Photogrammetric high resolution methods for nested parametrisation and validation of a physical-based soil erosion model

该项目的主要目标是通过摄影测量和优化技术以及不同尺度上基于物理的侵蚀建模方法来模拟土壤侵蚀。常用的基于过程的土壤侵蚀建模方法仅在确定参数的尺度上有效。由于侵蚀过程的复杂性、多变性和间断性，向其他时间和空间尺度转移的可能性有限。因此，所观察到的现象(细沟侵蚀或有轨电车线路中的集中流动)即使在复杂的模型方法中也不是或仅仅是初级的。摄影测量方法(例如，参考热像和三维表面重建方法)允许以一致的程序在不同的过程尺度上观察地表径流和改变土壤表面。侵蚀过程造成的土壤流失可以使用摄影测量数据进行验证。由于三维模型具有很高的空间和时间分辨率，因此基于物理的侵蚀模型不仅可以验证，而且可以通过随机方法进行参数化。此外，摄影测量观测(RGB和热)允许高度分辨的过程区分(飞溅、片状和细沟侵蚀、颗粒传输和沉积)。因此，采用适当的模型方法来模拟与尺度有关的过程是可能的(例如，分开考虑席流和细沟流)，并能够从根本上观察泥沙运动的相互联系以及事件频率和事件强度之间的关系。

The main objective of the project is the simulation of soil erosion by means of photogrammetric measurements and optimisation techniques as well as physically-based erosion modelling approaches on different scales. Common process based soil erosion modelling approaches are only valid on the scale at which parameters are identified. Due to the complexity, variability and discontinuity of erosion processes a transferability to other temporal and spatial scales is limited. Hence, observed phenomena (rill erosion or concentrated flow in tram lines) are not or only rudimentary covered even by complex model approaches. Photogrammetric measurements (e.g. referenced thermal images and 3D surface reconstruction methods) allow the observation of surface runoff and the altering of the soil surface across different process scales with a consistent procedure. Soil losses resulting from erosion processes can be validated using photogrammetric data. Physically based erosion models can not only be validated but also parameterised by means of stochastic approaches due to the great information density of 3D-models with high spatial and temporal resolution. Additionally, photogrammetric observations (RGB and thermal) allow for highly resolved process discrimination (splash-, sheet- and rill erosion, particle transport and deposition). Thus, an implementation of adapted model approaches for the simulation of scale-dependent processes (e.g. separate consideration of sheet and rill flow) is possible and enables a fundamentally new observation of the inter-connectivity of sediment transport as well as the relation between event frequency and event magnitude.