Search

In climates with a warm and a cold season, low flows are generated by different processes, which violates the homogeneity assumption of extreme value statistics. In this second part of a two-part series, we extend the mixed probability estimator of the companion paper (Laaha, 2023) to deal with dependency of seasonal events. We formulate a copula-based estimator for seasonal minima series and examine it in a hydrological context. The estimator is a valid generalization of the annual probability estimator and provides a consistent framework for estimating return periods of summer, winter, and annual events. Using archetypal examples we show that differences in the mixed estimator are always observed in the upper part of the distribution, which is less relevant for low-flow frequency analysis. The differences decrease as the return period increases so that both models coincide for the severest events. In a quantitative evaluation, we test the performance of the copula estimator on a pan-European data set. We find a large gain of both mixed distribution approaches over the annual estimator, making these approaches highly relevant for Europe as a whole. We then examine the relative performance gain of the mixed copula versus the mixed distribution approach in more detail. The analysis shows that the differences in the 100-year event are actually minimal. However, the differences in 2-year events are considerable in some of the catchments, with a relative deviation of −15 % to −25 % in the most affected regions. This points to a prediction bias of the mixed probability estimator that can be corrected using the copula approach. Using multiple regression models, we show that the performance gain can be well explained on hydrological grounds, with weak seasonality leading to a high potential for corrections and strong seasonal correlation reinforcing the need to take this potential into account. Accordingly, the greatest differences can be observed in mid-mountain regions in cold and temperate climates, where rivers have a strongly mixed low-flow regime. This finding is of particular relevance for event mapping, where regional severity can be misinterpreted when the seasonal correlation is neglected. We conclude that the two mixed probability estimators are quite similar, and both are conceptually more adequate than the annual minima approach for mixed summer and winter low-flow regimes. In regions with strong seasonal correlation the mixed copula estimator appears most appropriate and should be preferred over the mixed distribution approach.

A few low-frequent but extreme rainfall events generate substantial erosion amounts and, as a result, land degradation negatively affects soil health and agricultural productivity. Determining rainfall erosivity (from small to extreme events) that drive soil erosion will improve our knowledge and understanding of extreme erosion processes and our ability to develop and adapt adequate soil protection strategies. This study aims to analyze rainfall erosivity characteristics across Austria‘s main agricultural production zones using long-term (from 27 to 82 years) with high-resolution (5-minutes) rainfall data from 27 rain gauges. Erosive rainfall characteristics such as rainfall amount, event duration, and the maximum 30-minute intensity were investigated, and typical rainfall types were assessed through a clustering approach based on the k-means algorithm. Results identified three dominant types (clusters) of erosive rainfalls across Austria's agricultural areas. Seasonal distribution analysis showed that the predominant erosive events, characterized by short duration and high intensities, occur during the summer months from June to August. In contrast, the long-duration and low rainfall-intensity events are evenly distributed throughout the year. The spatial distribution of rainfalls related to the largest rainfall erosivities showed a pronounced occurrence in the southeastern pre-alpine areas. Knowledge of the occurrence of erosive rainfall events in space and time eventually supports the implementation of locally adapted Soil Conservation (SC) practices in Austria.

In our ACRP funded project EROS-A we aim at a comprehensive analysis of the erosive energy of rainfall, estimated soil loss, and reported damage using statistical and process-based methods. In this particular study, we evaluate the return periods of erosive rainfall events at 26 meteorological stations in Austria. The main focus is on daily cumulative rainfall and the maximum 30-min rainfall intensity (I30) as these were identified as key parameters for rainfall erosivity assessment. The extreme value series were obtained using both the Annual Maxima Series (AMS) and Peak Over Threshold (POT) approach, in order to assess which of the methods will be most accurate. The assumptions of stationarity and independence of the extreme value series were carefully checked using statistical trend and independence tests and no significant deviations were found.Generalized extreme value (GEV) and generalized Pareto (GPD) probability distributions were fitted using L-moment and maximum likelihood procedures. The GEV distribution is suited for AMS or block maxima data, whereas the GPD is suited for the POT series. For the obtained GEV and GPD models we examined extreme events with return periods of 2, 5, 10, 25, 50, and 100 years. We found that threshold selection is crucial for the POT, with diagnostic tools (such as mean residual life plots) not being fully decisive. Finally sensitivity analysis was performed where convergence of the fitted GPD to the GEV (AMS approach) helped determining robust thresholds for the GPD. The results show that the POT approach for daily cumulative precipitation is the most accurate in 69% of the cases and the AMS approach in 8% of the cases (different return periods and stations), while they have similar performance in 23% of the cases. Similar results are obtained for I30, were the success rates are 80% for the POT, 8% for the AMS and 12% for similar performance. In the next step, we will extend frequency analysis to a regional context, in order to map extreme rainfall erosivity across main agricultural production zones in Austria.

