Search

River floods cause extensive losses to economy, ecology, and society throughout the world. They are driven by the space‐time structure of catchment rainfall, which is determined by large‐scale, or even global‐scale, atmospheric processes. The identification of coherent, large‐scale atmospheric circulation structures that determine the moisture transport and convergence associated with rainfall‐induced flooding can help improve its predictability and phenomenology. In this paper, we extend a methodology, used for the analysis of extreme rainfall events, to high streamflow events (HSEs). The approach combines multiple machine learning methods to link HSEs to atmospheric circulation patterns. An application to the German streamflow network using reanalysis data for the period 1960 to 2012 is presented. Daily streamflow from 166 gauges, homogeneously distributed across Germany, are used. Geopotential height fields and integrated vapor transport (IVT) are derived from reanalysis data. An unsupervised neural network, Self Organizing Maps, is applied to geopotential height to identify a finite number of circulation patterns (CPs). Event synchronization between CPs and HSEs is used to establish if they are linked or not. If they are linked, the Event Synchronization method computes the delay between the occurrence of a CP and a HSE. Finally, local logistic regression is used to estimate the probability of occurrence of a HSE, as function of CP and IVT. We demonstrate that our approach is very effective to evaluate HSE probability occurrence across Germany. Key Points: Atmospheric circulation patterns that generate high streamflow events in Germany are identifiedEvent synchronization is used to determine synchronization and delay between atmospheric circulation patterns and high streamflow eventsThe occurrence probability of high streamflow events is conditioned on atmospheric circulation patterns and integrated water vapor transport [ABSTRACT FROM AUTHOR]

Especially for extreme precipitation or floods, there is considerable spatial and temporal variability in long term trends or in the response of station time series to large-scale climate indices. Consequently, identifying trends or sensitivity of these extremes to climate parameters can be marked by high uncertainty. When one develops a nonstationary frequency analysis model, a key step is the identification of potential trends or effects of climate indices on the station series. An automatic clustering procedure that effectively pools stations where there are similar responses is desirable to reduce the estimation variance, thus improving the identification of trends or responses, and accounting for spatial dependence. This paper presents a new hierarchical Bayesian approach for exploring homogeneity of response in large area data sets, through a multicomponent mixture model. The approach allows the reduction of uncertainties through both full pooling and partial pooling of stations across automatically chosen subsets of the data. We apply the model to study the trends in annual maximum daily stream flow at 68 gauges over Germany. The effects of changing the number of clusters and the parameters used for clustering are demonstrated. The results show that there are large, mainly upward trends in the gauges of the River Rhine Basin in Western Germany and along the main stream of the Danube River in the south, while there are also some small upward trends at gauges in Central and Northern Germany. [ABSTRACT FROM AUTHOR]

Groundwater abstraction and drainage are considered to be the main drivers for the salinization of low‐lying coastal groundwater systems, while the role of past boundary conditions is less clear. In this study, 3‐D paleo‐hydrogeological variable‐density groundwater flow and salt transport modeling ("paleo‐modeling") is applied to reconstruct the evolution of groundwater salinities during the Holocene, that is, the last 9,000 years, in Northwestern Germany. Novel aspects of this study include the consideration of highly resolved time‐variant boundary conditions in a 3‐D paleo‐modeling framework, for example, sea‐level rise, surface elevation and coastline changes, development of drainage networks and groundwater abstraction, as well as the quantification of isolated processes impacting salinization. Results show that salinization was a function of sea‐level rise from 9000 BP until 1300 CE. The creation of the dike line ∼1300 CE set the starting point for increasing anthropogenic control of the hydro(geo)logical system: changes in surface elevation and drainage of low‐lying marshes have become main drivers for salinization after 1600 CE when peat was artificially degenerated. Moreover, changes in the dike line caused by storm floods impacted the salinities. Model results for 2020 CE match well with present‐day salinity observations. Yet, salinization will continue in the future, as the hydro(geo)logical system has not reached an equilibrium. The presented paleo‐modeling framework can be viewed as a blueprint for similar low‐lying coastal groundwater systems, influenced by marine transgression and human development. Thereby, it enables the reconstruction of meaningful present‐day salinity distributions, serving as a vital basis for modeling future groundwater systems in a changing climate. Key Points: Salinization of a low‐lying coastal groundwater system during the Holocene, that is, 9000 BP until present‐day, was investigatedSea‐level rise and the evolution of paleogeography were major controls on groundwater salinization during most of the timeSurface elevation changes due to land cultivation, as well as the development of drainage networks, control salinization since ∼1600 CE [ABSTRACT FROM AUTHOR]

