Schirmer, Mario

2020-4, Ramgraber, Maximilian, Camporese, Matteo, Renard, Philippe, Salandin, Paolo, Schirmer, Mario

The increasing use of wireless sensor networks and remote sensing permits real‐time access to environmental observations. Data assimilation frameworks tap into such data streams to autonomously update and gradually improve numerical models. In hydrogeology, such methods are relevant in areas of long‐term interest in water quality and quantity, for example, in drinking water production. Unfortunately, accurate hydrogeological predictions often demand a degree of geological realism, which is difficult to reconcile with the operational limitations of many data assimilation frameworks. Alluvial aquifers, for example, are sometimes characterized by paleo‐channels of unknown extent and properties, which may act as preferential flow paths. Gradually optimizing such fields in real‐time or quasi‐real‐time settings is a formidable task. Besides subsurface properties, ill‐specified model forcings are a further source of predictive bias, which an optimizer could learn to compensate. In this study, we explore the use of a quasi‐online optimizer based on the iterative batch importance sampling framework for a groundwater model of a field site near Valdobbiadene, Italy. This site is characterized by the presence of paleo‐channels and heavily exploited for drinking water production and irrigation. We use Markov chain Monte Carlo steps to explore new parameterizations while maintaining consistency between states and parameters as well as conformance to a multipoint statistics training image. We also optimize a preprocessor designed to compensate for potential bias in the model forcing. We achieve promising and geologically consistent quasi‐real‐time optimization, albeit at the loss of parameter uncertainty.

, Diem, Samuel, Renard, Philippe, Schirmer, Mario

Obtaining a quantitative understanding of river–groundwater interactions is of high practical relevance, for instance within the context of riverbank filtration and river restoration. Modeling interactions between river and groundwater requires knowledge of the river’s spatiotemporal water level distribution. The dynamic nature of riverbed morphology in restored river reaches might result in complex river water level distributions, including disconnected river branches, nonlinear longitudinal water level profiles and morphologically induced lateral water level gradients. Recently, two new methods were proposed to accurately and efficiently capture 2D water level distributions of dynamic rivers. In this study, we assessed the predictive capability of these methods with respect to simulated groundwater residence times. Both methods were used to generate surface water level distributions of a 1.2 km long partly restored river reach of the Thur River in northeastern Switzerland. We then assigned these water level distributions as boundary conditions to a 3D steady-state groundwater flow and transport model. When applying either of the new methods, the calibration-constrained groundwater flow field accurately predicted the spatial distribution of groundwater residence times; deviations were within a range of 30% when compared to residence times obtained using a reference method. We further tested the sensitivity of the simulated groundwater residence times to a simplified river water level distribution. The negligence of lateral river water level gradients of 20–30 cm on a length of 200 m caused errors of 40–80% in the calibration-constrained groundwater residence time distribution compared to results that included lateral water level gradients. The additional assumption of a linear water level distribution in longitudinal river direction led to deviations from the complete river water level distribution of up to 50 cm, which caused wide-spread errors in simulated groundwater residence times of 200–500%. For an accurate simulation of groundwater residence times, it is therefore imperative that the longitudinal water level distribution is correctly captured and described. Based on the confirmed predictive capability of the new methods to estimate 2D river water level distributions, we can recommend their application to future studies that model dynamic river–groundwater systems.

2014-1-10, Diem, Samuel, Renard, Philippe, Schirmer, Mario

, Diem, Samuel, Renard, Philippe, Schirmer, Mario

River restoration measures are becoming increasingly popular and are leading to dynamic river bed morphologies that in turn result in complex water level distributions in a river. Disconnected river branches, nonlinear longitudinal water level profiles and morphologically induced lateral water level gradients can evolve rapidly. The modeling of such river-groundwater systems is of high practical relevance in order to assess the impact of restoration measures on the exchange flux between a river and groundwater or on the residence times between a river and a pumping well. However, the model input includes a proper definition of the river boundary condition, which requires a detailed spatial and temporal river water level distribution. In this study, we present two new methods to estimate river water level distributions that are based directly on measured data. Comparing generated time series of water levels with those obtained by a hydraulic model as a reference, the new methods proved to offer an accurate and faster alternative with a simpler implementation

2013-1-10, Diem, Samuel, Renard, Philippe, Schirmer, Mario