Pc and PRDGl=1/kIn (D1+I) +cLAId after rain rain model, as well as the mass M and fatigue based on the unit area of the litter. The dynamic interception model of rain in the rain Pc and PRD:Cl=1/kIn (D1+I) +aMb after rain drop model, the model simulated interception effect is good, MRE is between 10.7% and 23.3%. (3) the study found that canopy and litter cover can effectively reduce surface runoff and promote runoff infiltration. Compared with the bare land area, there are forest plots and litter coverage. The cover plot can postpone the water accumulation time and the flow time of 5 to 10 minutes, and the rate of runoff is 73% and 85% in bare land, respectively, the infiltration rate is 1.5 and 2.3 times of the bare land respectively, the total yield is 82% and 63% of the bare land respectively, the total infiltration is 1.5 and 1.6 times of the bare land, respectively. The rainfall intensity has a significant influence on the bare land area and the Lin Heku The total runoff is 33 to 83 times when the rainfall intensity increases from 5.7 to 75.6mmh-1, and the total runoff is 33 to 83 times more than that of the residential district. For the forest area, the total yield of the pine 1.0m x 1.0m plot is least because of the maximum leaf area index LAI and coverage. For the litter area, the total runoff yield Qg may be controlled by the mass M and the rain strong RI. In relation Qg=-7.24M+0.77RI (R=0.97). (4) after extending the CIDR model to the slope scale based on the structural parameters, it was found that the canopy and litter covered the slope in the subplot. The runoff and infiltration were all dynamic processes. The evolution law of the dynamic water balance could be attributed to the rainfall of 5.7 and 11.7 mm, and the infiltration process was within the various communities. Under the rainfall of 49.8 and 75.6mm, the hydrological process in each area is dominated by runoff, while the 25.2mm precipitation is dominated by the first 30 to 45 minute infiltration leading to the following 15 to 30 minutes. In terms of total water balance, the total infiltration of the forest area accounts for 47% of the precipitation, and the total runoff ratio is 49.6%. The amount of retention was only 3.4%, while the total runoff in the litter covered area was 45.3%, the total infiltration was 44.1%, the total interception was 10.6%. (5). The runoff and infiltration process model was constructed and verified. In the forest area, the runoff process model RDP:Qt= (-0.09 LAI+0.55) Pc (0.04LAI+1.13) based on the accumulated precipitation Pc was proposed, and the rainfall infiltration was carried out. The process model LIP:It= (-0.11LAI+1.04) Pc (0.11LAI+0.74 and heavy rain infiltration process model HIP:It= (0.44LAI+1.25) PcLRI (0.17LAI-0.02). Similarly, the litter area based on the litter type and its mass M, and the cumulative precipitation Pc, constructed and verified the runoff process model: Qt= (-0.39M+0.61). 0.64) Pc (0.17M+1.05) (You Song), and the infiltration process model IDP:It= (1.44 M2-1.60 M+1.08) PcM (-1.05M+1.34) +0.40 (Quercus variabilis) and It= (-0.96 M2+1.94M) pcM (You Song), the model can accurately simulate runoff and infiltration process, which is between 13 and 27%. (6) the main package of slope erosion and sediment yield under the cover of tree crown and litter. In the rapid increase stage of sediment yield and sediment concentration, the soil erosion can be effectively controlled by the forest and the litter area, and the rate of sediment yield is 63.3% and 16.4% in bare land, respectively, the total sediment yield is 58.8% in bare land and the 16.8%. rainfall intensity is significantly affected by the precipitation process, when the rain intensity increases from 5.7 to 75.6 mm H-1 The total sediment yield increased by nearly 30 to 77 times, and the canopy coverage C also had a certain effect on the sediment yield. The sediment yield SLR was related to it as follows: SLR=e-0.02C, while the total sediment yield S may be controlled by the coverage C and the rainfall intensity RI:S=-0.15 C+0.31 RI., and the increase of the litter mass M can reduce the total sediment yield by about 80%. The combined effect of sand quantity can be summarized as follows: S=-0.46 M+0.02 ruler. In comprehensive view, the 1 m x 1 m of Pinus tabulaeformis has the most significant anti-corrosion function in the forest area and the cork oak litter area. Furthermore, the erosion process model is established and verified, and the model can be more accurate to simulate the sand production process and the St= (0.04LAI2-0.21 LAI+0.60) Qr of the woodland covered slope (0.04LAI2-0.21 LAI+0.60) Qr. LAI016LA1+081), litter covered slope: Quercus variabilis and bare land: St= (-0.18 M+0.18) Qt (-0.34M+0.90), Pinus tabulaeformis: St= (-0.02M+0.05) Qt (b=-0.70M+1.32), the model overall simulation effect is better, MRE between 28%-35% between the model (slope plate) can better simulate the process of slope runoff and sediment, in the simulated runoff process (cumulative runoff). On the basis of the model, the prediction effect of the total runoff (CE0.85) is better than that of the erosion sediment yield (CE0.50). The conclusion is that the dynamic process of the canopy and litter structure on its hydrological and corrosion protection functions is quantified, and the directional control of the above functions can be realized by the dynamic process model containing the structural parameters, and it can also be a protective forest camp. Tree species selection, spatial allocation and structural optimization in the process of construction will provide some theoretical and data support.
【学位授予单位】：北京林业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S715
，

本文编号：2026126