Large‐scale groundwater models are required to estimate groundwater availability and to inform water management strategies on the national scale. However, parameterization of large‐scale groundwater models covering areas of major river basins and more is challenging due to the lack of observational data and the mismatch between the scales of modeling and measurements. In this work, we propose to bridge the scale gap and derive regional hydraulic parameters by spectral analysis of groundwater level fluctuations. We hypothesize that specific locations in aquifers can reveal regional parameters of the hydraulic system. We first generate ensembles of synthetic but realistic aquifers which systematically differ in complexity. Applying Liang and Zhang's (2013), https://doi.org/10.1016/j.jhydrol.2012.11.044, semi‐analytical solution for the spectrum of hydraulic head time series, we identify for each ensemble member and at different locations representative aquifer parameters. Next, we extend our study to investigate the use of spectral analysis in more complex numerical models and in real settings. Our analyses indicate that the variance of inferred effective transmissivity and storativity values for stochastic aquifer ensembles is small for observation points which are far away from the Dirichlet boundary. Moreover, the head time series has to cover a period which is roughly 10 times as long as the characteristic time of the aquifer. In deterministic aquifer models we infer equivalent, regionally valid parameters. A sensitivity analysis further reveals that as long as the aquifer length and the position of the groundwater measurement location is roughly known, the parameters can be robustly estimated. Plain Language Summary: We build large‐scale (regional) computer models of the subsurface flow conditions in order to quantify the long‐term shift in groundwater storage and response on the national level under changing climatic conditions and increasing human water demands. These models must be fed with hydrogeological parameters obtained from subsurface observation wells, drilling logs, and hydraulic tests in conjunction with (hydro)geological and geostatistical methods. In some regions these wells are sparsely distributed and derived parameters are representative only for small areas. We hypothesize that groundwater level records can reveal regional aquifer information when analyzed in the spectral domain. In order to bridge that scale gap and because groundwater level time series are generally available, we propose to infer regional parameters by analyzing the frequency content (spectrum) of long groundwater level time series. The required parameters were determined using mathematical formulations of the theoretical spectrum for simplified settings. We tested the methodology in computer models with limited complexity and found that the groundwater level time series indeed contain regional information if the time of observation is sufficiently long. Lastly, we apply the spectral analysis to real groundwater data to test the capability of the method to infer regional aquifer parameters in real aquifers. Key Points: We successfully tested the spectral analysis of groundwater level fluctuations in numerical models and obtained regional aquifer parametersIn a sensitivity analysis of the spectral analysis using field data, the storativity and the response times could be robustly estimatedThe application of the suggested methodology to the field data from a catchment in central Germany produced plausible results [ABSTRACT FROM AUTHOR]

Improving nitrogen (N) status in European water bodies is a pressing issue. N levels depend not only on current but also past N inputs to the landscape, that have accumulated through time in legacy stores (e.g., soil, groundwater). Catchment‐scale N models, that are commonly used to investigate in‐stream N levels, rarely examine the magnitude and dynamics of legacy components. This study aims to gain a better understanding of the long‐term fate of the N inputs and its uncertainties, using a legacy‐driven N model (ELEMeNT) in Germany's largest national river basin (Weser; 38,450 km2) over the period 1960–2015. We estimate the nine model parameters based on a progressive constraining strategy, to assess the value of different observational data sets. We demonstrate that beyond in‐stream N loading, soil N content and in‐stream N concentration allow to reduce the equifinality in model parameterizations. We find that more than 50% of the N surplus denitrifies (1480–2210 kg ha−1) and the stream export amounts to around 18% (410–640 kg ha−1), leaving behind as much as around 230–780 kg ha−1 of N in the (soil) source zone and 10–105 kg ha−1 in the subsurface. A sensitivity analysis reveals the importance of different factors affecting the residual uncertainties in simulated N legacies, namely hydrologic travel time, denitrification rates, a coefficient characterizing the protection of organic N in source zone and N surplus input. Our study calls for proper consideration of uncertainties in N legacy characterization, and discusses possible avenues to further reduce the equifinality in water quality modeling. Plain Language Summary: Lowering nitrogen (N) amounts in European surface waters is a pressing issue. N levels largely result from fertilizer application in agricultural areas, and deposition of atmospheric N coming from fossil fuel combustion. These N inputs to the landscape can accumulate below the ground surface in so‐called legacy stores (including the soil and aquifer), from which they can be released progressively through time. Therefore, N levels depend not only on the recent N inputs, but also on their history. Our modeling study aims to improve our understanding of the long‐term fate of the N inputs and its uncertainties in Germany's largest national river basin (Weser) over the period 1960–2015. It suggests that more than 50% of the N inputs to land is lost to the atmosphere (denitrification, 1480–2210 kg ha−1) and the stream export amounts to around 18% (410–640 kg ha−1), leaving behind as much as around 16% (264–820 kg ha−1) in the landscape (legacy). However, the uncertainties in these estimates remain large, partly due to a lack of observational data on internal (legacy) components and uncertainties in N inputs. Overall, our study calls for proper consideration of uncertainties in N legacy characterization, and discusses possible avenues to further reduce them. Key Points: We use a parsimonious model to examine the long‐term (1960–2015) fate of nitrogen inputs to the landscape including the legacy buildupMore than 50% of the nitrogen surplus denitrifies and around 491 kg ha−1 accumulates in legacy stores in Germany's largest river basinHydrologic travel time and denitrification rates largely contribute to the residual uncertainty in the simulated nitrogen legacies [ABSTRACT FROM AUTHOR]

