A novel segmental model to describe the complete soil water retention curve from saturation to oven dryness
A novel segmental model to describe the complete soil water retention curve from saturation to oven dryness
复制标题
DOI：
10.1016/j.jhydrol.2020.124649
复制
发表时间：
2020-05
影响因子：
6.4
通讯作者：
C. Du
中科院分区：
文献类型：
--
作者：
C. Du
使用残余含水量参数的传统保水模型无法准确描述土壤保水曲线（SWRC）的干端。我们开发了两段和三段平滑连续 SWRC 模型，承认存在三种饱和状态，适用于从饱和到绝干的含水量。两个模型中接近烘干端的方程都是基于土壤含水量与吸力对数之间比例关系的假设而获得的。两个模型在潮湿范围内的方程都是通过删除残余水含量参数从 van Genuchten 模型和/或 Brook-Corey 模型修改而来的。连接点的连续性和平滑性条件将三段和两段模型的拟合参数分别从 9 减少到 4，从 6 减少到 3。这两个模型都使用华盛顿土壤数据库和非饱和土壤水力数据库 (UNSODA) 的几乎整个范围内的保水数据进行了测试。这两个模型还与七个现有的 SWRC 模型进行了比较，并根据调整后的决定系数 (adj R2)、均方根误差 (RMSE) 和 Kashyap 信息标准 (KIC) 进行了评估。测试结果表明，三段模型在拟合干端SWRC方面表现优于两段模型。对于三段模型，adj R2值大于0.90，RMSE小于0.02，平均KIC值为-67.0，这些值优于其他模型。此外，当干燥端附近的数据不可用时，三段模型对于推断干燥范围内的值而言是稳健且稳定的。
Traditional water retention models that use the residual water content parameter fail to accurately describe the dry end of the soil water retention curve (SWRC). We developed two- and three-segment smooth, continuous SWRC models, acknowledging the existence of three saturation states, for water contents ranging from saturation to oven-dryness. The equations near the oven dryness end in both models were obtained based on the assumption of proportionality between the soil water content and the logarithm of suction. The equations in the wet range for both models were modified from the van Genuchten model and/or Brook-Corey model(s) by removing the residual water content parameter. The continuity and smoothness conditions at the junction points reduced the fitting parameters of the three- and two- segment models from 9 to 4 and from 6 to 3, respectively. Both models were tested with water retention data from the Washington soil database and the Unsaturated Soil Hydraulic Database (UNSODA) over nearly the entire range. Both models were also compared with seven existing SWRC models and were evaluated on the basis of the adjusted coefficient of determination (adj R2), the root mean square error (RMSE) and the Kashyap information criteria (KIC). The testing results showed that the three-segment model performed better than the two-segment model at fitting the SWRC at the dry end. For the three-segment model, the adj R2value was greater than 0.90, the RMSE was less than 0.02, and the mean KIC value was −67.0, and these values were superior to those of the other models. In addition, the three-segment model was robust and stable for extrapolation of values in the dry range when data near the dry end were unavailable.