Urban shallow groundwater is highly impacted in terms of hydrogeology and water quality by anthropogenic activities and infrastructure, such as heating and cooling, surface sealing, leaking sewage pipes, and underground buildings. For a sustainable management of urban water resources, a better understanding of biogeochemical processes and its dynamics on a spatial and temporal scale in the urban subsurface is needed. So far, data sets including a critical minimum number of key parameters and an appropriate resolution in space and time have often been missing. Here, we introduce a multi-tracer approach applied to assess the shallow aquifers in Vienna. Water samples were collected twice, in autumn 2021 and spring 2022, respectively, from 150 groundwater wells in the city limits of Vienna. A comprehensive set of parameters (e.g. major ions, nutrients, heavy metals, water and nitrate stable isotopes) were analyzed to evaluate the spatial and seasonal variations in water origin and quality. Statistical analysis revealed that driving factors influencing groundwater quality include aquifer properties, interactions between groundwater-surface water, and redox conditions. A combined interpretation of conservative tracers indicated zones influenced by surface water - groundwater interactions that also influenced the water chemistry. Microbial anaerobic processes govern groundwater quality. In particular, contamination of nitrate from septic water and manure is locally reduced by denitrification, as proven by compound-specific isotope analysis, improving water quality. At the same time, other anaerobic processes, such as iron and manganese reduction, sulfate reduction, and methanogenesis deteriorate water quality. Finally, groundwater temperatures, up to 27°C, were observed close to urban underground infrastructure, hinting at subsurface buildings and surface sealing as stressors in shallow groundwater. In conclusion, our high resolution spatial sampling with the large set of parameters will not only allow a better understanding of groundwater quality dynamics, but also allows to evaluate effects to groundwater biodiversity and develop predictive mathematical models.

Droughts are significant hydrological and environmental hazards that threaten the ecological functioning of water bodies. Low flow with increased water temperature leads to a cascade of hydrochemical processes. This is a particular cause of concern for regions like eastern Austria, where agricultural land use and the projected risk of low flows and increased water temperatures due to climate change are particularly high. Under these scenarios, nutrient release from river sediments may become the dominant factor for the water quality of Iotic ecosystems. The role of this remobilization-potential for water quality is assessed in the project DIRT based on a combination of laboratory experiments with at-site water quality monitoring and regionalized streamflow observations.Here we focus on space-time models of low flow and stream temperature, which are crucial for upscaling the remobilization potential along the river network. We present a study that compares different models for spatio-temporal low flow regionalization at the monthly scale in eastern Austria.We evaluate three different statistical models: (i) a tree-based boosting model, (ii) a simple linear regression model with 3-way interactions, and (iii) a combination of a non-linear boosting approach and Top-kriging. Our results show a very high performance for all models, with an overall R² of 0.88 and a median R² of 0.70. The best performance is reached by the combination of Top-kriging with a non-linear boosting approach. However, accuracy of the model is somewhat lower in headwater gauges, whereas non-headwater catchments are even better modeled by a simple spatio-temporal Top-kriging approach. In a next step, the model shall be integrated with laboratory experiments and water-quality monitoring to develop space-time models that can predict the remobilization of pollutants from river sediments.

Drought is a worldwide phenomenon that originates from a prolonged deficiency in precipitation, often combined with high evaporation, over an extended region. The resultant meteorological water balance deficiency may cause a hydrological drought to develop into below normal levels of streamflow, lakes, and groundwater. Contemporary knowledge and experiences from an international team of drought experts are consolidated in a textbook (Tallaksen et al., 2023), which builds on an earlier edition (URL 1), however with significant new material added. An updated synthesis was needed because of hydrological drought-issues that has been emerged over the last 15 years, particularly when much of the topic is currently dominated by climate and climatology approaches. The textbook consists of three parts; Part I (Drought as a natural hazard) discusses the drought phenomenon, its main features, regional diversity and controlling processes. Part II (Estimation methods) presents contemporary approaches to drought estimation, including data and hydrological drought characteristics, statistical analysis of drought series, incl. frequency analysis, time series analysis and regionalization procedures, as well as process-based modelling. Part III (Living with drought) addresses aspects related to the interactions between water and people. Topics include historical and future drought, how human interventions influence drought, drought impacts and Drought Early Warning Systems. Knowledge and experiences shared in the book are from regions all over the world although somewhat biased to Europe and rivers that flow most of the year.This presentation aims to introduce the textbook, its motivation and content to a wide audience. The textbook is supported with worked examples and self-guided tours that are elaborated more extensively on Github. Worked examples include online procedures, code, and details of the calculation procedure that enable readers to repeat calculations in a stepwise manner, either with their own data or by using online datasets, and we encourage user’s feedbacks and experiences in testing these. Self-guided tours are demonstrations of advanced methodologies that involve several calculation steps and are given as an online presentations. Four datasets are included on Github; an international, a regional and two local datasets. The international dataset illustrates the drought phenomenon and its diversity across the world, whereas regional data and local aspects of drought are studied using a combination of hydroclimatological time series and catchment information. Hopefully, the textbook will contribute to an increased awareness of one of our main natural hazards, and thereby increase the preparedness and resilience of society to drought. ReferencesURL 1: http://europeandroughtcentre.com/resources/hydrological-drought-1st-edition/ Tallaksen, L.M., Van Lanen, H.A.J., Hannaford, J., Hisdal, H., Kingston, D.G., Laaha, G., Prudhomme, C., Stagge, J.H., Stahl, K., Van Loon, A.F., Wanders, N., Barker, L.J., Blauhut, V., Bloomfield, J.P., Cammalleri, C., Engeland, K., Everard, N., Facer-Childs, K., Fendeková, M., Fry, M., Gauster, T., Harrigan, S., Ionita, M., Marsh, T., Muchan, K., Ngongondo, C., Parry, S., Rees, G., Sauquet, E., Vidal, J-P. and Vogt, J. (2023). Hydrological Drought. Processes and Estimation Methods for Streamflow and Groundwater. Elsevier Publisher.

