Explore the vast range of titles available.

Visual analysis of time series in hydrology is frequently seen as a crucial step to becoming acquainted with the nature of the data, as well as detecting unexpected errors, biases, etc. Human eyes, in particular those of a trained expert, are well suited to recognize irregularities and distinct patterns. However, there are limits as to what the eye can resolve and process; moreover, visual analysis is by definition subjective and has low reproducibility. Visual inspection is frequently mentioned in publications, but rarely described in detail, even though it may have significantly affected decisions made in the process of performing the underlying study. This paper presents a visual analysis of groundwater hydrographs that has been performed in relation to attempts to classify groundwater time series as part of developing a new concept for prediction in data-scarce groundwater systems. Within this concept, determining the similarity of groundwater hydrographs is essential. As standard approaches for similarity analysis of groundwater hydrographs do not yet exist, different approaches were developed and tested. This provided the opportunity to carry out a comparison between visual analysis and formal, automated classification approaches. The presented visual classification was carried out on two sets of time series from central Europe and Fennoscandia. It is explained why and where visual classification can be beneficial but also where the limitations and challenges associated with the approach lie. It is concluded that systematic visual analysis of time series in hydrology, despite its subjectivity and low reproducibility, should receive much more attention.

A new concept is proposed for describing, analysing and predicting the dynamic behaviour of groundwater resources based on classification and similarity. The concept makes use of the ideas put forward by the “PUB” (predictions in ungauged basins) initiative in surface-water hydrology. One of the approaches developed in PUB uses the principle that similar catchments, exposed to similar weather conditions, will generate a similar discharge response at the catchment outlet. This way, models developed for well-observed catchments can be used to make predictions for ungauged catchments with similar properties (topography, land use, etc.). The concept proposed here applies the same idea to groundwater systems, with the goal to make predictions of the dynamic behaviour of groundwater in poorly observed systems using similarities to well-observed and understood systems. This paper gives an overview of the main ideas, the methodological background, the progress so far, and the challenges that the authors regard as most crucial for further development. One of the main goals of this article is thus to raise interest for this new concept within the groundwater community. There are a multitude of highly interesting aspects to investigate, and a community effort, as with PUB, is required. A second goal is to foster and exchange ideas between the groundwater and surface water research communities who, while often working on similar problems, have often missed the opportunity to learn from each other.

This study aims to identify common hydrogeological patterns and to gain a deeper understanding of the underlying similarities and their link to physiographic, climatic, and anthropogenic controls of coastal groundwater. The most striking aspects of GWL dynamics and their controls were identified through a combination of statistical metrics, calculated from about 8,000 groundwater hydrographs, and pattern recognition, classification, and explanation using machine learning techniques and SHapley Additive exPlanations (SHAP). Overall, four different GWL dynamics patterns emerge, independent of the different seasons, time series lengths, and periods. We show in this study that similar GWL dynamics can be observed around the world with different combinations of site characteristics, but also that the main factors differentiating these patterns can be identified. Three of the identified patterns exhibit high short-term and interannual variability and are most common in regions with low terrain elevation and shallow groundwater depth. Climate and soil characteristics are most important in differentiating these patterns. This study provides new insights into the hydrogeological behavior of groundwater in coastal regions and guides systematic and holistic groundwater monitoring and modelling, motivating to consider various aspects of GWL dynamics when, for example, estimating climate-driven GWL changes – especially when information on potential controls is limited.

Groundwater level (GWL) dynamics result from a complex interplay between groundwater systems and the Earth system. This study aims to identify common hydrogeological patterns and to gain a deeper understanding of the underlying similarities and their link to physiographic, climatic, and anthropogenic controls of groundwater in coastal regions. The most striking aspects of GWL dynamics and their controls were identified through a combination of statistical metrics, calculated from about 8000 groundwater hydrographs, pattern recognition using clustering algorithms, classification using random forest, and SHapley Additive exPlanations (SHAPs). Hydrogeological similarity was defined by four clusters representing distinct patterns of GWL dynamics. These clusters can be observed globally across different continents and climate zones but simultaneously vary regionally and locally, suggesting a complicated interplay of controlling factors. The main controls differentiating GWL dynamics were identified, but we also provide evidence for the currently limited ability to explain GWL dynamics on large spatial scales, which we attribute mainly to uncertainties in the explanatory data. Finally, this study provides guidance for systematic and holistic groundwater monitoring and modeling and motivates a consideration of the different aspects of GWL dynamics, for example, when predicting climate-induced GWL changes, and the use of explainable machine learning techniques to deal with GWL complexity – especially when information on potential controls is limited or needs to be verified.