Debris floods most commonly occur in steep mountain channels and on their alluvial fans but can also occur on small gravel bed rivers with watershed areas up to several hundred square kilometers. This became obvious during July 2021 and November 2021 debris floods in northwestern Germany and southwestern British Columbia, Canada. We subdivide debris floods into three categories: those triggered by a supra‐critical bed shear stress ratio, form by dilution from debris flows, and resulting from outbreak floods. This trichotomy challenges traditional hazard assessments; debris floods classify as a fluvial process, yet their destructive mechanics are difficult to characterize. Key hazards interact spatially and temporally in debris floods: inundation, scour, sediment transport and deposition and bank erosion. We describe approaches to quantify hazards by systematically accounting for these processes and introduce a novel approach for hazard quantification and mapping in which flow velocity, depth, and presumed fluid density are combined with the annual event frequency for all event scenarios. The derivative "composite hazard maps" are equally valid for debris floods and debris flows. Isolines of bank erosion based on a probabilistic analysis of a physically based model are added to capture the potential of debris floods to abruptly widen their channels. Substantial challenges remain, specifically in the reliable prediction of sediment transport and progressive bank erosion. Our intention is to homogenize debris‐flood hazard assessments and especially mapping methodologies. This could allow for systematic integration with landuse policies assuring consistency among approaches executed by practitioners acting in this field. Plain Language Summary: Debris floods can be highly destructive. We provide a comprehensive methodology to assess debris‐flood hazards with the goal of developing a consistent approach, allowing for better input to quantitative risk assessments and the selection of appropriate mitigation measures. We suggest a method for debris‐flood mapping that combines the recurrence interval of various debris flood scenarios with their impact force and bank erosion to support a multi‐faceted hazard assessment. Key Points: Debris flood assessments are under‐researched and poorly constrained. This contribution attempts to rectify that situationBeyond debris flood inundation, this contribution outlines methods for sediment bulking, sediment transport, and bank erosionA new method for hazard mapping is proposed based on impact force and event frequency. This allows direct comparison of study sites [ABSTRACT FROM AUTHOR]

The dynamic relationships between water flux and storage, together with the associated water ages and speed of hydrological responses (as proxies for velocity and celerity respectively) are fundamental to understanding how catchments react to hydroclimate perturbations, such as floods and droughts. Using results from a calibrated, tracer‐aided ecohydrological model (EcH2O‐iso) we analyzed the dynamics of storage‐flux‐age‐response time (RT) interactions at scales that resolve the internal heterogeneity of these non‐stationary relationships. EcH2O‐iso has previously shown an adequate representation of ecohydrological flux partitioning and storage dynamics (celerity), and water ages (velocity) over 11‐year at Demnitzer Millcreek catchment (DMC, 66 km2), a drought‐sensitive, lowland catchment in Germany. The 11‐year period had marked hydroclimatic contrasts facilitating the evaluation of flux‐storage‐age‐RT dynamics under different wetness anomalies. Our results show that the spatio–temporal variability of soil moisture and ecohydrological partitioning dynamics reflect both land use (especially forest cover) and distinct soil units (i.e., brown earth vs. podzolic soils). Spatial differences in RTs of storage were driven by rapid soil evaporation and transpiration responses to rainfall, which revealed a divergence of transpiration ages from RTs. RTs of groundwater and streamflow were fast (days), but mediation by soil water storage dynamics caused marked separation from water ages (years‐decades) of deeper flow paths. Analysis of RTs and ages revealed a degradation of process representation with coarsening model spatial resolution. This study uses novel analysis of the spatio‐temporal interactions of flux‐storage‐age‐RT from a model to understand the sensitivity and resilience of catchment functionality to hydroclimatic perturbations. Key Points: Spatio‐temporal variability of soil storage and ecohydrological partitioning was modulated by vegetation characteristicsTranspiration, groundwater, and streamflow response times were distinct from water ages, with spatial differences driven by vegetation unitsLower model resolution reduced spatial variability and increased the difference of catchment response and water age of fluxes and storages [ABSTRACT FROM AUTHOR]