Extreme value statistics are well established for floods and are also receiving increasing attention in drought hydrology. They allow the user to characterize the severity of an event by a statistical probability or return period, a concept that is well understood in the scientific, policy, and public arenas. Frequency analysis is usually carried out on the basis of annual extreme event series, which is straightforward in its application and interpretation. However, in seasonal climates with a warm and a cold season, the low-flows can be generated by different processes, which violates the basic assumptions of extreme value statistics and can lead to inaccurate conclusions.Here we assess the value of a mixed distribution approach for low-flows to perform frequency analysis in catchments with a mixed summer/winter regime. We first present the theoretical concept of the mixed probability estimator for low-flows. We then illustrate the characteristics of the model for archetypal low-flow regimes, from pronounced summer and winter regimes to flow mixtures with weak seasonality. We successively evaluate the gain in performance from the mixed distribution model for a range of low flow regimes, based on a comprehensive Austrian dataset. We finally scrutinize the assumption of the mixed probability estimator and review the added value of using an extended, Copula-based framework. The results show large differences of event return periods, and suggest that the mixed estimator is relevant not only for mountain forelands, but for a much wider range of catchment typologies across Europe. These even include typical summer regimes when only single winter low-flows are mixed in. We conclude that the mixed distribution approaches outperform the conventional frequency estimator and should be used by default in seasonal climates where summer and winter low flows occur.

Statistical learning methods offer a promising approach for low-flow regionalization. We examine seven statistical learning models (Lasso, linear, and nonlinear-model-based boosting, sparse partial least squares, principal component regression, random forest, and support vector regression) for the prediction of winter and summer low flow based on a hydrologically diverse dataset of 260 catchments in Austria. In order to produce sparse models, we adapt the recursive feature elimination for variable preselection and propose using three different variable ranking methods (conditional forest, Lasso, and linear model-based boosting) for each of the prediction models. Results are evaluated for the low-flow characteristic Q95 (Pr(Q>Q95)=0.95) standardized by catchment area using a repeated nested cross-validation scheme. We found a generally high prediction accuracy for winter (RCV2 of 0.66 to 0.7) and summer (RCV2 of 0.83 to 0.86). The models perform similarly to or slightly better than a top-kriging model that constitutes the current benchmark for the study area. The best-performing models are support vector regression (winter) and nonlinear model-based boosting (summer), but linear models exhibit similar prediction accuracy. The use of variable preselection can significantly reduce the complexity of all the models with only a small loss of performance. The so-obtained learning models are more parsimonious and thus easier to interpret and more robust when predicting at ungauged sites. A direct comparison of linear and nonlinear models reveals that nonlinear processes can be sufficiently captured by linear learning models, so there is no need to use more complex models or to add nonlinear effects. When performing low-flow regionalization in a seasonal climate, the temporal stratification into summer and winter low flows was shown to increase the predictive performance of all learning models, offering an alternative to catchment grouping that is recommended otherwise.

In seasonal climates with a warm and a cold season, low flows are generated by different processes so that the annual extreme series will be a mixture of summer and winter low-flow events. This leads to a violation of the homogeneity assumption for all statistics derived from the annual series and gives rise to inaccurate conclusions. In this first part of a two-paper series, a mixed distribution approach to perform frequency analysis in catchments with mixed low-flow regimes is proposed. We formulate the theoretical basis of the mixed distribution approach for the lower extremes based on annual minima series. The main strength of the model is that it allows the user to estimate return periods of summer low flows, winter low flows, and annual return periods in a theoretically sound and consistent way. Using archetypal examples, we show how the model behaves for a range of low-flow regimes, from distinct winter and summer regimes to mixed regimes where seasonal occurrence in summer and winter is equally likely. The examples show in a qualitative way the loss in accuracy one has to expect with conventional extreme value statistics performed with the annual extremes series. The model is then applied to a comprehensive Austrian data set to quantify the expected gain of using the mixed distribution approach compared to conventional frequency analysis. Results indicate that the gain of using a mixed distribution approach is indeed large. On average, the relative deviation is 21 %, 39 %, and 63 % when estimating the low flow with a 20-, 50-, and 100-year return period. For the 100-year event, 75 % of stations show a performance gain of >10 %, 41 % of stations > 50 %, and 25 % of stations > 80.6 %. This points to a broad relevance of the approach that goes beyond highly mixed seasonal regimes to include the strongly seasonal ones. We finally correlate the performance gain with seasonality indices in order to show the expected gain conditional to the strength of seasonality expressed by the ratio of average summer and winter low flow seasonality ratio (SR). For the 100-year event, the expected gain is about 70 % for SR=1.0, 20 % for SR=1.5, and 10 % for SR=2.0. The performance gain is further allocated to the spatial patterns of SR in the study area. The results suggest that the mixed estimator is relevant not only for mountain forelands but to a much wider range of catchment typologies. The mixed distribution approach provides one consistent approach for summer, winter, and annual probabilities and should be used by default in seasonal climates with a cold winter season where summer and winter low flows can occur.

In climates with a warm and a cold season, low-flows are generated by different processes, which violates the homogeneity assumption of extreme value statistics. In this second part of a two-part series, we extend the mixed probability estimator of the companion paper (Laaha, G., 2022) to deal with dependency of seasonal events. We formulate a copula-based estimator for seasonal minima series and examine it in a hydrological context. The estimator is a valid generalization of the annual probability estimator and provides a consistent framework for estimating return periods of summer, winter and annual events. Using archetypal examples we show that differences to the mixed estimator are always observed in the upper part of the distribution, which is less relevant for low-flow frequency analysis. The differences decrease as the return period increases so that both models coincide for the severest events. In a quantitative evaluation, we test the performance of the copula estimator on a pan-European data set. We find a large gain of both mixed distribution approaches over the annual estimator, making these approaches highly relevant for Europe as a whole. We then examine in more detail the relative performance gain of the mixed copula versus the mixed distribution approach. The analysis shows that the gain for the 100-year event is actually minimal. However, the gain for 2-year events is considerable in some of the catchments, with a relative deviation of -15 to -23 % in the most affected regions. This points to a prediction bias of the mixed probability estimator that can be corrected using the copula approach. Using multiple regression models, we show that the performance gain can be well explained on hydrological grounds, with weak seasonality leading to a high potential for corrections and strong seasonal correlation reinforcing the need to take it into account. Accordingly, the greatest differences can be observed in mid-mountain regions in cold and temperate climates, where rivers have a strongly mixed low-flow regime. This finding is of particular relevance for event mapping, where regional severity can be misinterpreted when the seasonal correlation is neglected. We conclude that the two mixed probability estimators are quite similar, and both are more accurate as the annual minima approach. In regions with strong seasonal correlation the mixed copula estimator is most accurate and should be preferred over the mixed distribution approach.

Accurate predictions of seasonal low flows are critical for a number of water management tasks that require inferences about water quality and the ecological status of water bodies. This paper proposes an extreme gradient tree boosting model (XGBoost) for predicting monthly low flow in ungauged catchments. Particular emphasis is placed on the lowest values (in the magnitude of annual low flows and below) by implementing the expectile loss function to the XGBoost model. For this purpose, we test expectile loss functions based on decreasing expectiles (from τ=0.5 to 0.01) that give increasing weight to lower values. These are compared to common loss functions such as mean and median absolute loss. Model optimization and evaluation are conducted using a nested cross-validation (CV) approach that includes recursive feature elimination (RFE) to promote parsimonious models. The methods are tested on a comprehensive dataset of 260 stream gauges in Austria, covering a wide range of low-flow regimes. Our results demonstrate that the expectile loss function can yield high prediction accuracy, but the performance drops sharply for low expectile models. With a median R2 of 0.67, the 0.5 expectile yields the best-performing model. The 0.3 and 0.2 perform slightly worse, but still outperform the common median and mean absolute loss functions. All expectile models include some stations with moderate and poor performance that can be attributed to some systematic error, while the seasonal and annual variability is well covered by the models. Results for the prediction of low extremes show an increasing performance in terms of R2 for smaller expectiles (0.01, 0.025, 0.05), though leading to the disadvantage of classifying too many extremes for each station. We found that the application of different expectiles leads to a trade-off between overall performance, prediction performance for extremes, and misclassification of extreme low-flow events. Our results show that the 0.1 or 0.2 expectiles perform best with respect to all three criteria. The resulting extreme gradient tree boosting model covers seasonal and annual variability nicely and provides a viable approach for spatiotemporal modeling of a range of hydrological variables representing average conditions and extreme events.

In this paper future changes of surface water availability in Austria are investigated. We use an ensemble of downscaled and bias-corrected regional climate model simulations of the EURO-CORDEX initiative under moderate mitigation (RCP4.5) and Paris agreement (RCP2.6) emission scenarios. The climatic water balance and its components (rainfall, snow melt, glacier melt and atmospheric evaporative demand) are used as indicators for surface water availability and we focus on different altitudinal classes (lowland, mountainous and high alpine) to depict a variety of processes in complex terrain. Apart from analysing the mean changes of these components we also pursue a hazard risk approach by estimating future changes in return periods of drought events of a given magnitude as observed in the reference period. The results show in general wetter conditions over the course of the 21st century over Austria on an annual basis compared to the reference period 1981–2010. Considering seasonal differences, winter and spring are getting wetter due to an increase in precipitation and a higher fraction of rainfall as a consequence of rising temperatures. In summer only little changes in the mean of the climatic water balance conditions are visible across the model ensemble. On the contrary, by analysing changes in return periods of drought events, an increasing risk of moderate and extreme drought events during summer is apparent, a signal emerging within the climate system along increasing warming.

ZusammenfassungDas Thema Auffindbarkeit von FAH wirft nach wie vor eine Reihe unbeantworteter Forschungsfragen hinsichtlich der Instinktbewegung von Fischen, ausgelöst durch Schlüsselreize, auf. Demzufolge ist eine systematische Abklärung für bestehende und insbesondere für zukünftige FAH Projekte zielführend. Die Auffindbarkeit von FAH wird in verschiedenen Leitfäden, neben der Positionierung der Lage des Einstiegs, primär am Parameter Strömung (Rheotaxis) festgemacht. Aus hydraulischer Sicht hat jedoch die am Einstieg austretende Strömung schon von der Größe des Ausbreitungsfeldes her lediglich eine maximal untergeordnete Wirkung. Andere, äußere, die Wanderung beeinflussende Faktoren wie z. B. Licht, Temperatur, Sauerstoffgehalt, akustische Impulse u. a. m. wurden bislang vollkommen bzw. weitgehend außer Acht gelassen. Felduntersuchungen, die das Aufstiegsverhalten der Grundgesamtheit der migrierenden Individuen mit den Umgebungsreizen zeitgenau in Zusammenhang bringen, fehlen bisher. Ziel dieses Papers ist die systematische Abklärung der Wirkung alternativer Schlüsselreize auf die Auffindbarkeit von Fischaufstiegshilfen im Rahmen eines mehrstufigen Forschungsprojekts der VERBUND Hydro Power GmbH. Bei der hier beschriebenen Studie wurden die Daten des Frühjahrs- und des Herbstmonitorings 2019 in der FAH am Kraftwerksstandort Edling/Drau analysiert. Neben den Reizen in Form der verschiedenen Strömungsparameter werden künstliche Reize, welche auf die Phonotaxis, also die Orientierung aufgrund von akustischen Reizen, und die Phototaxis, die Orientierung an Lichtreizen abzielen, gesetzt. Auf der Grundlage der zeitgenauen Erfassung jedes einzelnen aufsteigenden Individuums wurde mittels technischer, fischökologischer und mathematischer Methoden der Zusammenhang zwischen den Schlüsselreizen und der Auffindbarkeit bzw. den Aufstiegsraten in der FAH untersucht. Zum Einsatz kamen hierbei neben dem FishCam-System zur zeitgenauen Erfassung der Wanderbewegungen die FishNet-Software sowie statistische Methoden zur Variablenselektion auf Basis von Regressionsmodellen mit unterschiedlichen Pönalisierungstermen wie LASSO, SCAD und MCP. Die anthropogen angebotenen zusätzlichen Schlüsselreize Licht und Akustik am Einstieg der FAH KW Edling bewirkten keine Steigerung der Aufstiegsraten und somit keine Verbesserung der Auffindbarkeit der FAH.