We present a method for risk assessment of groundwater drawdown induced land subsidence when planning for sub-surface infrastructure. Since groundwater drawdown and related subsidence can occur at large distances from the points of inflow, the large spatial extent often implies heterogeneous geological conditions that cannot be described in complete detail. This calls for estimation of uncertainties in all components of the cause-effect chain with probabilistic methods. In this study, we couple four probabilistic methods into a comprehensive model for economic risk quantification: a geostatistical soil-stratification model, an inverse calibrated groundwater model, an elasto-plastic subsidence model, and a model describing the resulting damages and costs on individual buildings and constructions. Groundwater head measurements, hydraulic tests, statistical analyses of stratification and soil properties and an inventory of buildings are inputs to the models. In the coupled method, different design alternatives for risk reduction measures are evaluated. Integration of probabilities and damage costs result in an economic risk estimate for each alternative. Compared with the risk for a reference alternative, the best prior alternative is identified as the alternative with the highest expected net benefit. The results include spatial probabilistic risk estimates for each alternative where areas with significant risk are distinguished from low-risk areas. The efficiency and usefulness of this modelling approach as a tool for communication to stakeholders, decision support for prioritization of risk reducing measures, and identification of the need for further investigations and monitoring are demonstrated with a case study of a planned railway tunnel in Varberg, Sweden.

This paper presents the results of the 2022 Groundwater Time Series Modelling Challenge, where 15 teams from different institutes applied various data-driven models to simulate hydraulic-head time series at four monitoring wells. Three of the wells were located in Europe and one was located in the USA in different hydrogeological settings in temperate, continental, or subarctic climates. Participants were provided with approximately 15 years of measured heads at (almost) regular time intervals and daily measurements of weather data starting some 10 years prior to the first head measurements and extending around 5 years after the last head measurement. The participants were asked to simulate the measured heads (the calibration period), to provide a prediction for around 5 years after the last measurement (the validation period for which weather data were provided but not head measurements), and to include an uncertainty estimate. Three different groups of models were identified among the submissions: lumped-parameter models (three teams), machine learning models (four teams), and deep learning models (eight teams). Lumped-parameter models apply relatively simple response functions with few parameters, while the artificial intelligence models used models of varying complexity, generally with more parameters and more input, including input engineered from the provided data (e.g. multi-day averages). The models were evaluated on their performance in simulating the heads in the calibration period and in predicting the heads in the validation period. Different metrics were used to assess performance, including metrics for average relative fit, average absolute fit, fit of extreme (high or low) heads, and the coverage of the uncertainty interval. For all wells, reasonable performance was obtained by at least one team from each of the three groups. However, the performance was not consistent across submissions within each group, which implies that the application of each method to individual sites requires significant effort and experience. In particular, estimates of the uncertainty interval varied widely between teams, although some teams submitted confidence intervals rather than prediction intervals. There was not one team, let alone one method, that performed best for all wells and all performance metrics. Four of the main takeaways from the model comparison are as follows: (1) lumped-parameter models generally performed as well as artificial intelligence models, which means they capture the fundamental behaviour of the system with only a few parameters. (2) Artificial intelligence models were able to simulate extremes beyond the observed conditions, which is contrary to some persistent beliefs about these methods. (3) No overfitting was observed in any of the models, including in the models with many parameters, as performance in the validation period was generally only a bit lower than in the calibration period, which is evidence of appropriate application of the different models. (4) The presented simulations are the combined results of the applied method and the choices made by the modeller(s), which was especially visible in the performance range of the deep learning methods; underperformance does not necessarily reflect deficiencies of any of the models. In conclusion, the challenge was a successful initiative to compare different models and learn from each other. Future challenges are needed to investigate, for example, the performance of models in more variable climatic settings to simulate head series with significant gaps or to estimate the effect of drought periods.