The quest for improved hydrological models is one of the big challenges in hydrology. When discrepancies are observed between simulated and measured discharge, it is essential to identify which algorithms may be responsible for poor model behavior. Particularly in complex hydrological models, different process representations may dominate at different moments and interact with each other, thus highly complicating this task. This paper investigates the analysis of the temporal dynamics of parameter sensitivity as a way to disentangle the simulation of a hydrological model and identify dominant parameterizations. Three existing methods (the Fourier amplitude sensitivity test, the extended Fourier amplitude sensitivity test, and Sobol's method) are compared by applying them to a TOPMODEL implementation in a small mountainous catchment in the tropics. For the major part of the simulation period, the three methods give comparable results, while the Fourier amplitude sensitivity test is much more computationally efficient. This method is also applied to the complex hydrological model WaSiM-ETH implemented in the Weisseritz catchment, Germany. A qualitative model validation was performed on the basis of the identification of relevant model components. The validation revealed that the saturation deficit parameterization of WaSiM-ETH is highly susceptible to parameter interaction and lack of identifiability. We conclude that temporal dynamics of model parameter sensitivity can be a powerful tool for hydrological model analysis, especially to identify parameter interaction as well as the dominant hydrological response modes. Finally, an open source implementation of the Fourier amplitude sensitivity test is provided. [ABSTRACT FROM AUTHOR]

In this paper we investigate the use of hydrological models as learning tools to help improve our understanding of the hydrological functioning of a catchment. With the model as a hypothetical conceptualization of how dominant hydrological processes contribute to catchment-scale response, we investigate three questions: (I) During which periods does the model (not) reproduce observed quantities and dynamics? (2) What is the nature of the error during times of bad model performance? (3) Which model components are responsible for this error? To investigate these questions, we combine a method for detecting repeating patterns of typical differences between model and observations (time series of grouped errors, TIGER) with a method for identifying the active model components during each simulation time step based on parameter sensitivity (temporal dynamics of parameter sensitivities, TEDPAS). The approach generates a time series of occurrence of dominant error types and time series of parameter sensitivities. A synoptic discussion of these time series highlights deficiencies in the assumptions about the functioning of the catchment. The approach is demonstrated for the Weisseritz headwater catchment in the eastern Ore Mountains. Our results indicate that the WaSiM-ETH complex grid-based model is not a sufficient working hypothesis for the functioning of the Weisseritz catchment and point toward future steps that can help improve our understanding of the catchment. [ABSTRACT FROM AUTHOR]

The reliability of the machine learning model prediction for a given input can be assessed by comparing it against the actual output. However, in hydrological studies, machine learning models are often adopted to predict future or unknown events, where the actual outputs are unavailable. The prediction accuracy of a model, which measures its average performance across an observed data set, may not be relevant for a specific input. This study presents a method based on metamorphic testing (MT), adopted from software engineering, to assess the prediction reliability where the actual outputs are unknown. In this method, the predictions for a group of related inputs are considered consistent only if the input and output follow certain relations that are deduced from the properties of the system being modeled. For instance, the predicted runoff volume should increase in a rainfall‐runoff model as the rainfall magnitude of an input increases. In this study, the MT‐based method was applied to assess the predictions made by various machine learning models that were trained to predict the magnitude of flood events in Germany. Surprisingly, the prediction accuracy of a model and its ability to provide consistent predictions were found to be uncorrelated. This study further investigated the factors influencing the assessment result of a given input, such as its similarity to observed data. Overall, this research shows that MT is an effective and simple method for detecting inconsistent model predictions and is recommended when a model is employed to making predictions under new conditions. Plain Language Summary: In hydrological studies, machine learning models are often built to learn the statistical correlations between random variables of interests. In these models, predictions are made based on learned statistical correlations, and the involved hydrological processes are not explicitly considered. Thus, it is difficult to examine the mechanisms that lead to the prediction made for a given input. Metamorphic testing (MT) is a technique to assess the correctness of the predicted output for a given input that does not require knowledge of the actual output. In MT, the relation among a group of related inputs and the corresponding outputs is examined, and the predictions are considered consistent if this relation satisfies certain criteria that are related to the properties of the system being modeled. This study applied MT to assess various machine learning models that predicted the magnitude of flood events in Germany. MT was found to be very effective in detecting inconsistent predictions. Many models failed to capture the positive correlation between precipitation magnitude and the magnitude of flood events, even for those with high prediction accuracies. The MT‐based method presented in this study is useful for assessing the reliability of model predictions, especially when the actual outcomes are unknown. Key Points: This study presents a metamorphic testing‐based method to assess the reliability of model predictions when the actual outcomes are unknownSpecific tests can be designed using hydrological knowledge to check if a model preserves certain properties of the system being modeledSome models with high prediction accuracies can often respond incorrectly when input changes and thereby produce inconsistent predictions [ABSTRACT FROM AUTHOR]