Low flow estimation is a crucial part in water management. Prediction of low flow in ungauged basins is often performed through statistical models. This can be either regionalization approaches, where homogeneous regions are used for modeling, or single model frameworks that range from simple linear models to more complex as random forest, support vector regression or deep learning approaches. Although there are large sample studies for the US (e.g. Tyralis et al. 2021) or Australia (e.g. Worland et al. 2018), we are not aware of a study that combines different large datasets and analyzing the effect on prediction accuracy. We are hypothesing that the heterogeneity of many datasets together can improve prediction accuracy for tree-based models relative to linear models. Hence, we propose to combine several similar datasets and analyze the effect on prediction accuracy for estimating Q95 by a simple random forest model.Our study uses four large hydrological datasets – CAMELS-GB (Coxon et al. 2020), CAMELS-US (Addor et al. 2017), CAMELS-AUS (Fowler et al. 2021) and LamaH-CE (Klinger et al., 2021). We are applying a random forest model to ensure that interactions and non-linearity can be captured. Prediction accuracy is evaluated by leave one out cross-validation (LOOCV) and several performance metrics, e.g. median absolute error (MDAE), or root mean squared error (RMSE). LOOCV is used for each individual dataset and in one run for the merged dataset. Results indicate that merging datasets can improve prediction accuracy, but models fail to correctly predict low flows around zero.ReferencesAddor, N., Newman, A. J., Mizukami, N., and Clark, M. P.: The CAMELS data set: catchment attributes and meteorology for large-sample studies, Hydrol. Earth Syst. Sci., 21, 5293–5313, https://doi.org/10.5194/hess-21-5293-2017, 2017. Fowler, K. J. A., Acharya, S. C., Addor, N., Chou, C., and Peel, M. C.: CAMELS-AUS: hydrometeorological time series and landscape attributes for 222 catchments in Australia, Earth Syst. Sci. Data, 13, 3847–3867, https://doi.org/10.5194/essd-13-3847-2021, 2021. Coxon, G., Addor, N., Bloomfield, J. P., Freer, J., Fry, M., Hannaford, J., Howden, N. J. K., Lane, R., Lewis, M., Robinson, E. L., Wagener, T., and Woods, R.: CAMELS-GB: hydrometeorological time series and landscape attributes for 671 catchments in Great Britain, Earth Syst. Sci. Data, 12, 2459–2483, https://doi.org/10.5194/essd-12-2459-2020, 2020. Klingler, C., Schulz, K., and Herrnegger, M.: LamaH-CE: LArge-SaMple DAta for Hydrology and Environmental Sciences for Central Europe, Earth Syst. Sci. Data, 13, 4529–4565, https://doi.org/10.5194/essd-13-4529-2021, 2021. Tyralis, H.; Papacharalampous, G.; Langousis, A.; Papalexiou, S.M. Explanation and Probabilistic Prediction of Hydrological Signatures with Statistical Boosting Algorithms. Remote Sens. 2021, 13, 333. https://doi.org/10.3390/rs13030333 Worland, S. C., Farmer, W. H., and Kiang, J. E.: Improving predictions of hydrological low-flow indices in ungaged basins using machinelearning, Environmental modelling & software, 101, 169–182, https://doi.org/10.1016/j.envsoft.2017.12.021, 2018.

Habitat fragmentation is one of the drivers for amphibian population declines globally. Especially in industrialized countries roads disrupt the seasonal migration of amphibians between hibernation and reproduction sites, often ending in roadkills. Thus, a timely installing of temporary mitigation measures is important for amphibian conservation. We wanted to find out if plant phenology can be a proxy in advance to determine the start of amphibian migration, since both phenomena are triggered by temperature. We analysed data of 3751 amphibian and 7818 plant phenology observations from citizen science projects in Austria between 2000 and 2018. Using robust regression modelling we compared the migration of common toads (Bufo bufo) and common frogs (Rana temporaria) with the phenology of five tree, one shrub, and one herb species. Results showed close associations between the migration of common frogs and phenological phases of European larch, goat willow and apricot. Models based on goat willow predict migration of common frog to occur 21 days after flowering, when flowering was observed on 60th day of year; apricot based models predict migration to occur 1 day after flowering, observed on the 75th day of year. Common toads showed weaker associations with plant phenology than common frogs. Our findings suggest that plant phenology can be used to determine the onset of temporary mitigation measures for certain amphibian species to prevent roadkills.

