基于Hydro-DNDC模型的紫色土典型小流域氮素损失途径与通量分析 | |
Alternative Title | Modeling of Nitrogen Loss Pathways and Fluxes in a Small Watershed of Purple Soil Based on Hydro-DNDC Model |
胡磊 | |
Subtype | 硕士 |
Thesis Advisor | 朱波 |
2017 | |
Degree Grantor | 中国科学院大学 |
Place of Conferral | 北京 |
Degree Discipline | 环境工程 |
Keyword | 紫色土 小流域 Hydro-DNDC模型 氮流失 含氮气体排放 |
Abstract | 定量描述小流域氮素迁移与转化是保证农业生态系统持续发展并协调环境友好的基础。由于小流域代表了一个典型区域的社会经济与生态环境特征，是开展区域环境与经济活动评估的基本单元，且流域内地形、土地利用和社会经济活动等较为复杂，因此在传统的点位观测基础上，开展典型小流域氮循环的模拟研究，是未来氮素循环与管理研究的热点。本文选择经典的生物地球化学模型（DNDC）与地表径流方程（SCS）和修改侵蚀模型（MUSLE）耦合而建立的生物地球化学-水土流失耦合模型（Hydro-DNDC）对四川盆地典型紫色土小流域—万安小流域开展氮素损失途径与通量的模拟，分析氮素随地表径流和壤中流的流失及氮素气体的排放过程与通量的空间特征及其尺度差异。主要研究结果和结论如下： （1）通过对2015~2016年完整的“小麦-玉米”轮作周期内土壤N2O排放、地表径流和壤中流的观测得到了典型坡地农田几种方式下氮素损失的量。在整个轮作周期内土壤N2O排放总量为1.16±0.35 kgN·hm?2，排放系数为0.26%；通过地表径流和壤中流流失的流失的铵态氮总量为0.44kg N·hm?2，流失系数为0.16%，硝态氮总量为26.77 kg N·hm?2，流失系数为9.56%。从观测的已知氮素不同途径损失量来看，硝态氮淋溶损失是四川盆地丘陵区典型坡地农田最大的流失途径。（2）为确定模型模拟地块尺度氮素损失途径各通量的参数及其模拟的精度，利用小流域内部典型农田（地块）尺度2013~2016年“小麦-玉米”轮作系统的观测数据对模型模拟的产量、N2O排放和氮素径流损失分别进行参数率定和模型精度评价。结果表明：具备了准确模拟产量的能力；模型能够较准确捕捉到施肥初期和大降雨过程中土壤N2O的排放峰值，能够较准确模拟N2O的年迹累积排放量，但是对于实际低排放通量的观测模拟结果还不是很理想；Hydro-DNDC能准确的模拟紫色土典型坡地农田地表径流的季节变化趋势（R2=0.99，P<0.01，n=10；ME=0.85），Hydro-DNDC能准确的模拟紫色土典型坡地农田壤中流的季节变化趋势（R2=0.92，P<0.01，n=18；ME=0.86），土壤侵蚀的季节变化特征（R2=0.99，P<0.01，n=10；ME=0.82），总氮N流失的季节变化特征（R2=0.98，P<0.01，n=10；ME=0.87）和硝态N淋溶的季节变化趋势（R2=0.98，P<0.01，n=18；ME=0.79）。模型的敏感性分析结果显示：降雨量是典型坡地农田径流、泥沙迁移和氮素流失最为敏感的环境因子；氮肥施用量的多少对硝态氮淋溶影响较大，但对总氮流失影响较小；坡度大小显著影响泥沙迁移过程，从而对总氮流失的影响较大。（3）通过2013~2014年截流小流域出口月径流量、泥沙侵蚀量和总氮负荷分别检验Hydro-DNDC模型的模拟能力，结果表明：月径流量模拟结果与实测值的相关性和精度较高（R2=0.91，P<0.01，n=24；ME=0.87），年平均径流的模拟偏差为11.34%；月泥沙侵蚀模拟结果与实测值的相关分析和精度较高（R2=0.92，P<0.01，n=24；ME=0.89），年平均泥沙侵蚀量的模拟偏差为8.90%；模型能够捕捉到雨季所引起的N素流失峰值（R2=0.89，P<0.01，n=24；ME=0.79），月平均总N流失的模拟偏差为10.3%。验证结果表明Hydro-DNDC能够准确模拟小流域中氮素流失空间上的“热点”区域，并捕捉时间上的“热点”时段。（4）基于分布式输入数据库的输入，模型对2016年万安小流域模拟的结果表明：不同土地利用方式下，由于坡度和土壤属性的不同造成不同氮素损失途径的强度差异明显；在空间上，万安小流域全年施用氮肥量约为158679kg，万安小流域全年因径流和泥沙迁移引起的总N流失总量为11366 kg N，其中约64.3%来源于农田；模型模拟的N淋溶总量为：20238 kg N，其中约94.4%来源于农田；模型模拟的含N气体排放总量为18177 kg N，其中约98.9%来源于农田；含氮气体中NH3挥发为排放的主要形式，排放量为17014kg N，占所有含氮气体的93.6%。时间上，万安小流域总氮流失和氮素淋溶损失主要集中在5~9月，其中流失量与淋溶量分别占年总量的80.5%和75.9%，降雨量是导致总氮淋失和氮素淋溶的最主要原因。 |
Other Abstract | Quantitative description of the migration and transformation of nitrogen in small watershed is the basis for ensuring the sustainable development of agroecosystem. Due to the large spatial scale and the complexity of the terrain and land use types of the small watershed, the traditional point observation can not satisfy the study of nitrogen cycle in small watershed. In this paper, the coupling model of DNDC and surface runoff (SCS) and sediment erosion (MUSLE) was used to simulate the nitrogen loss path and flux in the typical small watershed - Wanan in purple soil area. Simulated nitrogen loss with surface runoff, soil flow and nitrogen gas emissions. The main findings and conclusions are as follows:(1) The N2O emission, surface runoff and soil leach in the "wheat - maize" rotation cycle from 2015 to 2016 were used to obtain the amount of nitrogen loss in several ways. The total N2O emission in the whole rotation period was 1.16 ± 0.35 kg N ? hm-2 and the emission coefficient was 0.26%. The total amount of ammonium nitrogen dissolved by surface runoff and soil loss was 0.44 kg N ? hm-2 and the loss coefficient was 0.16%. The total amount of nitrate nitrogen was 26.77 kg N ? hm-2, and the loss coefficient was 9.56%. From the observed loss of different ways of nitrogen loss, nitrate leaching loss is the largest loss path of typical farmland in hilly area of Sichuan Basin.