Despite reduction measures, nitrate and aluminum concentrations remain high in aquifers in northwestern Europe. To evaluate the effectiveness of groundwater protection policies, the long‐term fate of these contaminants in groundwater needs to be understood. The groundwater catchment of the Haren water works, NW Germany, was characterized hydrogeochemically in the late 1990s, which provides an opportunity to study the solute fronts over a two‐decade period and conduct a post‐audit of the predicted front movement. Results indicate that, despite a significant reduction of the atmospheric acid loads, the acidification of soil and groundwater at the forest site persists. Removal of sorbed aluminum is required to induce a noticeable improvement, which will take at least several decades. The unexpected appearance of nitrate at the site, caused by a land use change in 1998, highlights the need for long‐term monitoring. Core data at the agricultural site show that the denitrification front has moved very little between 1998 and 2017, in accordance with previous forecasts. Denitrification by‐products, mainly sulfate and nitrogen, have migrated from the upper into the lower aquifer. A reactive transport model demonstrated how the link between the regional groundwater flow, pyrite oxidation, and the temporal variability of the nitrate concentration in recharge water, as reconstructed from age tracers, result in the observed vertical distribution of sulfate and nitrogen. This study demonstrates how long‐term monitoring, aided by model‐based data interpretation, can be used to successfully study and predict the fate of contaminants in groundwater. Key Points: Position of reaction fronts for nitrate and acidification in aquifers checked after 20+ yearsAcidification of soil and groundwater still strong, despite reductions of atmospheric inputNitrate front has moved very slowly over last 20 years but denitrification products have migrated into lower aquifer [ABSTRACT FROM AUTHOR]

The ecosystem function of vegetation to attenuate export of nutrients is of substantial importance for securing water quality. This ecosystem function is at risk of deterioration due to an increasing risk of large‐scale forest dieback under climate change. The present study explores the response of the nitrogen (N) cycle of a forest catchment in the Bavarian Forest National Park, Germany, in the face of a severe bark beetle (Ips typographus Linnaeus) outbreak and resulting large‐scale forest dieback using top‐down statistical‐mechanistic modeling. Outbreaks of bark beetle killed the dominant tree species Norway spruce (Picea abies (L.) H.Karst.) in stands accounting for 55% of the catchment area. A Bayesian hierarchical model that predicts daily stream NO3 concentration (C) over three decades with discharge (Q) and temperature (T) (C‐Q‐T relationship) outperformed alternative statistical models. A catchment model was subsequently developed to explain the C‐Q‐T relationship in top‐down fashion. Annually varying parameter estimates provide mechanistic interpretations of the catchment processes. Release of NO3 from decaying litter after the dieback was tracked by an increase of the nutrient input parameter cs0. The slope of C‐T relation was near zero during this period, suggesting that the nutrient release was beyond the regulating capacity of the vegetation and soils. Within a decade after the dieback, the released N was flushed out and nutrient retention capacity was restored with the regrowth of the vegetation. Key Points: Pulse of nitrate export from a forest catchment in response to bark beetle infestation followed by recovery of nutrient retention capacityTop‐down, data‐driven Bayesian hierarchical model assists mechanistic interpretation of hydrochemical processesConcentration‐discharge‐temperature relationship is shaped by spatial heterogeneity of nutrient and seasonality of biogeochemical reactions [ABSTRACT FROM AUTHOR]

The main goal of this study is to explore whether the ideas established by surface water hydrologists in the context of predictions in ungauged basins can be useful in hydrogeology. The concrete question is whether it is possible to create predictive models for groundwater systems with no or few observations based on knowledge derived from similar groundwater systems which are well observed. To do so, this study analyzes the relationship between temporal dynamics of groundwater levels and climatic and physiographic characteristics. The analysis is based on data from 341 wells in Southern Germany with 10‐year daily groundwater hydrographs. Observation wells are used in confined and unconfined sand and gravel aquifers from narrow mountainous valleys as well as more extensive lowland alluvial aquifers. Groundwater dynamics at each location are summarized with 46 indices describing features of groundwater hydrographs. Besides borehole log‐derived geologic information, local and regional morphologic characteristics as well as topography‐derived boundary and climatic descriptors were derived for each well. Regression relationships were established by mining the data for associations between dynamics and descriptors with forward stepwise regression at a confidence level >95%. The most important predictors are geology and boundary conditions and, secondarily, climate, as well as some topographic features, such as regional convergence. The multiple regression models are in general agreement with process understanding linked to groundwater dynamics in unconfined aquifers. This systematic investigation suggests that statistical regionalization of groundwater dynamics in ungauged aquifers based on map‐derived physiographic and climatic controls may be feasible. Key Points: Important physiographic and climatic controls on groundwater dynamics in Southern Germany are quantifiedEmpirical models are successfully constructed to explain groundwater dynamics based on climatic and physiographic system characteristicsThe constructed empirical models are generally consistent with physical process understanding [ABSTRACT FROM AUTHOR]