Statistical learning methods offer a promising approach for low flow regionalization. We examine seven statistical learning models (lasso, linear and non-linear model based boosting, sparse partial least squares, principal component regression, random forest, and support vector machine regression) for the prediction of winter and summer low flow based on a hydrological diverse dataset of 260 catchments in Austria. In order to produce sparse models we adapt the recursive feature elimination for variable preselection and propose to use three different variable ranking methods (conditional forest, lasso and linear model based boosting) for each of the prediction models. Results are evaluated for the low flow characteristic Q95 (Pr(Q>Q95) = 0.95) standardized by catchment area using a repeated nested cross validation scheme. We found a generally high prediction accuracy for winter (R2CV of 0.66 to 0.7) and summer (R2CV of 0.83 to 0.86). The models perform similar or slightly better than a Top-kriging model that constitutes the current benchmark for the study area. The best performing models are support vector machine regression (winter) and non-linear model based boosting (summer), but linear models exhibit similar prediction accuracy. The use of variable preselection can significantly reduce the complexity of all models with only a small loss of performance. The so obtained learning models are more parsimonious, thus easier to interpret and more robust when predicting at ungauged sites. A direct comparison of linear and non-linear models reveals that non-linear relationships can be sufficiently captured by linear learning models, so there is no need to use more complex models or to add non-liner effects. When performing low flow regionalization in a seasonal climate, the temporal stratification into summer and winter low flows was shown to increase the predictive performance of all learning models, offering an alternative to catchment grouping that is recommended otherwise.

This study investigates the atmospheric drivers of severe precipitation deficits in the Greater Alpine Region during the last 210 years utilizing a daily atmospheric circulation type reconstruction. Precipitation deficit tends to be higher during periods with more frequent anticyclonic (dry) and less frequent cyclonic (wet) circulation types, as would be expected. However, circulation characteristics are not the main drivers of summer precipitation deficit. Dry soils in the warm season tend to limit precipitation, which is particularly the case for circulation types that are sensitive to a soil moisture‐precipitation feedback. This mechanism is of specific relevance in explaining the major drought decades of the 1860s and 1940s. Both episodes show large negative precipitation anomalies in spring followed by increasing frequencies of circulation types sensitive to soil moisture precipitation feedbacks. The dry springs of the 1860s were likely caused by circulation characteristics that were quite different from those of recent decades as a consequence of the large spatial extent of Arctic sea ice at the end of the Little Ice Age. On the other hand, the dry springs of the 1940s developed under a persistent positive pressure anomaly across Western and Central Europe, triggered by positive sea surface temperatures in the western subtropical Atlantic., Key Points Winter and spring droughts are strongly related to the Eastern Atlantic‐Western Russia atmospheric variability patternSummer droughts are enhanced by prevailing dry spring conditions through a positive soil moisture‐precipitation feedbackThe major dry episodes in the 1860s and 1940s are particularly driven by this feedback mechanismThe dry springs of the 1860s were likely caused by unusual circulation regimes, synchronized with large spatial extents of Arctic sea ice at the end of the Little Ice AgeThe dry springs of the 1940s are due to positive sea surface temperature anomalies in the western subtropical Atlantic, which are related to a positive Eastern Atlantic‐Western Russia pattern

Aiming for refined drought characterization, the validation of targeted drought indices is of vital importance. In this study, we compared the performance of established drought indices – the SPI (Standardized Precipitation Index) and the SPEI (Standardized Precipitation Evapotranspiration Index) – with standardized drought indices using a recently developed, vine copula based method for the computation of multivariate drought indices (here addressed as VCI). For our validation study, we used several environmental drought indicators: monthly streamflow anomalies and streamflow drought events from a network of 332 catchments across Europe, as well as gross primary production (GPP) and net ecosystem exchange (NEE) for Germany. The novel multivariate VC-Indices can combine two or more user-selected, drought relevant variables to model different drought types, depending on the user-application. Validation with streamflow data showed that the maximum probability of drought detection values for SPEI, SPI and VCI was observed for 12.0%, 25.9% and 62.0% of the catchments, and the minimum false alarm rate values for SPEI, SPI and VCI was observed for 20.5%, 33.4% and 46.1% of the catchments, respectively. Validation with carbon flux data showed that the average R2 values of a pixel-wise linear regression for the growing season for the period 1980 to 2010 between SPEI, SPI and VCI with NEE were 0.26, 0.07 and 0.37, respectively. Similarly, the average R2 values for SPEI, SPI and VCI with GPP were 0.03, 0.04 and 0.14, respectively. Our results emphasize using the VCI as an additional source of information in order to allow better understanding of drought characterization.

We deliver the first analysis of the 2021 cold spell in Texas which combines temperature dependent load estimates with temperature dependent estimates of power plant outages to understand the frequency of loss of load events, using a 71 year long time series of climate data. The expected avoided loss from full winterization is 11.74bn\$ over a 30 years investment period. We find that large-scale winterization, in particular of gas infrastructure and gas power plants, would be profitable, as related costs for winterization are substantially lower. At the same moment, the necessary investments involve risk due to the low-frequency of events – the 2021 event was the largest and we observe only 8 other similar ones in the set of 71 simulated years. Regulatory measures may therefore be necessary to enforce winterization.

