Abstract. Temperate regions across Europe, such as Denmark, are projected to be subjected to substantial changes in the hydrological cycle due to climate change. Changes in climate can materialize as general changes in long term means, extremes, or in seasonal patterns, e.g., dampening or intensification of the seasonal contrasts. Changes in seasonal patterns can affect the hydrological cycle in various ways, due to the interlinkage between hydrological compartments. To detect, track and quantify the impact of changes in climate and seasonal patterns, integrated hydrological modelling is needed. This makes Denmark an ideal test case due to the established integrated and physically based National Hydrological Model of Denmark (DK-model). Utilizing climate projections from 17 RCP8.5 climate models, downscaled and bias-corrected for Denmark, we calculate climate change impacts on both overall values and seasonality for the variables soil moisture, streamflow, shallow and deeper groundwater to the end of the century. Moreover, standardized hydrological drought indices are calculated for the same variables. Climate change projections point towards a future with higher annual precipitation, mainly due to wetter winters, while climatic water balance deficits increase during summer; thus, intensifying the seasonal contrast. The increased contrast is reappearing in the fast-responding hydrological variable, soil moisture; while streamflow and shallow groundwater clearly reproduce increase during the wetter winter, the summer signal differs. The deep groundwater systems experience higher future groundwater heads across the entire year. Common for all variables is a larger seasonality, defined as contrast between intra-annual low and highs. Notably, the ensemble of hydrological projections is more in agreement regarding the seasonal contrast than on the direction of absolute change, with results agreeing for 85 % to 99 % of the area of Denmark on increased seasonality, whereas only agreeing for 50 % to 98 % on the absolute direction of that change. The drought indices exhibit a similar seasonal change, with more droughts during summer and more wet anomalies during winter for soil moisture, while summer droughts for streamflow and shallow groundwater partially are buffered by wetter winters and the related recharge increase. In summary, the results indicate that despite considerable increases in precipitation, projected climate change for Denmark is expected to enhance hydrological seasonality instead of producing a uniform transition to a wetter regime, potentially impacting the climate adaptation and mitigation effort, agricultural yields, and water supply.

This method article presents a local refinement framework for a national-scale, machine learning-based groundwater model that predicts typical summer and winter water table depth at 10 × 10 m resolution at national scale of Denmark. While the existing baseline model provides high-resolution national coverage and is suitable for screening purposes, its accuracy remains insufficient for local groundwater management applications. The proposed method integrates new groundwater observations into the existing training dataset and retrains the baseline model to enhance local fidelity while preserving large-scale consistency. The approach is demonstrated for a 212 km² case study area using 100 synthetic groundwater observations. Cross-validation results show consistent improvements in mean error, mean absolute error, and root mean squared error compared to the baseline model, particularly when increased weights are assigned to new observations. The refined models produce improved precision in areas with new observations while maintaining baseline behaviour elsewhere, demonstrating its suitability for combining national datasets with local monitoring networks for improved decision support in climate adaptation, nature conservation and other applications.

Most operational flood forecasting systems provide predictions of pluvial and fluvial floods, often neglecting groundwater flooding. Groundwater-induced floods can occur when prolonged rainfall, high river stages or elevated sea levels raise the groundwater table above the surface of the land, often occurring in low-lying areas or areas with specific soil and land-surface conditions. This study presents an operational, national-scale, integrated flood forecasting system that combines surface water and groundwater components – such as river discharge and high groundwater levels – to assess flood risk in Denmark. The system has been proven to effectively capture peak river flows and elevated groundwater levels, as it did across the country during the winter of 2024, and provide local-scale insights, as exemplified during a specific flood event in Varde, west Denmark. This study demonstrates how groundwater flooding, often neglected in operational forecasting, can be effectively incorporated at a national scale to support more informed flood management.