This study proposes a new process‐based framework to characterize and classify runoff events of various magnitudes occurring in a wide range of catchments. The framework uses dimensionless indicators that characterize space–time dynamics of precipitation events and their spatial interaction with antecedent catchment states, described as snow cover, distribution of frozen soils, and soil moisture content. A rigorous uncertainty analysis showed that the developed indicators are robust and regionally consistent. Relying on covariance‐ and ratio‐based indicators leads to reduced classification uncertainty compared to commonly used (event‐based) indicators based on absolute values of metrics such as duration, volume, and intensity of precipitation events. The event typology derived from the proposed framework is able to stratify events that exhibit distinct hydrograph dynamics even if streamflow is not directly used for classification. The derived typology is therefore able to capture first‐order controls of event runoff response in a wide variety of catchments. Application of this typology to about 180,000 runoff events observed in 392 German catchments revealed six distinct regions with homogeneous event type frequency that match well regions with similar behavior in terms of runoff response identified in Germany. The detected seasonal pattern of event type occurrence is regionally consistent and agrees well with the seasonality of hydroclimatic conditions. The proposed framework can be a useful tool for comparative analyses of regional differences and similarities of runoff generation processes at catchment scale and their possible spatial and temporal evolution. Key Points: A new process‐based framework for characterizing runoff events of diverse sizes is proposed and applied to a wide set of catchmentsNovel indicators combining the space–time dynamics of event precipitation and catchment state are used for classificationThe derived event typology captures distinct shapes of event hydrographs even if streamflow data are not directly used for classification [ABSTRACT FROM AUTHOR]

Flood loss data collection and modeling are not standardized, and previous work has indicated that losses from different flood types (e.g., riverine and groundwater) may follow different driving forces. However, different flood types may occur within a single flood event, which is known as a compound flood event. Therefore, we aimed to identify statistical similarities between loss‐driving factors across flood types and test whether the corresponding losses should be modeled separately. In this study, we used empirical data from 4,418 respondents from four survey campaigns studying households in Germany that experienced flooding. These surveys sought to investigate several features of the impact process (hazard, socioeconomic, preparedness, and building characteristics, as well as flood type). While the level of most of these features differed across flood type subsamples (e.g., degree of preparedness), they did so in a nonregular pattern. A variable selection process indicates that besides hazard and building characteristics, information on property‐level preparedness was also selected as a relevant predictor of the loss ratio. These variables represent information, which is rarely adopted in loss modeling. Models shall be refined with further data collection and other statistical methods. To save costs, data collection efforts should be steered toward the most relevant predictors to enhance data availability and increase the statistical power of results. Understanding that losses from different flood types are driven by different factors is a crucial step toward targeted data collection and model development and will finally clarify conditions that allow us to transfer loss models in space and time. Key Points: Survey data of flood‐affected households show different concurrent flood types, undermining the use of a single‐flood‐type loss modelThirteen variables addressing flood hazard, the building, and property level preparedness are significant predictors of the building loss ratioFlood type‐specific models show varying significance across the predictor variables, indicating a hindrance to model transferability [ABSTRACT FROM AUTHOR]

Flood damage processes are complex and vary between events and regions. State‐of‐the‐art flood loss models are often developed on the basis of empirical damage data from specific case studies and do not perform well when spatially and temporally transferred. This is due to the fact that such localized models often cover only a small set of possible damage processes from one event and a region. On the other hand, a single generalized model covering multiple events and different regions ignores the variability in damage processes across regions and events due to variables that are not explicitly accounted for individual households. We implement a hierarchical Bayesian approach to parameterize widely used depth‐damage functions resulting in a hierarchical (multilevel) Bayesian model (HBM) for flood loss estimation that accounts for spatiotemporal heterogeneity in damage processes. We test and prove the hypothesis that, in transfer scenarios, HBMs are superior compared to generalized and localized regression models. In order to improve loss predictions for regions and events for which no empirical damage data are available, we use variables pertaining to specific region‐ and event‐characteristics representing commonly available expert knowledge as group‐level predictors within the HBM. Key Points: Hierarchical Bayesian model (HBM) captures spatiotemporal variability in flood damage processes better than established depth‐damage functionsRegion‐ and event‐specific characteristics that explain the variability in damage processes improve spatiotemporal transfer of the HBMThe HBM is tested for transferability using empirical damage data from six events in the Elbe, Danube, Rhine and Oder catchments in Germany [ABSTRACT FROM AUTHOR]

Discharge from multiple wastewater treatment plants (WWTPs) distributed in urbanized river basins contributes to impairments of river water‐quality and aquatic ecosystem integrity, with size and location of WWTPs determined by population distribution within a river basin. Here we used geo‐referenced data for WWTPs in Germany to investigate the spatial organization of three attributes of interest in this study: population, population equivalents (the aggregated population served by each WWTP), and the number/sizes of WWTPs. To this end, we selected as case studies three large urbanized river basins (Weser, Elbe, and Rhine), home to about 70% of the population in Germany. We employed fractal river networks as structural platforms to examine the spatial patterns from two perspectives: spatial hierarchy (stream order) and patterns along longitudinal flow paths (width function). Moreover, we proposed three dimensionless scaling indices to quantify (1) human settlement preferences by stream order, (2) non‐sanitary flow contribution to total wastewater treated at WWTPs, and (3) degree of centralization in WWTPs locations. Across the three river basins, we found scale‐invariant distributions for each of the three attributes with stream order, quantified using extended Horton scaling ratios. We found a weak downstream clustering of population in the three basins. Variations in population equivalent clustering among different class‐sizes of WWTPs reflected the size, number, and locations of urban agglomerations in these river basins. We discussed the applicability of this approach to other large urbanized basins to analyze spatial organization of population and WWTPs. Key Points: Spatial hierarchies for population, population equivalents, and the number of wastewater treatment plants are scale‐invariantUnique patterns for cities size and number affect clustering location and degree for different class‐sizes of wastewater treatment plantsThree scaling indices characterize correlations in spatial patterns of population, population equivalents, and wastewater treatment plants [ABSTRACT FROM AUTHOR]

