摘要
[目的]揭示土壤、地形和作物参数对DNDC模型模拟的茶园地上部生物量与氮磷流失量的影响程度,识别各模型输出的最敏感参数,为模型的率定提供指导方案。[方法]基于天目湖流域茶园径流小区数据和DNDC模型,应用定性的Morris法和定量的Sobol法对模拟的茶园地上部生物量、氮磷流失量(包括氮磷淋失量和含氮气体排放量)和水分渗漏量进行敏感性分析,评估了各个模型输出结果的敏感参数,并借助蒙特卡洛方法量化模型输出的不确定性。[结果]对地上部生物量产生最主要影响的参数为成熟作物中籽粒占总生物量之比;对氮淋失量产生最主要影响的参数为土壤黏粒含量;对磷淋失量产生最主要影响的参数为土壤容重;对水分渗漏量产生最主要影响的参数为坡度;对N_(2)O排放量产生最主要影响的参数为初始土壤有机碳含量;对NO排放量和对NH3排放量产生最主要影响的参数为土壤pH。各输出结果中,磷淋失量的不确定性最大,变异系数为43.84%;水分渗漏量的不确定性最小,变异系数为2.36%。Morris法和Sobol法相比,Morris法优在方便省时,Sobol法优在准确全面,两种方法得到的参数敏感度排序总相关性为0.86,具有一定可替代性。[结论]在率定模型之前,先应用Morris法确定敏感参数的范围,再应用Sobol法精确量化,从而快速准确地筛选出模型敏感参数。在此基础上,固定非敏感参数,先重点校正敏感作物参数,再校正敏感土壤参数。本研究可为DNDC模型参数的本地化提供科学支撑,提高该模型对天目湖流域茶园生态系统的模拟精度和效率。
[Objective]The objectives of this study are to assess the influence of soil,topography,and crop parameters on the simulation accuracy of aboveground biomass and losses of nitrogen and phosphorus in tea gardens using the DNDC model,to identify the most sensitive parameters for each model output,and to provide guidance for calibrating the model.[Methods]Based on the data of tea garden plot in the Tianmu Lake Basin and DNDC model,qualitative Morris method and quantitative Sobol method were applied to carry out sensitivity analysis on the simulated aboveground biomass,nitrogen and phosphorus loss(including nitrogen and phosphorus leaching and nitrogenous gas emissions)and water leaching of tea garden,and to evaluate the sensitive parameters of each output.The uncertainty of model output was quantified by Monte Carlo method.[Results]The most important parameter affecting aboveground biomass is the grain fraction of total biomass at maturity,the most important parameter affecting nitrogen leaching is soil clay fraction,the most important parameter affecting phosphorus leaching is soil bulk density,the most important parameter affecting water leaching is slope gradient,the most important parameter affecting N_(2)O emission is the initial soil organic carbon(SOC)content,and the most important parameter affecting NO emission and NH 3 emission is soil pH value.Among the output results,the uncertainty of phosphorus leaching loss is the largest,and the coefficient of variation is 43.84%.The uncertainty of water leakage is the least,and the coefficient of variation is 2.36%.Compared with Sobol method,Morris method is convenient and time-saving,while Sobol method is accurate and comprehensive.The correlation comparison of parameter sensitivity ranking obtained by the two methods is 0.86,which has certain substitutability.[Conclusion]Before calibrating the model,it is advisable to employ the Morris method for preliminary identification of sensitive parameters and subsequently utilize the Sobol method for precise quantification.This approach enables efficient and accurate screening of the model′s sensitive parameters.Based on this premise,the correction of sensitive crop parameters is prioritized while keeping non-sensitive parameters fixed,followed by addressing sensitive soil parameters.The results of this study can provide scientific support for the localization of DNDC model and improve the simulation accuracy and efficiency of the tea garden ecosystem in the Tianmu Lake Basin.
作者
叶志成
张莉莉
廖凯华
朱青
赖晓明
郭长强
Ye Zhicheng;Zhang Lili;Liao Kaihua;Zhu Qing;Lai Xiaoming;Guo Changqiang(Key Laboratory of Watershed Geographic Sciences,Nanjing Institute of Geography and Limnology,CAS,Nanjing 210008,China;College of Nanjing,University of Chinese Academy of Sciences,Nanjing 211135,China)
出处
《水土保持研究》
CSCD
北大核心
2024年第2期43-51,59,共10页
Research of Soil and Water Conservation
基金
国家自然科学基金面上项目(42171077)
江苏省碳达峰碳中和科技创新专项资金项目(BK20220042)
中国科学院青年创新促进会基金(2020317)
中国科学院南京地理与湖泊研究所自主部署科研项目(NIGLAS2022GS10)。
