Process-based modeling of soil hydrology and its verification at the Biosphere 2 Landscape Evolution Observatory

土壤水分运动是提供和调节多种生态系统服务的关键过程。然而，通过数学模型进行准确预测仍然是一个挑战，因为所有模型组件都存在很大的不确定性，即使使用像Richards方程这样基于过程的描述也是如此。数据同化方法提供了将来自不确定模型和不确定测量的信息组合成一种改进的一致状态描述的可能性，前提是不确定性可以正确量化。最大的不确定性通常与土壤的水力特性有关。将相关的水力参数纳入增强状态允许数据同化方法来估计它们。这甚至可以在存在模型误差（如优先流）的情况下实现，如果适当考虑这些误差的话。到目前为止，这种一致的描述只在一维土壤剖面的较小尺度上得到证实。在更大的尺度上，详细的基于过程的描述尚未完成。这是由于缺乏关于非均质土壤水力特性的信息，以及高度非线性和相互作用的过程。位于生物圈2号的景观演变观测站（Landscape Evolution Observatory， LEO）是进行山坡土壤水文实验研究的独特研究基础设施。它由三个人工山坡和一个广泛的传感器和采样网络组成。为了提高我们在更大的山坡尺度上的理解，本项目的目标是对低LEO土壤水分运动的一致和基于过程的描述，包括土壤水力参数的异质性和演化的表示。这将沿着以下思路解决：(i)通过数据同化在选定的山坡剖面上推断土壤水力参数的时间演变，（ii）通过水力实验和正演模拟确定土壤水力特性的异质性及其影响，以及（iii）通过数据同化方法获得并验证对山坡部分的一致描述。这个项目将回答当前的监测技术是否足以获得山坡水文的一致和有用的准确表示，如果是这样，如何实现描述以及有多大的不确定性。除此之外，我希望对LEO的异质性形成有定量的了解。

Soil water movement is a key process in several provisioning and regulating ecosystem services. However, it’s accurate prediction through mathematical models remains a challenge, because of large uncertainties in all model components, even when using process-based descriptions like the Richards equation.Data assimilation methods offer the possibility to combine information from uncertain models and uncertain measurements into an improved consistent description of the state, provided the uncertainties can be quantified correctly. The largest uncertainties are typically associated with the hydraulic properties of the soil. Including the pertinent hydraulic parameters into an augmented state allows data assimilation methods to estimate them. This can even be achieved in the presence of model errors like preferential flow, if these errors are considered appropriately. So far, such consistent descriptions have only been demonstrated on smaller scales up to one-dimensional soil profiles. On larger scales detailed process-based descriptions have not yet been accomplished. This is due to lack of information on the heterogeneous soil hydraulic properties in combination with the highly nonlinear and interacting processes.A unique research infrastructure for the experimental study of soil hydrology on hillslopes is the Landscape Evolution Observatory (LEO) at Biosphere 2. It consists of three artificial hillslopes with an extensive sensor and sampler network.To improve our understanding at this larger hillslope scale, the objective of this project is the consistent and process-based description of soil water movement at LEO including the representation of heterogeneity and evolution of soil hydraulic parameters. This will be addressed along the following lines: (i) deduce the evolution of soil hydraulic parameters in time through data assimilation at selected profiles in the hillslopes, (ii) determine the heterogeneity of soil hydraulic properties and their effects through hydraulic experiments and forward simulations, and (iii) gain and verify a consistent description of parts of the hillslope through data assimilation methods. This project will answer if current monitoring techniques suffice to obtain a consistent and usefully accurate representation of hillslope hydrology and if so, how the description is achieved and with what uncertainty. Beyond this, I expect quantitative insight into the formation of heterogeneity at LEO.