River systems have undergone a massive transformation since the Anthropocene. The natural properties of river systems have been drastically altered and reshaped, limiting the use of management frameworks, their scientific knowledge base and their ability to provide adequate solutions for current problems and those of the future, such as climate change, biodiversity crisis and increased demands for water resources. To address these challenges, a socioecologically driven research agenda for river systems that complements current approaches is needed and proposed. The implementation of the concepts of social metabolism and the colonisation of natural systems into existing concepts can provide a new basis to analyse the coevolutionary coupling of social systems with ecological and hydrological (i.e., ‘socio-ecohydrological’) systems within rivers. To operationalize this research agenda, we highlight four initial core topics defined as research clusters (RCs) to address specific system properties in an integrative manner. The colonisation of natural systems by social systems is seen as a significant driver of the transformation processes in river systems. These transformation processes are influenced by connectivity (RC 1), which primarily addresses biophysical aspects and governance (RC 2), which focuses on the changes in social systems. The metabolism (RC 3) and vulnerability (RC 4) of the social and natural systems are significant aspects of the coupling of social systems and ecohydrological systems with investments, energy, resources, services and associated risks and impacts. This socio-ecohydrological research agenda complements other recent approaches, such as ‘socio-ecological’, ‘socio-hydrological’ or ‘socio-geomorphological’ systems, by focusing on the coupling of social systems with natural systems in rivers and thus, by viewing the socioeconomic features of river systems as being just as important as their natural characteristics. The proposed research agenda builds on interdisciplinarity and transdisciplinarity and requires the implementation of such programmes into the education of a new generation of river system scientists, managers and engineers who are aware of the transformation processes and the coupling between systems.

This paper aims to develop a spatiotemporal model to estimate monthly low flow quantiles Q95 [P(QIn a first step we are adapting a spatiotemporal model initially designed for modeling air pollution data. This approach is based on empirical orthogonal functions (EOF), that should capture the temporal structure of the spatiotemporal model. The EOFs are weighted by regression coefficients estimated by universal kriging. We extend the model by using GLM-boosting, LASSO, Principal Component Regression (PCR) and Random Forest (RF) for selecting the regression coefficients of the EOFs. Furthermore, we do not limit the kriging structure of the residual field to geographical coordinates but use a broader approach of physiographic kriging. In a second step we implement separate models for the mean parts of the model and the residual parts of the model. The mean field is defined by statistical learning methods as RF, GAM-boosting, LASSO and Support Vector Machines (SVM). For the residual field we define two different approaches, either the method developed in the first step or spatiotemporal kriging.Model performance is evaluated by cross validation and the best model is selected by the mean squared error (MSE).

In this contribution future changes of surface water availability over the Austrian domain is investigated. We use an ensemble of downscaled and bias-corrected regional climate model simulations of the EURO-CORDEX initiative under moderate mitigation (RCP4.5) and Paris agreement (RCP2.6) emission scenarios. The climatic water balance and its components (rainfall, snow melt, glacier melt and potential evapotranspiration) are used as indicators for surface water availability and we focus on different altitudinal classes (lowland, mountainous and high alpine) to depict a variety of processes in complex terrain. Apart from analysing the mean changes of these quantities we also pursue a hazard risk approach by estimating changes in return periods of drought events of a given magnitude as observed in the reference period. The results show in general wetter conditions over the course of the 21st century over Austria. Considering seasonal differences, winter and spring will be getting wetter due to an increase in precipitation along with a higher rainfall/snowfall fraction as a consequence of rising temperatures. In summer only little changes in the ensemble median of the climatic water balance are visible, hence uncertainties are large due to a considerable ensemble spread. However, by analysing changes in return periods of drought events, a robust signal of increasing risk of moderate and extreme drought events during summer is apparent. It emerges from an increase in interannual variability of the climatic water balance, which likely stems from intensified land-atmosphere coupling under climate change sustaining and intensifying spring preconditions towards even wetter or dryer summers.

Climate data sets are widely used for renewable power simulation. While previous generations of global reanalysis data including MERRA-2 and ERA-Interim have been widely assessed for their suitability to simulate variable renewable power systems, more recent datasets such as ERA5 and ERA5-Land lack validation, particularly in regions outside of Europe.Here, we assess the accuracy and bias of wind power simulation using ERA5 wind speeds in Brazil and New Zealand as well as solar photovoltaic power simulation accuracy using ERA5-Land solar radiation and temperature data in Chile. We compare the performance of ERA5 and ERA5-Land to MERRA-2 based renewable power generation. The reference data sets are capacity factors derived from data measured at individual installations in each country and the performance indicators include the pearson’s correlation coefficient, mean bias error (MBE) and root mean square error (RMSE). For wind power simulation, we also assess a bias correction method using the Global Wind Atlas.Since models applying the resulting datasets are based on different spatial and temporal scales, we also aim at finding a relation between the spatial and temporal resolution and simulation quality. We assess the simulation results applying spatial aggregation ranging from individual installations to the country level and temporal aggregation varying from hours to months. This aids to evaluate the reliability of the simulated renewable power generation time series on various spatiotemporal scales for future simulation efforts.Overall, we find that both datasets, ERA5 and ERA5-Land, perform well in wind and solar photovoltaic power simulations. For wind power simulation, ERA5 shows improved performance compared to MERRA-2 based wind power simulation, while for solar photovoltaic the improvements of ERA5-Land compared to MERRA-2 are minor. Correlation of wind power generation is around 0.8 without correction and MBEs around -0.1. Mean bias correction with the Global Wind Atlas does not consistently improve simulation results. For the solar photovoltaic power simulation, we find correlations above 0.75, while the MBE is between -0.05 and 0.1.