Uncertainties in hydrological models arise primarily from the subsurface architecture and hydrological properties of each unit. This paper presents a method to simulate structural uncertainties in large-scale hydrostratigraphic or geological models by generating multiple realizations constrained by quantified uncertainties. The simulation is based on the geology-driven modeling (GDM) method. GDM requires quantified uncertainty estimates for each interpretation point and assumes that the model domain is sufficiently small for stationarity to hold. However, the stationarity assumption does not hold at the national scale. To address this limitation, an upscaling method for GDM is proposed, which subdivides the domain into regions with similar structural properties. These regions can be simulated independently using Gaussian simulation, then combined and layer-corrected to produce a cohesive three-dimensional model ensemble. This approach is computationally efficient, flexible, and scalable, making national-scale modeling feasible. The upscaled version, GDM-National, is introduced, and its effectiveness is demonstrated by generating an ensemble of realizations for Denmark’s national hydrostratigraphic layer model. The method provides a valuable tool for propagating interpretation uncertainties from hydrostratigraphic layers to hydrological models.

Accurate water level predictions are increasingly crucial for mitigating flood risks. Hydrological and hydrodynamic models provide water level predictions, but their accuracy depends on detailed information about stream cross-sections and floodplain topography, which are data that are difficult to obtain at larger scale, especially in regions with perennial river systems. Stream discharge is a variable that is more straightforward to predict by conventional hydrological models. However, the relationship between discharge and water level is complex, depending on cross-section geometry and channel roughness. Here machine learning models offer an alternative opportunity to predict water level by ingesting readily available topographic data derived from high-resolution digital elevation models in combination with simulated stream discharge, thereby skipping the need to explicitly define rating curves or to run complex hydrodynamic simulations. The idea is that stream discharge provides information about the temporal variability, whereas the topographic data provides static information in the cross-section geometry. First, we present a method for extracting stream geomorphometry from a high-resolution (40 cm) digital elevation model in Denmark. The methodology is based on analyzing elevation changes along cross-sections throughout the entire Danish river network. Stream widths are estimated by identifying the most probable bank positions through a probabilistic count of all possible configurations within 100-meter stream reaches. The resulting dataset has been validated against 2,000 measured cross-sections along Danish rivers, showing similar spatial patterns across reach to river scales. Moreover, the slope and elevation of the water level as well as channel area and depth are derived from the high-resolution DEM for 100-meter stream reaches.Second, we present the development of a machine learning-based model that utilizes the derived stream geomorphometry in combination with stream discharge simulated by the National Hydrological Model of Denmark to predict daily stream water levels. Timeseries of daily stream water level of 40 gauging stations are used to train a Long Short Term Memory network. The results demonstrate that incorporating topography-derived information of mean water level and slope, stream channel width, area, and depth, enhance the accuracy of the water level estimates. Overall, our approach provides a versatile approach providing crucial information on flood risks that can easily be scaled up to national scale.

Redox conditions play a crucial role in determining the fate of geogenic and anthropogenic contaminants in groundwater, impacting ecosystem services vital for both the aquatic environment and human water supply. For example, investigating the reduction of nitrate underscores the importance of data on redox conditions since denitrification takes places in anoxic environments. Specifically, knowledge of the depth to the uppermost reduced layer, i.e., first redox interface, can inform water and land management by identifying agricultural areas vulnerable or robust to nitrate leaching. Assessing redox processes is complicated by geological heterogeneities, resulting in complexities of local to regional groundwater flow paths. Geospatial machine learning techniques have previously successfully mapped redox conditions based on sediment color or water chemistry observations. This study introduces a novel approach that combines both data sources to enhance understanding of subsurface redox conditions in Denmark. In the first step, depth to the first redox interface is mapped using sediment color information from 26,800 boreholes. This depth is derived from sediment color changes, transitioning from oxic to reduced colors of quaternary sediments. The mapping utilizes a regression-based gradient boosting with decision tree algorithm trained against sediment color data and 20 covariates, encompassing information on hydrogeology, lithology, topography, and hydrology. In the second step, the depth of the first redox interface is compared against groundwater chemistry to classify continuous and discontinuous redox conditions. Continuous conditions exhibit the absence of oxic groundwater below the first redox interface, while discontinuous conditions show oxic groundwater below the interface. This classification is performed using a gradient boosting with decision tree algorithm utilizing the same 20 covariate maps and 21,800 classified groundwater samples. Both models undergo comprehensive cross-validation and feature importance analysis. The depth to the first redox interface is modeled with a mean error of 0.001 m and a root-mean-squared error of 8.3 m. The continuous/discontinuous classification attains an accuracy of 69.5 %. Both variables are mapped at a 25 m spatial resolution at the national scale of Denmark. Results indicate a mean depth to the first redox interface of 9.4 m and a standard deviation of 5.7 m, with spatial patterns largely driven by the groundwater table. 66.0% of Denmark is classified as discontinuous, indicating complex redox conditions, predominantly collocated with moraine clay. These maps contribute significantly to understanding subsurface redox processes, supporting national-scale land and water management.

