土壤氮素行为及其模拟模型不确定性的Monte-Carlo分析

项目名称： 土壤氮素行为及其模拟模型不确定性的Monte-Carlo分析

项目编号： No.40871104

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 无线电电子学、电信技术

项目作者： 吕军

作者单位： 浙江大学

项目金额： 46万元

中文摘要： 不确定性是土壤氮素行为及其面源污染过程的重要基本特征。根据课题申请书的研究计划，针对氮素在水田土壤迁移转化过程中具有高度时空变异性，本课题主要进行了免耕稻田土壤养分变异特征的分析。结果表明，免耕条件下土壤氮、磷养分含量的时空变异具有很强烈的随机和轨迹变化的组合特征，其变化方向（增或减）与土壤的粘粒含量有密切关系；由于水田土壤有机质含量普遍较高，因而有机质含量变化对供试土壤的氮素养分的变化不明显。从氮素流失和输出田间的情况分析，稻田土壤氮素的迁移转化过程，受人为的土地利用方式和农田管理措施、自然的流域水文循环等影响更大。因此，我们扩展了研究内容，对集水区域（小流域）尺度的氮素迁移转化特征进行了模拟和不确定性分析。在两个实验区用研究都表明，对土壤氮素迁移影响最大的是水循环；TN、TP的输出量和入河量均随流量的增加而增加（P<0.01）。Monte Carlo分析结果表明，河流段末流量参数的皮尔曼等级相关系数（SRCC）是模型其它输入参数的SRCC值的两倍以上。在SWAT模型中，区域水质最敏感的参数是SCS径流曲线数（CN2）。其次是段末浓度和背景浓度的影响，而流速和综合降解系数的影响很小。

中文关键词： 土壤氮素迁移转化;时空变异;不确定性;氮素流失;水文循环

英文摘要： Uncertainty is an important essential characteristics of nitrogen behavior in field and in processes of nitrogen diffuse pollution. According to the subject research plan, the natural spatiotemporal variations of soil nutrient contents in the paddy soil with zero tillage were measured and analyzed. Results showed that the spatiotemporal variability of the soil nitrogen and phosphorous contents are resulted from the combination of both the stochastic and tracking factors, and their variation direction (increase or decrease) is closely related to clay content in the paddy soil. However, the variation of the content of soil organic matter (OM) is not related to the variation of soil nitrogen content due to all paddy soil sampels have quite high OM in the research area. In the view of soil nitrogen lose rate and nitrogen export from the fields, soil nitrogen migration and transformation process in paddy field is more strongly effected by human activities such as land use type and farmland management practice, and natural hydrologic cycling, etc. We expanded our researches to the catchment (or watershed) scale for the nitrogen migration and its simulation uncertainty analysis in the research project. We find that the most important factor to influence soil nitrogen migration is water cycling. TN and TP export from field and into the river were significantly increased with runoff (P&lt;0.01). Results based on Monte-Carlo modelling showed that the Spearman Rank Correlation Coefficient（SRCC）for the rivers flow at the end of the reach is more than twice as much as the that value from other moddelling parameters in the simulation model. In the SWAT model, the most sensitive parameter to influence the soil nitrogen migtation is the SCS runoff curve number (CN2), next important factor is the nitrogen concentration in the runoff and background in the flow, but flow velocity and the comprehensive degradation coefficient are only small influences.

英文关键词： Migration and transformation of soil nitrogen; spatiotemporal variability; uncertainty; nitrogen loss; hydrological cycling