Abstract: Transformation and retention of nitrogen and other biologically reactive solutes in the hyporheic zones of running water contribute to an essential ecosystem service. However, the synoptic impact of intense agricultural or urban land‐uses, elevated nutrient loading, flow alterations, riparian clear‐cutting, and channelization on the source‐sink behavior of solutes in hyporheic zones remains largely uncharacterized and unquantified. Therefore, we studied nutrient dynamics in a hydromorphologically and chemically modified stream reach using a new monitoring approach allowing the simultaneous measurement of nutrient and water flux through a screened area in the subsurface of rivers (hyporheic passive flux meter, HPFM). With HPFMs we directly assessed time‐integrated lateral hyporheic nitrate fluxes during early spring and midsummer covering different temperature and discharge regimes. Contrary to our expectations, higher stream discharge coincided with substantially lower hyporheic exchange rates. While in streams featuring a natural morphology, bed form induced exchange commonly increases with surface flow, the influence of groundwater level was dominant in this reach. Furthermore, in contrast to less impacted environments, where progressive substrate depletion with depths reduces metabolic rates in the subsurface, we identified not the upper, but the intermediate layer of the hyporheic zone as hot spot of nutrient turnover. Overall, the hyporheic zone at the study site functioned partly as nitrate source, partly as a sink. Neither of the commonly used determinants redox state and residence time could explain this source or sink function. Our results give clear evidence to carefully transfer the knowledge of hyporheic zone processes from “natural” systems to anthropologically modified streams. [ABSTRACT FROM AUTHOR]

A quantitative understanding of groundwater-surface water interactions is vital for sustainable management of water quantity and quality. The noble gas radon-222 (Rn) is becoming increasingly used as a sensitive tracer to quantify groundwater discharge to wetlands, lakes, and rivers: a development driven by technical and methodological advances in Rn measurement. However, quantitative interpretation of these data is not trivial, and the methods used to date are based on the simplest solutions to the mass balance equation (e.g., first-order finite difference and inversion). Here we present a new implicit numerical model (FINIFLUX) based on finite elements for quantifying groundwater discharge to streams and rivers using Rn surveys at the reach scale (1-50 km). The model is coupled to a state-of-the-art parameter optimization code Parallel-PEST to iteratively solve the mass balance equation for groundwater discharge and hyporheic exchange. The major benefit of this model is that it is programed to be very simple to use, reduces nonuniqueness, and provides numerically stable estimates of groundwater fluxes and hyporheic residence times from field data. FINIFLUX was tested against an analytical solution and then implemented on two German rivers of differing magnitude, the Salzach (∼112 m3 s−1) and the Rote Main (∼4 m3 s−1). We show that using previous inversion techniques numerical instability can lead to physically impossible negative values, whereas the new model provides stable positive values for all scenarios. We hope that by making FINIFLUX freely available to the community that Rn might find wider application in quantifying groundwater discharge to streams and rivers and thus assist in a combined management of surface and groundwater systems. [ABSTRACT FROM AUTHOR]

We examined 3 years of measured daily values of all major water budget components (precipitation P, potential evapotranspiration PET, actual evapotranspiration ET, and runoff R) and volumetric soil water content θ of a small, forested catchment located in the west of Germany. The spatial distribution of θ was determined from a wireless sensor network of 109 points with 3 measurement depths each; ET was calculated from eddy-covariance tower measurements. The water budget was dominantly energy limited, with ET amounting to approximately 90% of PET, and a runoff ratio R/P of 56%. P, ET, and R closed the long-term water budget with a residual of 2% of precipitation. On the daily time scale, the residual of the water budget was larger than on the annual time scale, and explained to a moderate extent by θ ( R2 = 0.40). Wavelet analysis revealed subweekly time scales, presumably dominated by unaccounted fast-turnover storage terms such as interception, as a major source of uncertainty in water balance closure. At weekly resolution, soil water content explained more than half ( R2 = 0.62) of the residual. By means of combined empirical orthogonal function and cluster analysis, two slightly different spatial patterns of θ could be identified that were associated with mean θ values below and above 0.35 cm3/cm3, respectively. The timing of these patterns as well as the varying coherence between PET, ET, and soil water content responded to changes in water availability, including a moderate response to the European drought in spring 2011. [ABSTRACT FROM AUTHOR]