The National Hydrological Model for Denmark (DK-model) is a distributed, integrated hydrological model coupling 3D groundwater flow to descriptions of root zone processes, overland flow and river routing, including anthropogenic interference with the hydrological cycle. It covers all of Denmark (~43,000km2) at 500m and 100m grid scale. Its constant development over the last three decades has both been driven by research projects and projects for public authorities. It is being used for various tasks such as water resource assessments, climate change impact assessments, hydrological real-time monitoring and nutrient transport studies. Recently, we endeavored novel ways to calibrate and parameterize the DK-model. The model is placed on the edge between research interest and practical applications, with a demand for adequately representing various aspects of the hydrological cycle across the entirety of the model domain. In combination with its large-scale distributed nature and high computational demand, conventional (groundwater) model optimization techniques are challenged: The complex nature and versatile applications of the DK-model require suitable parametrization schemes and inclusion of diverse calibration and evaluation data, beyond conventional groundwater head observations and streamflow. This also leads to trade-offs between the multiple objective functions. Hence, we moved beyond previously used single solution, gradient-based optimization algorithms. The Pareto Archived Dynamically Dimensioned Search (PADDS) algorithm allows us to use a global parameter optimization, effective even at a few hundred model runs. Another major advantage of PADDS is that it does not require the a-priori weighting of objective function groups – instead, it explores the tradeoffs (pareto front) between the different objective function groups, allowing weighting after gaining knowledge about tradeoffs during the optimization process. Also, all solutions explored during the optimization are stored and remain open to analysis after finished optimization. This not only sheds light on tradeoffs between different objective functions in a unique manner, but also supports understanding of parameter sensitivity and uncertainty in a manner which otherwise is hard to achieve due to computational constraints. Moreover, we included evapotranspiration patterns from satellite products as well as a machine learning based estimate of artificial drain flow as novel spatial data in the model evaluation. This helps us constraining some of the model processes crucial for e.g. nutrient transport, but otherwise poorly constrained by conventional data such as streamflow (305 stations) and groundwater heads (24,000 wells) covering practically the entire model domain. We explored the benefits of this optimization setup applied to the DK-model, advancing not only the calibration process itself, but also our understanding of model process representation and performance.

Large-scale datasets of hydrometeorological time series and catchment attributes are essential for advancing the understanding of hydrological processes, advancing hydrological model development, and enabling performance benchmarking. CAMELS (Catchment Attributes and Meteorology for Large-sample Studies) datasets have already been published for various regions worldwide covering a wide range of hydrometeorology and physiography. We introduce a CAMELS-style dataset for Denmark (CAMELS-DK) containing predominantly lowland, groundwater-influenced, and small-sized catchments. With respect to already published CAMELS datasets, we see this as a valuable extension that enlarges the variability of catchments. Moreover, this is the first CAMELS dataset to include both gauged and ungauged catchments as well as detailed groundwater information. CAMELS-DK comprises dynamic and static variables for 3,330 catchments across Denmark, derived from diverse hydrogeological datasets, meteorological observations, and simulated variables provided by the National Hydrological Model of Denmark. From the latter, a comprehensive list of simulated groundwater related variables like phreatic depth or groundwater-surface water interactions, are included. Streamflow observations are available for 304 catchments, while simulated streamflow data are provided for a total of 3,330 catchments. The dataset spans 30 years (1989–2019) at a daily temporal resolution. Additionally, the dataset includes variables capturing human impacts on Denmark's water resources, such as groundwater abstraction and irrigation.By providing streamflow at almost full spatial coverage of Denmark, and not being limited to gauged sites, along with various simulation outputs from a distributed, process-based hydrological model, CAMELS-DK significantly enhances the utility of CAMELS datasets. This includes supporting the development of data-driven and hybrid/physically informed modeling frameworks.The dataset is accessible via Koch et al. (2024) and the paper describing the dataset is currently under review (Liu et al., 2024).Koch, J., Liu, J., Stisen, S., Troldborg, L., Højberg, A. L., Thodsen, H., Hansen, M. F. T., and Schneider, R. J. M.: CAMELSDK: Hydrometeorological Time Series and Landscape Attributes for 3330 Catchments in Denmark, https://doi.org/doi:10.22008/FK2/AZXSYP.Liu, J., Koch, J., Stisen, S., Troldborg, L., Højberg, A. L., Thodsen, H., Hansen, M. F. T., and Schneider, R. J. M.: CAMELS-DK: Hydrometeorological Time Series and Landscape Attributes for 3330 Catchments in Denmark, Earth Syst. Sci. Data Discuss. [preprint], https://doi.org/10.5194/essd-2024-292, in review, 2024.