(2) In order to determine the simulation accuracy of the simulation of the nitrogen loss path in the typical farmland, the observation data of the "wheat-maize" rotation system (from 2013 to 2016) of the typical farmland (land) scale in the small watershed were simulated by model including the N2O emission and nitrogen runoff loss were evaluated by parameter rate and model accuracy. The results show that the model can accurately simulate the yield. The model can accurately capture the emission peak of N2O in the early stage of fertilization and rainfall, and can accurately simulate the cumulative emission of N2O, but for the actual low emission flux. Hydro-DNDC can accurately simulate the seasonal variation of surface runoff (R2=0.99, P<0.01, n=10; ME = 0.85), and Hydro-DNDC can accurately simulate the seasonal variation of soil flow (R2=0.92, P<0.01, n=18; ME=0.86), the seasonal variation of soil erosion (R2=0.99, P<0.01, n=10; ME=0.82) , the seasonal variation of total N runoff (R2=0.98, P<0.01, n=10; ME=0.87) and the seasonal variation of nitrate N leaching (R2=0.98, P<0.01, n=18; ME=0.79). The results of the sensitivity analysis showed that the rainfall was the most sensitive environmental factor in the runoff, sediment migration and nitrogen loss at typical slope farmland. The effect of nitrogen application on the leaching of NO3--N was larger, but there was little effect on total nitrogen loss and the slope significantly affected the sediment migration process, which had a great effect on the total nitrogen loss.(3) The simulation capacity of Hydro-DNDC model was tested by monthly runoff, sediment erosion and total nitrogen load from 2013 to 2014 of the JieLiu small watershed. The results show that the correlation and accuracy of the monthly flow between the simulated results and the measured values are match better (R2=0.90, P<0.01, n=24; ME=0.89) and the simulated variation of annual mean runoff was 25.3%. The simulated monthly sediment erosion values were fit the observations accurately (R2=0.91, P<0.01, n=24; ME=0.88) and the simulated deviation of average annual sediment erosion is 16.01%. The Hydro-DNDC model was able to capture the peak N loss caused by the rainy season (R2=0.89, P<0.01, n=24; ME=0.79) and the simulated deviation of the mean monthly total N loss is 8.1%. The results show that Hydro-DNDC can accurately simulate the "hot spot" area in the nitrogen loss space of the small watershed and capture the "hot spot" time.(4) Based on the input of the distributed input database, the model shows that under different land use patterns, the difference of the intensity of nitrogen loss is obvious due to the different slope and soil properties. In space, the total amount of nitrogen loss in the Wan’an small watershed is about 158679kg, and the total amount of N loss caused by runoff and sediment erosion in the Wan’an small watershed is 11366 kg N, of which about 64.3% is from farmland. The total amount is 20238 kg N, of which about 94.4% is from farmland. The total simulated N-containing gas emission is 18177 kg N, of which about 98.9% is from farmland. NH3 volatile is the main form of nitrogen gas emission, the emissions were 17014 kg N, accounting for 93.6% of all nitrogenous gases. In the time, the loss of total nitrogen and the leaching loss of nitrogen in Wan’an small watershed were mainly concentrated from May to September, which taking up 80.5% and 75.9% percent of total N runoff and leaching nitrate nitrogen, respectively. The rainfall was the most important reason of the total N runoff and nitrogen leaching. |
Pages | 82 |
Language | 中文 |
Document Type | 学位论文 |
Identifier | http://ir.imde.ac.cn/handle/131551/24614 |
Collection | 山地表生过程与生态调控重点实验室 |
Affiliation | 中国科学院成都山地灾害与环境研究所 |
First Author Affilication | 中国科学院水利部成都山地灾害与环境研究所 |
Recommended Citation GB/T 7714 | 胡磊. 基于Hydro-DNDC模型的紫色土典型小流域氮素损失途径与通量分析[D]. 北京. 中国科学院大学,2017. |
Files in This Item: | ||||||
File Name/Size | DocType | Version | Access | License | ||
基于Hydro-DNDC模型的紫色土典型（2719KB） | 学位论文 | 开放获取 | CC BY-NC-SA | View Application Full Text |
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment