Reactive solute transport with a variable selectivity coefficient in an undisturbed soil column
Reactive solute transport with a variable selectivity coefficient in an undisturbed soil column
复制标题
在原状土柱中具有可变选择性系数的反应性溶质迁移
DOI：
--
复制
发表时间：
1997
期刊：
影响因子：
--
通讯作者：
R. Berndtsson
中科院分区：
文献类型：
--
作者：
K. Momii;Y. Hiroshiro;K. Jinno;R. Berndtsson
主要离子浓度的空间分布限制了田间土壤溶质运移过程的可预测性。因此，根据田间土壤试验和数值模拟结果，结合化学反应分析溶质运移具有重要意义。建立了一个考虑阳离子交换反应的模拟模型，并将其应用于未扰动土壤的溶质运移分析。采用Levenberg-Marquardt非线性最小二乘回归技术估计对流-弥散方程中的化学反应项，以同时满足物理和化学过程。利用KCl溶液连续入渗原状土柱后测定的Ca 2+、Mg 2+、Na +和K +的液相和固相浓度空间分布对模型的可靠性进行了检验。实验结果表明，Ca-Na和Ca-Mg交换的选择性系数基本保持不变，而Ca-K交换的选择性系数随固相中K +的当量分数的增加而增加。根据模拟结果讨论了交换选择性系数对反应性溶质运移的影响。当选择性系数为常数时，该模型不能预测固相中阳离子浓度的空间分布。因此，可以通过使用可变而不是恒定的选择性系数来改进模型预测。
The spatial distribution of major ion concentrations limits the predictability of solute transport processes in field soils. Therefore, it is important to analyze solute transport with chemical reactions based on results obtained from field soils and numerical simulation. A simulation model with cation-exchange reactions was developed and applied to solute-transport analysis of an undisturbed field soil. Chemical reaction terms in the convective-dispersive equation were estimated by the Levenberg-Marquardt nonlinear least-squares regression technique to satisfy physical and chemical processes simultaneously. The reliability of the model was tested with liquid-phase and solid-phase concentrations of measured spatial distributions of Ca 2+ , Mg 2+ , Na + , and K + after continuous infiltration of KCl solution into an undisturbed soil column. The experimental results revealed that the selectivity coefficients for Ca-Na and Ca-Mg exchange could be kept constant, while those for Ca-K exchange increased with the equivalent fraction of K + in the solid phase. The effects of the exchange selectivity coefficient on reactive solute transport are discussed based on the simulation results. When a constant selectivity coefficient was used, the model failed to predict the spatial distributions of cation concentrations in the solid phase. Thus, model predictions can be improved by use of variable instead of constant selectivity coefficients.