Large samples of hydrometeorological time series and catchment attributes are critical for improving the understanding of complex hydrological processes, hydrological model development, and performance benchmarking. CAMELS (Catchment Attributes and Meteorology for Large-sample Studies) datasets have been developed in several countries and regions around the world, providing valuable data sources and test beds for hydrological analysis and new frontiers in data-driven hydrological modeling. Regarding the lack of samples from lowland, groundwater-dominated, small-sized catchments, we develop an extensive repository of a CAMELS-style dataset for Denmark (CAMELS-DK). This CAMELS addition is the first containing both gauged and ungauged catchments as well as detailed groundwater information. The dataset provides dynamic and static variables for 3330 catchments covering all of Denmark from various hydrogeological datasets, meteorological observations, and a well-established national-scale hydrological model. For 304 of those catchments, streamflow observations are provided, whereas simulated streamflow is provided for all 3330 catchments. The dataset contains time series spanning 30 years (1989–2019) with a daily time step, and the data will be updated once new observations and model simulations become available. The dense and full spatial coverage for all 3330 catchments, instead of only gauged catchments, together with the addition of various simulation data from a distributed, process-based model, enhances the applicability of such CAMELS data, for example, for the development of data-driven and hybrid physically informed modeling frameworks or other cases where consistent full spatial coverage is required. We also provide quantities related to the human impact on the hydrological system in Denmark, such as groundwater abstraction and irrigation. The CAMELS-DK dataset is freely available at https://doi.org/10.22008/FK2/AZXSYP (Koch et al., 2024).

Redox conditions play a crucial role in determining the fate of many contaminants in groundwater, impacting ecosystem services vital for both the aquatic environment and human water supply. Geospatial machine learning has previously successfully modelled large-scale redox conditions. This study is the first to consolidate the complementary information provided by sediment color and water chemistry to enhance our understanding of redox conditions in Denmark. In the first step, the depth to the first redox interface is modelled using sediment color from 27,042 boreholes. In the second step, the depth of the first redox interface is compared against water chemistry data at 22,198 wells to classify redox complexity. The absence of nitrate containing water below the first redox interface is referred to as continuous redox conditions. In contrast, discontinuous redox conditions are identified by the presence of nitrate below the first redox interface. Both models are built using 20 covariate maps, encompassing diverse hydrologically relevant information. The first redox interface is modelled with a mean error of 0.0 m and a root-mean-squared error of 8.0 m. The redox complexity model attains an accuracy of 69.8 %. Results indicate a mean depth to the first redox interface of 8.6 m and a standard deviation of 6.5 m. 60 % of Denmark is classified as discontinuous, indicating complex redox conditions, predominantly collocated in clay rich glacial landscapes. Both maps, i.e., first redox interface and redox complexity are largely driven by the water table and hydrogeology. The developed maps contribute to our understanding of subsurface redox processes, supporting national-scale land-use and water management.

Groundwater constitutes 99% of the Earth’s liquid freshwater and is crucial for human health, economic development, and ecosystem sustainability. This study assesses groundwater sustainability in Denmark by employing a comprehensive hydrological model and a set of ensemble indicators. The paper describes the methodology and the results based on nine selected indicators. Three indicators focus on recharge capture and aquifer sustainability, one focuses on groundwater level and wetland capture, two focus on baseflow and drainage flow capture, and three focus on eco flow capture. Our findings highlight that while overall exploitable groundwater resources are estimated at 1.1 billion m3/year, significant regional disparities exist, with certain areas, notably Zealand, facing over-exploitation rates exceeding 250% of sustainable limits. The indicators developed not only provide a framework for assessing current groundwater resource limits, but also serve as a basis for future monitoring and adaptive management strategies. This research underscores the need for stakeholder engagement and integrated approaches to ensure the sustainability of groundwater resources in the face of growing anthropogenic pressures and climate change. Our work contributes to the ongoing discourse on sustainable water management and offers a robust methodology for assessing groundwater sustainability.

Denmark's complete reliance on groundwater for water supply presents a unique case study in management of natural resources, urban planning, and water resilience in the face of climate change. This paper examines the groundwater management strategies in Denmark in general, focusing on Denmark's four largest cities—Copenhagen, Aarhus, Odense, and Aalborg— each facing distinct challenges due to their demographic, geographical, hydrogeological, and economic characteristics. Through analysis of these cities' approaches to groundwater management, this research contributes to the global discourse on sustainable urban water supply systems. As coastal groundwater cities (CGC), these urban areas must navigate the complexities of sustaining growing populations, mitigating climate change impacts, and coastal processes while ensuring the long-term viability of their groundwater resources. Copenhagen and Aalborg, built atop semi-confined fractured and locally karstic carbonate rocks, highlights the specific challenges associated with karstic groundwater systems, while, Aarhus, and Odense built on glaciofluvial aquifers faces different issues. The different groundwater challenges in these cities underscores the importance of integrating urban development with water resource management and environmental sustainability, offering valuable insights and lessons learned for other regions facing similar challenges. This study, thus not only sheds light on Denmark's groundwater management practices, but also emphasizes the need for innovative solutions to ensure the resilience of urban water supply systems in a changing climate and increasing pressures of emerging organic contaminants and elevated concentrations of geogenic elements induced by water abstraction and fluctuating water tables. Advanced Danish monitoring and modelling tools applied to support decision-making and innovation within the water sector are continuously developed and improved to support resilient and sustainable management of the available water resources.

Abstract. Large samples of hydrometeorological time series and catchment attributes are critical for improving the understanding of complex hydrological processes, hydrological model development and performance benchmarking. CAMELS (Catchment Attributes and Meteorological time series for Large Samples) datasets have been developed in several countries and regions around the world, providing valuable data sources and testbeds for hydrological analysis and new frontiers in data-driven hydrological modelling. Regarding the lack of samples from low-land, groundwater-dominated, small-sized catchments, we develop an extensive repository of a CAMELS-style dataset for Denmark (CAMELS-DK). This CAMELS addition is the first containing both, gauged and ungauged catchments as well as detailed groundwater information. The dataset provides dynamic and static variables for 3330 catchments from various hydrogeological datasets, meteorological observations, and a well-established national-scale hydrological model. The dataset is enhanced with streamflow observations in 304 of those catchments. The spatially dense and full spatial coverage, supplying data for 3330 catchments, instead of only gauged catchments, together with the addition of simulation data from a distributed, process-based model enhance the applicability of such CAMELS data. This is especially relevant for the development of data-driven and hybrid physical informed modelling frameworks. We also provide quantities related to human impact on the hydrological system in Denmark, such as groundwater abstraction and irrigation. The CAMELS-DK dataset is freely available at https://doi.org/10.22008/FK2/AZXSYP (Koch et al., 2024).

