【作者】 孙美；

【导师】 冯绍元； 霍再林；

【作者基本信息】 中国农业大学 ， 农业水土工程， 2014， 博士

【摘要】 本文以华北平原典型集约化种植区(北京地区)为研究区域,通过田间试验数据的采取与收集,对根区水质模型(RZWQM2)进行了参数优化估计,分析了模型参数及输出响应的不确定性；采用率定的RZWQM2模型开展了冬小麦—夏玉米轮作系统中农田土壤水氮运移规律的定量研究；建立了基于土壤基本理化性状指标的大兴区土壤饱和导水率传递函数模型,从空间尺度上分析了大兴区农田土壤氮素含量及累积量的影响因子,以及氮素淋失的空间分布特征及其对农田管理措施、气象条件变化的响应。论文的主要研究内容与结论如下：(1)通过区域布点取样测定和分析,研究了大兴区农田土壤颗粒组成与硝态氮的空间变异特征,建立了土壤饱和导水率传递函数并应用于大兴区饱和导水率预测中；分析了不同种植类型、土壤质地、作物种植对大兴区不同深度土壤氮素含量及累积量空间分布特征的影响。结果表明,研究区大部分取样点硝态氮在土壤剖面分布上表现为随深度增加而降低；不同的种植类型对土壤中的硝态氮积累影响显著,露地菜田与设施菜田的硝态氮含量及累积量均较高,大兴区长子营、采育镇等果蔬密集种植区是硝态氮淋失污染风险的主要地区。(2)基于田间喷灌试验资料,采用PEST对RZWQM2模型进行率定及验证,不同处理模拟值与实测值吻合良好,与传统试错法的对比,该方法具有快速准确的优点,表明RZWQM2模型可以用于冬小麦—夏玉米轮作作物生长与土壤水氮迁移转化过程的模拟。(3)应用基于拉丁超立方抽样的普似然参数估计法(GLUE)对RZWQM2模型不确定性进行分析,提出了整合不同观测指标的带权重联合似然函数来确定模型输出响应变量的不确定性范围。结果表明,土壤剖面1m内氮素累积量对硝化反应速率参数、夏玉米苗期生长特性参数、冬小麦春化作用特性及光周期敏感特性参数较敏感；土壤饱和导水率、氮素的硝化、反硝化、水解速率参数对作物的产量起较关键作用；采用联合似然函数的方法计算得到的模型输出结果不确定性范围总体上可更好地包含各输出响应的观测值。(4)不同水氮组合对作物产量、根区氮素淋失的影响研究表明,当作物灌水量较小时,作物产量、根区硝态氮淋失量对施氮量并不敏感；硝态氮的淋洗主要发生在夏玉米生育期。高水高肥条件地下水面处土壤硝态氮淋失通量占氮肥施用量的比例较高；中水中肥条件下作物的水氮利用指标总体较高。长系列相同灌水条件不同施肥条件下,土壤中硝态氮运移时间大体一致,淋洗时间随土壤深度增加产生明显滞后；不同灌水条件下,灌水量越大淋洗至深层的时间越短。(5)将RZWQM2模型与ArcGIS的结合,对大兴区作物产量和土壤氮素淋失对农田管理措施及气象条件变化的响应进行了分布式模拟。结果表明,水氮量增加、温度升高、降雨增大,都可以不同程度上提高作物产量及土壤硝态氮淋失量；在空间尺度上,地下水面处硝态氮淋失量对不同影响因素的影响因土壤质地的差异表现不同,西部庞各庄等砂性地区地下水的污染风险较高。更多还原

【Abstract】 Based on the background of typical intensive plant area in the North China Plain, Daxing district of Beijing, the parameters of Root Zone Water quality Model Version2.0(RZWQM2) was estimated by the collected data of field experiment irrigated with sprinkler. The uncertainty of these model input parameters and model outputs were also analyzed. The calibrated RZWQM2model was used to quantify the relevance of the transport of water and nitrogen in agricultural field and the pollution load of nitrate leaching in groundwater. The pedo-transfer function for Daxing district was established based on the soil saturated conductivity and soil physical characteristics. The influence factor of nitrogen content and nitrogen accumulation of soil profile in Daxing district, and the response of spatial distribution of nitrogen leaching to varying agricultural management practices and meteorological conditions were analyzed on a spatial scale. The main objectives and conclusions are as follows:(1) Choosing the Daxing district in Beijing as the study area, the spatial variation of particle composition and soil nitrate content were studied based on the measurement and analysis of regional sampling points. Pedo-transfer function of Daxing district was established and applied to predict the saturated hydraulic conductivity other sampling points. The effects of different soil plant types, soil textures and crop plant on the spatial distribution characteristics of soil nitrogen content and accumulation within different soil depth in Daxing district. The results showed that, soil nitrate content of most sampling points was decreased with increasing depth of soil profile. The effect of different plant types on nitrate content was obvious. Outdoor and protected vegetable field contained more nitrate nitrogen in soil profile. The towns of Panggezhuang and Beizangcun where soil contained more sand, and the towns of Zhangziying and Caiyu where planted intensively fruits and vegetables were the main district susceptible to nitrate pollution.(2) Based on the data of field experiment, PEST was used to calibrated and validated RZWQM2model. The simulations using PEST optimation showed a good fit with observations, and was obtained in a more efficient and precise way. The results indicated that the RZWQM2model could be used in simulating the crop growth and transport process of water and nitrogen in the double-rotation crop system of waxy maize and winter maize.(3) The sensitivity and uncertainty of model parameters, the uncertainty of model outputs were analyzed using General Likelihood Uncertainty Estimation (GLUE) method which based on Latin Hypercube Sampling (LHS) method. The weighted combined likelihood function aggregated different observation variables was proposed to determine the uncertainty range of RZWQM2model outputs. The results showed that the nitrogen accumulation with1m depth of soil profile was sensitive to the parameters of nitrification rate coefficient, thermal time of emergence stage for maize, and vernalization and photoperiod coefficients for wheat. The parameters of soil saturated hydraulic conductivity, nitrification and denitrification and hydrolysis of rate coefficient played a key role to crop yield. The uncertainty ranges for different model outputs calculated by combined likelihood function included generally better the observations.(4) The research on the effect of different application rate of irrigation and nitrogen fertilizer on crop yield and nitrogen leaching of root zone showed that, the crop yield and nitrate leaching was not sensitive to the fertilizer amount when the irrigation was very less. Nitrate leaching was mainly occurred in the growth period of waxy maize due to more precipitation during this time. When the irrigation and fertilizer for waxy maize and winter wheat were both40cm and400kg ha-1, respectively, the nitrate leaching into groundwater equaled to47.9%of application amount of nitrogen fertilizer. In the condition of normal irrigation and fertilization, NUPE, NUE, NPFP, and Wpi was relatively higher. In the long-term time, the time of nitrate leaching from upper soil (lm) to lower soil (2m) or groundwater (15m) was basically the same under same irrigation condition, but nitrate leaching had a increasing time-lag with soil depth. Under different irrigation conditions, nitrate leaching time was shorter when irrigation amount was larger.(5) The responses of nitrogen leaching in different depth and crop yield to varying agricultural management practices and meteorological conditions were simulated by combing the RZWQM2model and ArcGIS software. As the results showed, all the increasing of irrigation and fertilizer, temperature, and precipitation could enhance crop yield and nitrogen leaching. On spatial scale, the effects of different influence factors to nitrate leaching at depth of groundwater lever were not same due to the various soil texture. The groundwater of sandy soil area in the western Daxing district are vulnerable to nitrage pollution.更多还原