A procedure is presented for valuation of information analysis (VOIA) to determine the need for additional information when assessing the effect of several design alternatives to manage future disturbances in hydrogeological systems. When planning for groundwater extraction and drawdown in areas where risks—such as land subsidence, wells running dry and drainage of streams and wetlands—are present, the need for risk-reducing safety measures must be carefully evaluated and managed. The heterogeneity of the subsurface calls for an assessment of trade-offs between the benefits of additional information to reduce the risk of erroneous decisions and the cost of collecting this information. A method is suggested that combines existing procedures for inverse probabilistic groundwater modelling with a novel method for VOIA. The method results in (1) a prior analysis where uncertainties regarding the efficiency of safety measures are estimated, and (2) a pre-posterior analysis, where the benefits of expected uncertainty reduction deriving from additional information are compared with the costs for obtaining this information. In comparison with existing approaches for VOIA, the method can assess multiple design alternatives, use hydrogeological parameters as proxies for failure, and produce spatially distributed VOIA maps. The method is demonstrated for a case study of a planned tunnel in Stockholm, Sweden, where additional investigations produce a low number of benefits as a result of low failure rates for the studied alternatives and a cause-effect chain where the resulting failure probability is more dependent on interactions within the whole system rather than on specific features.

This paper presents the results of the 2022 Groundwater Time Series Modelling Challenge, where 15 teams from different institutes applied various data-driven models to simulate hydraulic-head time series at four monitoring wells. Three of the wells were located in Europe and one was located in the USA in different hydrogeological settings in temperate, continental, or subarctic climates. Participants were provided with approximately 15 years of measured heads at (almost) regular time intervals and daily measurements of weather data starting some 10 years prior to the first head measurements and extending around 5 years after the last head measurement. The participants were asked to simulate the measured heads (the calibration period), to provide a prediction for around 5 years after the last measurement (the validation period for which weather data were provided but not head measurements), and to include an uncertainty estimate. Three different groups of models were identified among the submissions: lumped-parameter models (three teams), machine learning models (four teams), and deep learning models (eight teams). Lumped-parameter models apply relatively simple response functions with few parameters, while the artificial intelligence models used models of varying complexity, generally with more parameters and more input, including input engineered from the provided data (e.g. multi-day averages).

This paper presents the results of the 2022 Groundwater Time Series Modelling Challenge, where 15 teams from different institutes applied various data-driven models to simulate hydraulic-head time series at four monitoring wells. Three of the wells were located in Europe and one was located in the USA in different hydrogeological settings in temperate, continental, or subarctic climates. Participants were provided with approximately 15 years of measured heads at (almost) regular time intervals and daily measurements of weather data starting some 10 years prior to the first head measurements and extending around 5 years after the last head measurement. The participants were asked to simulate the measured heads (the calibration period), to provide a prediction for around 5 years after the last measurement (the validation period for which weather data were provided but not head measurements), and to include an uncertainty estimate. Three different groups of models were identified among the submissions: lumped-parameter models (three teams), machine learning models (four teams), and deep learning models (eight teams). Lumped-parameter models apply relatively simple response functions with few parameters, while the artificial intelligence models used models of varying complexity, generally with more parameters and more input, including input engineered from the provided data (e.g. multi-day averages). The models were evaluated on their performance in simulating the heads in the calibration period and in predicting the heads in the validation period. Different metrics were used to assess performance, including metrics for average relative fit, average absolute fit, fit of extreme (high or low) heads, and the coverage of the uncertainty interval. For all wells, reasonable performance was obtained by at least one team from each of the three groups. However, the performance was not consistent across submissions within each group, which implies that the application of each method to individual sites requires significanteffort and experience. In particular, estimates of the uncertainty interval varied widely between teams, although some teams submitted confidence intervals rather than prediction intervals. There was not one team, let alone one method, that performed best for all wells and all performance metrics. Four of the main takeaways from the model comparison are as follows: (1) lumped-parameter models generally performed as well as artificial intelligence models, which means they capture the fundamental behaviour of the system with only a few parameters. (2) Artificial intelligence models were able to simulate extremes beyond the observed conditions, which is contrary to some persistent beliefs about these methods. (3) No overfitting was observed in any of the models, including in the models with many parameters, as performance in the validation period was generally only a bit lower than in the calibration period, which is evidence of appropriate application of the different models. (4) The presented simulations are the combined results of the applied method and the choices made by the modeller(s), which was especially visible in the performance range of the deep learning methods; underperformance does not necessarily reflect deficiencies of any of the models. In conclusion, the challenge was a successful initiative to compare different models and learn from each other. Future challenges are needed to investigate, for example, the performance of models in more variable climatic settings to simulate head series with significant gaps or to estimate the effect of drought periods.