Accurate streamflow estimation is essential for effective water resource management and adapting to extreme events in the face of changing climate conditions. Hydrological models have been the conventional approach for streamflow interpolation and extrapolation in time and space for the past few decades. However, their large-scale applications have encountered challenges, including issues related to efficiency, complex parameterization, and constrained performance. Deep learning methods, such as long short-term memory (LSTM) networks, have emerged as a promising and efficient approach for large-scale streamflow estimation. In this study, we have conducted a series of experiments to identify optimal hybrid modeling schemes to consolidate physically based models with LSTM aimed at enhancing streamflow estimation in Denmark. The results show that the hybrid modeling schemes outperformed the Danish National Water Resources Model (DKM) in both gauged and ungauged basins. While the standalone LSTM rainfall–runoff model outperformed DKM in many basins, it faced challenges when predicting the streamflow in groundwater-dependent catchments. A serial hybrid modeling scheme (LSTM-q), which used DKM outputs and climate forcings as dynamic inputs for LSTM training, demonstrated higher performance. LSTM-q improved the mean Nash–Sutcliffe efficiency (NSE) by 0.22 in gauged basins and 0.12 in ungauged basins compared to DKM. Similar accuracy improvements were achieved with alternative hybrid schemes, i.e., by predicting the residuals between DKM-simulated streamflow and observations using LSTM. Moreover, the developed hybrid models enhanced the accuracy of extreme events, which encourages the integration of hybrid models within an operational forecasting framework. This study highlights the advantages of synergizing existing physically based hydrological models (PBMs) with LSTM models, and the proposed hybrid schemes hold the potential to achieve high-quality large-scale streamflow estimations.

<p>Operational forecasts of groundwater levels provide critical real-time knowledge during extreme events, such as floods and droughts. This study proposes a Long Short-Term Memory based Encoder-Decoder (LSTM-ED) model for multi-step-ahead groundwater level forecasting. The LSTM-ED is a well-suited architecture for sequence-to-sequence modelling tasks but has not yet been applied to forecast groundwater levels. The proposed LSTM-ED model is designed in the context of the Danish online monitoring system <em>grundvandsstanden.dk</em> to serve as operational groundwater level forecasting system. In the encoder LSTM model, sequences of past precipitation, temperature and groundwater levels are processed to initialize the decoder LSTM model which, in addition takes in forecast sequences of precipitation and temperature to output a sequence of groundwater levels. We train LSTM-ED models individually for each well, with all data aggregated to daily timescale. We demonstrate the performance of the LSTM-ED architecture for numerous wells from <em>grundvandsstanden.dk</em> and test varying lead times of up to 30 days. The LSTM-ED model forecasts are contrasted with simple benchmark models as well as with a sequence-to-sequence LSTM model that does not incorporate forecasts of precipitation and temperature for outputting the groundwater sequence. Initial results underpin that integrating forecasts of precipitation and temperature is a crucial component, especially for wells with shallow intakes where surface and sub-surface processes are well connected. The sequence-to-sequence LSTM model yields similar accuracy as the simple benchmark models, whereas accuracy clearly improves for the LSTM-ED model. Overall, this study highlights the potential of LSTM-ED models as an operational tool for multi-step-ahead forecasting of groundwater levels.</p>

Abstract Chalk and limestone aquifers contribute one-third of the drinking water supply in Denmark, and one-sixth of that national groundwater resource is assessed as having a ‘Poor’ status in terms of the quantitative Water Framework Directive due to intensive abstraction. This paper describes the national groundwater level monitoring network with regard to the following three applications: (1) when used for the annual surveying and reporting of groundwater resources and impacts from climate and groundwater abstraction; (2) as part of real-time monitoring and modelling for daily and seasonal forecasting; and (3) for tracking long-term climate change impacts on groundwater levels. Groundwater level monitoring provides a particularly important indicator of abstraction pressure and sustainable balance compared with recharge. Many larger chalk and limestone groundwater bodies in Denmark are only monitored by local water companies and not represented in the national groundwater level network. This raises the concern that current national groundwater level monitoring does not fully support integrated modelling and assessment purposes for chalk and limestone groundwater bodies. This also implies that, for tracking long-term climate change and anthropogenic impacts on groundwater levels, the national groundwater monitoring network especially lacks long-term records with complete 30-year time-series for many intensively exploited large chalk and limestone aquifers.

