Monitoring and modelling of non-equilibrium soil water dynamics and lateral subsurface flow in hillslope soils

在排水良好的非饱和土壤中，水主要是垂直流动的。横向流动开始于土壤接近含水饱和和毛细力消失的位置。即使使用空间分布的三维数值模型，也不能真实地描述山坡上沿阻碍土壤水平边界的横向流动和其他非均质性的发生。目前还没有一个基于过程的斜坡土非饱和带瞬态横向流动模型概念。发展这一概念的一个主要困难是，由于结构非均质性，田间土壤中的水动力学表现出非平衡效应和滞后。因此，横向流动已经在接近于零的局部水势处触发，即远远早于通常假设的完全含水饱和发生。另一个困难是需要2D或3D山坡表示，以及相应的对数据和计算能力的高要求。在这个项目中，我们开发了一个概念框架，以物理一致的方式描述一维垂直流动的非平衡动力学和滞后。该分析将基于VAMOS渗滤仪系统和TERENO-SoilCan渗滤仪网络提供的独特数据集，该网络自2013年以来监测不同农田土壤（3D）和渗滤仪（1D）的含水量和基质电位。升级到山坡尺度将通过由局部含水饱和度（即零电位）触发的垂直一维柱的动态横向耦合来完成。这将允许描述横向流动在大范围内大大降低复杂性。项目结构为(1)统一土壤水滞回和水力非平衡模型的概念（Vogel），(2)斜坡尺度侧流的开始（Gerke），(3)斜坡尺度模型的开发和模型结构的评价（Wöhling）。为了验证模型，我们将设计联合现场和实验室实验。所提出的模型框架有望预测非饱和土中横向流动的开始和动力学。因此，期望形成一个更现实的基础来量化在山坡和集水区尺度上临时变化的流动路径和旅行时间，这是理解和预测在可变饱和地下溶质运输的一个臭名昭著的问题。

In well-drained, unsaturated soils, water moves predominantly vertically. Lateral flow is initiated at locations where the soil approaches water saturation and capillary forces vanish. The onset of lateral flow along impeding soil horizon boundaries and other heterogeneities in hillslopes cannot be described realistically even with spatially-distributed 3D numerical models. A process-based model concept for transient lateral flow in the unsaturated zone of hillslope soils is still missing. One major difficulty to develop such a concept is that water dynamics in field soil exhibit non-equilibrium effects and hysteresis due to structural heterogeneities. Consequently, lateral flow is triggered already at local water potentials close to zero, i.e. far before complete water saturation occurs as is commonly assumed. Another difficulty is the need for a 2D or 3D representation of the hillslope and the corresponding high demand of both data and computing power.In this project, we develop a conceptual framework to described non-equilibrium dynamics and hysteresis for 1D vertical flow in a physically consistent way. The analysis will be based on unique data sets provided by the VAMOS lysimeter system and the TERENO-SoilCan lysimeter network, which monitor water contents and matric potentials in different field soils (3D) and lysimeters (1D) since 2013. Upscaling to the hillslope-scale will be accomplished by a dynamic lateral coupling of vertical 1D columns triggered by local water saturation (i.e. zero potential). This will allow to describe lateral flows at large scales with considerably reduced complexity.The project structure is (1) unified concepts to model soil water hysteresis and hydraulic non-equilibrium (Vogel), (2) onset of lateral flow at hillslope scale (Gerke), (3) hillslope scale model development and evaluation of model structures (Wöhling). For model validation we will design joint field and lab experiments.The proposed model framework is expected to predict the onset and dynamics of lateral flow in unsaturated soils. Thus, it is expected to form a more realistic basis to quantify the temporarily changing flow paths and travel times at the scale of hillslopes and catchments which is a notorious problem for understanding and predicting the transport of solutes in the variably saturated subsurface.