We investigate the usefulness of complex flood damage models for predicting relative damage to residential buildings in a spatial and temporal transfer context. We apply eight different flood damage models to predict relative building damage for five historic flood events in two different regions of Germany. Model complexity is measured in terms of the number of explanatory variables which varies from 1 variable up to 10 variables which are singled out from 28 candidate variables. Model validation is based on empirical damage data, whereas observation uncertainty is taken into consideration. The comparison of model predictive performance shows that additional explanatory variables besides the water depth improve the predictive capability in a spatial and temporal transfer context, i.e., when the models are transferred to different regions and different flood events. Concerning the trade-off between predictive capability and reliability the model structure seem more important than the number of explanatory variables. Among the models considered, the reliability of Bayesian network-based predictions in space-time transfer is larger than for the remaining models, and the uncertainties associated with damage predictions are reflected more completely. [ABSTRACT FROM AUTHOR]

The understanding of streamflow generation processes is vitally important in the management of water resources. In the absence of the data required to achieve this, Integrated Surface-Subsurface Hydrological Models (ISSHM) can be used to assist with the development of this understanding. However, the standard outputs from these models only enable elicitation of information about hydrological drivers and hydrological responses that occur at the same time. This generally limits the applicability of ISSHMs for the purposes of obtaining an improved understanding of streamflow generation processes to catchment areas that do not exhibit significant storage, travel times or flow depletion mechanisms. In order to overcome this limitation, a previously published Hydraulic Mixing-Cell (HMC) method is improved so that it can be used to follow surface water derived from direct rainfall and groundwater discharge to the stream and adjacent overland flow areas. The developed approach was applied to virtual experiments (based on the Lehstenbach catchment in southeastern Germany), which are composed of two ISSHMs of contrasting scales: (1) a riparian wetland of area 210 m2 and (2) a catchment of area 4.2 km2. For the two models, analysis of modeling results for a large storm event showed complex spatiotemporal variability in streamflow generation and surface water-groundwater interaction. Further analysis with the HMC method elucidated in-stream and overland flow generation mechanisms. This study showed within a modeling framework that identification and quantification of in-stream and overland flow generation better informed understanding of catchment functioning through decomposition of streamflow hydrographs, and analysis of spatiotemporal variability of flow generation mechanisms. [ABSTRACT FROM AUTHOR]

Riparian zones are important buffer zones for streams as they are hotspots of nitrate transformation and removal in agricultural catchments. However, mixing of water from different sources and various transformation processes can complicate the quantification of nitrate turnover in riparian zones. In this study, we analyzed nitrate concentration and isotope data in riparian groundwater along a 2‐km stream section in central Germany. We developed a mathematical model combining end‐member mixing and isotope modeling to account for mixing of river water and groundwater and quantify nitrate transformation in riparian groundwater. This enabled us to explicitly determine the extent of denitrification (as process leading to permanent nitrate removal from riparian groundwater) and transient nitrate removal by additional processes associated with negligible isotope fractionation (e.g., plant uptake and microbial assimilation) and to perform an extensive uncertainty analysis. Based on the nitrogen isotope data of nitrate, the simulations suggest a mean removal of up to 27% by additional processes and only about 12% by denitrification. Nitrate removal from riparian groundwater by additional processes exceeded denitrification particularly in winter and at larger distance from the river, underlining the role of the river as organic carbon source. This highlights that nitrate consumption by additional processes predominates at the field site, implying that a substantial fraction of agricultural nitrogen input is not permanently removed but rather retained in the riparian zone. Overall, our model represents a useful tool to better compare nitrogen retention to permanent nitrogen removal in riparian zones at various temporal and spatial scales. Plain Language Summary: Nitrogen is an important nutrient for agricultural crops. However, excessive nitrogen input into surface water in the form of nitrate can lead to algae blooms and lack of oxygen. The riparian zones of rivers are important buffer zones where groundwater is connected to soils, which are rich in soil organisms and organic matter pools fueling reaction processes. Hence, plants and bacteria can remove nitrate from riparian groundwater before it reaches the river. Bacterial consumption of nitrate (denitrification) leads to complete removal of nitrogen via release of nitrogen gas into the atmosphere. In contrast, other biogeochemical processes such as nitrate uptake by plants merely result in nitrogen retention within riparian zones. To quantify the role of denitrification relative to other processes, we developed a novel model combining concentration and isotope data of nitrate and applied it to a groundwater study site in Central Germany. We found that nitrate removal from riparian groundwater by additional processes largely exceeded denitrification. Hence, a major fraction of nitrogen inputs was retained in the riparian zone and may eventually end up in the river. Such information is highly relevant for many river ecosystems at risk of eutrophication because of high nitrogen inputs from agriculture. Key Points: We present a model using concentration and isotope data to distinguish riparian denitrification from additional nitrate removal processesThe model was applied to concentration and dual‐element isotope data of nitrate from riparian groundwater wellsNitrate removal by additional processes greatly exceeded denitrification, particularly at larger distance from the river and in winter [ABSTRACT FROM AUTHOR]