Groundwater dating by radioactive cosmogenic tracers such as 39Ar relies on the decay rate from a known initial atmospheric activity (100%modern). Thereby, it is assumed that cosmogenic 39Ar production in the subsurface is negligible at depths below the water table and that contributions from natural rock radioactivity are minor or missing. Here we present 39Ar data from aquifers located in quaternary glacial sediments and tertiary limestones in Denmark, which unequivocally demonstrate that cosmogenic production can induce considerable age biases. 39Ar values larger than 100%modern are observed at relatively shallow groundwater depths in non-radiogenic rocks. These activities are compared to calculations based on previously assessed depth-dependent production rates in rocks and realistic estimates of the emanated fractions to the water phase. The water residence time distribution with depth, which was determined by numerical flow modeling and particle tracking, underpinned the significance of muon-induced 39Ar production. The short-lived isotope 37Ar is produced by similar processes as 39Ar and demonstrated its usefulness as an indicator of local underground production in an aquifer. The significance of cosmogenic underground production in other possible recharge scenarios was then assessed by explicitly simulating the radioargon accumulation and decay in a 2D synthetical numerical model. These simulations demonstrated that underground production is negligible when the water infiltrates freely in a porous aquifer. However, in the presence of a confining layer impeding the infiltration at shallow depths (<30 m), as is the case in our study site in Denmark for instance, over-modern 39Ar activities (>100%modern) may occur. The age concluded from the dissolved activities is then possibly biased towards young values. Special attention should thus be paid to the recharge rates when using 39Ar for dating groundwater. 37Ar activities provide complementary information about the strength and mechanisms of underground production.

In this study, we upscale and simplify hydrostratigraphic information from a detailed model for Denmark to a pan-European scale. This is part of a larger project to develop a harmonised overview of the volume and depth of groundwater resources in a quasi-3D European groundwater resource model. A 10 km grid and a maximum of c. 10 hydrostratigraphic layers were chosen as the common scale for the European database. The Danish information is based on the national water resources model (the DK-model), where the information is significantly more detailed (100 m grid and up to 26 layers). Information was transferred from the DK-model to the quasi-3D model by a method involving computations of mean volumes and expert assessment to reduce layers in each cell. In this process, detailed hydrostratigraphic information is lost, which could otherwise be used for local groundwater flow modelling in Denmark. However, the strength of the quasi-3D model is that it still contains the volumes of all hydrostratigraphic units, both the saturated and unsaturated parts. Hence, the upscaled model can contribute to a relatively precise calculation of European groundwater resources for the quantitative assessment of groundwater status across Europe at a 10 × 10 km scale.

Post audits of hydrological or groundwater models are the last part of the modelling protocol, where the original model predictions are tested using new data obtained after a certain period. The evaluation of model predictions and associated predictive uncertainty was performed by comparing an original hydrological model, a model with post audited geology, and a model with post audited geology and calibrated against new types of observation data. The post audit showed original model predictions close to what was observed (in terms of abstracted volumes necessary to lower a shallow groundwater table). In contrast to the robust original model predictions, the original model underestimated the predictive uncertainty compared to the assessments of uncertainty using the new and updated post audit model. To ensure a robust model evaluation, we propose a four-step post audit protocol, including (1) testing the validity of the original model predictions with new data, (2) estimating the predictive uncertainty of the original model, (3) producing a new post audit model(s) based on revising the conceptual model and calibration, and (4) assessing the predictive uncertainty of the new post audit models. The work presented here was motivated by the lack of studies that, after a certain time, have re-evaluated model predictions (post audit) with new data.

The paper analyzes the national DK-model hydrological information and prediction (HIP) system and HIP portal viewed as a ‘digital twin’ and how the introduction of real-time dynamic updating of the DK-model HIP simulations can make room for plug-in submodels with real-time boundary conditions made available from an HIP portal. The possible feedback to a national real-time risk knowledge base during extreme events (flooding and drought) is also discussed. Under climate change conditions, Denmark is likely to experience more rain in winter, more evapotranspiration in summer, intensified cloudbursts, drought, and sea level rise. These challenges were addressed as part of the Joint Governmental Digitalization Strategy 2016–2020 for better use and sharing of public data about the terrain, water, and climate to support climate adaptation, water management, and disaster risk reduction. This initiative included the development of a new web-based data portal (HIP portal) developed by the Danish Agency for Data Supply and Infrastructure (SDFI). GEUS delivered 5 terabytes of hydrological model data to the portal, with robust calibration methods and hybrid machine learning (ML) being key parts of the deliverables. This paper discusses the challenges and potentials of further developing the HIP digital twin with ‘plug-in digital twins’ for local river basins, including feedback to the national level.