数值反演模型模拟层状土壤水力性质
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

S152

基金项目:

国家自然科学基金项目(41571214)资助


Numerical Inversion-based Simulation of Hydraulic Properties of Stratic Soi
Author:
Affiliation:

Fund Project:

National Natural Science Foundation of China (No. 41571214)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    土壤水力性质主要包括土壤水力传导率和水分特征曲线,是进行土壤水分运动数值模拟必不可少的基础数据。自然界中的土壤大都呈层状分布,模拟中涉及到的水力性质参数较均质土壤更多。通过室内土柱入渗试验,利用Hydrus-1D软件对垂向一维水分运动进行数值模拟,以入渗过程中不同实测变量(累积入渗量、压力水头及二者组合)为目标函数开展数值反演,优化得到各层土壤水力性质。根据决定系数(R2)、均方根误差(RMSE)、变异系数(CVs)、优化参数相关矩阵及优化参数标准误差等评价指标,探讨了不同测定变量对解非唯一性、准确性等的影响。结果表明,目标函数仅含有累积入渗量时,虽然反演模拟值与实测值符合度较高,R2达到0.999,RMSE小于0.005 62 cm,但用压力水头实测值验证参数准确性,效果较差,R2最高仅为0.338,RMSE小于0.096 5 cm。使用压力水头作为目标函数进行参数优化求解时,拟合度R2达到0.855以上,RMSE小于0.038 4 cm;用累计入渗量实测值进行验证,效果较好,R2可达到0.99以上,RMSE小于0.038 56 cm。累积入渗量和压力水头联合反演求土壤水力性质拟合度较高,R2达到0.905以上,RMSE小于0.035 4 cm。虽然累积入渗量和压力水头联合反演时R2和RMSE指标次于累积入渗量,但累积入渗量和压力水头联合反演可以一定程度降低优化参数的相关性和标准误差,从而降低解的非唯一性,提高解的准确性。以上结论可为层状土壤瞬态流试验反演参数过程中,测量变量的选取提供科学依据。

    Abstract:

    [Objective] Soil hydraulic properties refer mainly to soil hydraulic conductivity and water characteristic curve, which are essential basic data for numerical simulation of soil water movement. Most soils in nature are present in layers, but in the simulation experiments so far done, the soils were assumed to be homogeneous in texture. [Method] Through an indoor soil column infiltration test, vertical one-dimensional water movement was numerically simulated with the aid of Hydrous-1D software, and numerical inversions were carried out with the variables (cumulative infiltration volume, pressure head and combination of both) measured in the infiltration process in the indoor soil column infiltration test as objective functions to optimize the parameters obtained of the hydraulic properties of each layers of the static soil. In the light of the evaluation indices, such as determination coefficient (R2), root mean square error (RMSE), coefficient of variation (CVS), correlation matrix of the optimized parameters and standard error of the optimized parameters, effects of different measured variables on non-uniqueness and accuracy of the solution were discussed. [Result] Results show that when the objective functions contain only the cumulative infiltration, although the value simulated through inversion is in good agreement with the measured value, R2 reached up to 0.999 and RMSE was less than 0.005 62 cm, but the parameters verified with the measured value of pressure head was not so good in accuracy, with R2 being only 0.338, and RMSE less than 0.096 5 cm. When the pressure head was used as the objective function for parameter optimization and solution, the fitting degree R2 reached up to 0.855 and RMSE was less than 0.038 4 cm; the use of the measured value of cumulative infiltration in verification was quite good in effect, with R2 reaching up to 0.99 and RMSE being less than 0.038 56 cm. The simulation of hydraulic properties of the soil through inversion using cumulative infiltration and pressure head in combination was quite high in goodness of fit, with R2 reaching more than 0.905 and RMSE less than 0.035 4 cm. Although in the inversion, the indices, R2 and RMSE, were inferior to that using cumulative infiltration, the inversion reduced correlation and standard error of the optimized parameters to a certain extent, thereby reducing the solution Non-uniqueness, and improving accuracy of the solution. [Conclusion] The above conclusions may serve as reference for selection of measured variables for use in the inversion for parameters of transient flow in the static soil.

    参考文献
    相似文献
    引证文献
引用本文

周晓冰,栾永霞,林青,徐绍辉.数值反演模型模拟层状土壤水力性质[J].土壤学报,2021,58(5):1214-1223. DOI:10.11766/trxb202005030212 ZHOU Xiaobing, LUAN Yongxia, LIN Qing, XU Shaohui. Numerical Inversion-based Simulation of Hydraulic Properties of Stratic Soi[J]. Acta Pedologica Sinica,2021,58(5):1214-1223.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2020-05-03
  • 最后修改日期:2020-12-29
  • 录用日期:2021-03-16
  • 在线发布日期: 2021-03-18
  • 出版日期: 2021-09-11