Intermittent rivers are prevalent in many countries across Europe, but little is known about the temporal evolution of intermittence and its relationship with climate variability. Trend analysis of the annual and seasonal number of zero-flow days, the maximum duration of dry spells and the mean date of the zero-flow events is performed on a database of 452 rivers with varying degrees of intermittence between 1970 and 2010. The relationships between flow intermittence and climate are investigated using the standardized precipitation evapotranspiration index (SPEI) and climate indices describing large-scale atmospheric circulation. The results indicate a strong spatial variability of the seasonal patterns of intermittence and the annual and seasonal number of zero-flow days, highlighting the controls exerted by local catchment properties. Most of the detected trends indicate an increasing number of zero-flow days, which also tend to occur earlier in the year, particularly in southern Europe. The SPEI is found to be strongly related to the annual and seasonal zero-flow day occurrence in more than half of the stations for different accumulation times between 12 and 24 months. Conversely, there is a weaker dependence of river intermittence with large-scale circulation indices. Overall, these results suggest increased water stress in intermittent rivers that may affect their biota and biochemistry and also reduce available water resources., This research is part of the COST action CA15113 SMIRES, Science and Management of Intermittent Rivers and Ephemeral Streams https://www.smires.eu/, a short-term scientific mission that has been funded for Agnieszka Rutkowska. Agnieszka Rutkowska was also supported by the Ministry of Science and Higher Education of the Republic of Poland (grant DS 3371). The different data providers are gratefully acknowledged. Some of the station data were retrieved from the Global Runoff Data Centre (56068 Koblenz, Germany) and from the HyMeX programme database. The authors would like to thank two anonymous reviewers, Kendra Kaiser and the Associate Editor, Stephanie Kampf, for their constructive comments.

Restoring the longitudinal connectivity of rivers through fish passes is of great importance for achieving good ecological status of surface waters. However, the key stimuli determining the findability of the entrance of these structures is still subject to debate. In this article, the influence of water temperature, light and acoustic stimuli in addition to flow parameters on fish movement is assessed. Analyses are based on a comprehensive dataset of 40,000 fish migrations covering time accurate recording of ascending individuals during the observation periods from 2015 to 2020 in various fish passes on the river Drava in Carinthia, Austria. The data are assessed by technical, fish ecological and statistical methods. Results indicate that the effect of the water temperature gradient between the fish pass and the main river has most impact on the ascent rates, whereas the two factors of light and acoustic, as well as most of the flow parameters, have no or negligible effect on the findability. A favourable thermal environment can be important to ensure efficient upstream migration and thus facilitate the findability of fish passes.

Standardized drought indices such as SPI are frequently used around the world to assess drought severity across a continent or a larger region covering different meteorological regimes. But how standard are the standardized indices? In this paper we quantify the uncertainty of SPI and SPEI based on an Austrian data set to shed light on what are the main sources of uncertainty in the study area. Here we analyze the uncertainty contributions by a linear mixed model that employs a restrictive maximum likelihood estimator in order to produce unbiased variance and covariance components. Five factors that either defy the control of the analyst (record length, observation period), or need to be subjectively decided during the steps of the calculation (choice of the distribution, parameter estimation method, and GOF-test of the fitted distribution) are considered. The results show that, overall, the choice of the distribution and the observational window are the most important sources of uncertainty. We quantify the relative uncertainty contributions in greater detail in order to give guidance how to make estimates most accurate for a given data set. We finally analyze the total uncertainty of SPI and SPEI to shed light on our main question whether the indices are skillful enough to provide a quantification of atmospheric drought that is standardized enough to allow the intended comparisons across various data situations and meteorological regimes.

Pluvial flooding is getting increasing attention around the world, but there are very limited studies on how to estimate pluvial flood damage risk at a regional scale. In this paper, we evaluate the value of statistical methods for pluvial flood risk estimation. One objective is to assess if statistical methods can deliver significant relationships between precipitation indices, soil and landscape characteristics, and pluvial flood damage risk in agricultural areas. Additionally, we want to explore if the statistical methods could be a cost-efficient alternative to deterministic surface run-off models. The analyses are performed for agricultural landscapes in Upper Austria where flood damage location reports of the Austrian hail insurance are available for the 2007 to 2013 period. We assess the relative performances of a generalized linear model (GLM) and a random forest (RF) model to estimate pluvial flood occurrence based on presence/absence raster data set, where the imbalance of the binary target variable and multicollinearity of predictors pose specific challenges. The analysis reveal that within the study area the most severe agricultural damages were triggered by intense rain events, but in one case the damage was caused by longer lasting low intense rain. By means of log-linear models and random forests the relationships between the location characteristics and damage events are investigated. A specific focus is laid on comparing the skills of continuous versus classified parameters, and on methods for dealing with collinearity. We compare the performances of both models based on cross-validation statistics and the process-realism of results. We finally discuss how far the statistical methods appear appropriate for quantifying pluvial flood damage risk at a regional scale.

Background The Spanish slug (Arion vulgaris, also known as A. lusitanicus) is considered one of the most invasive species in agriculture, horticulture and private gardens all over Europe. Although this slug has been problematic for decades, there is still not much known about its occurrence across private gardens and the underlying meteorological and ecological factors. One reason for this knowledge gap is the limited access of researchers to private gardens. Here we used a citizen science approach to overcome this obstacle and examined whether the occurrence of Arionidae in Austrian gardens was associated with meteorological (air temperature, precipitation, global solar radiation, relative humidity) or ecological factors (plant diversity, earthworm activity). Occurrence of the invasive A. vulgaris versus the similar-looking native A. rufus was compared using a DNA-barcoding approach. Results Slugs were collected from 1061 gardens from the dry Pannonian lowland to the wet alpine climate (altitudinal range 742 m). Slug abundance in gardens was best explained and negatively associated with the parameters “sum of the mean air temperature in spring”, “number of frost days in the previous winter” and “mean daily global solar radiation on the day of data collection”. Precipitation, plant diversity and earthworm activity were also related to slug abundance, but positively. Out of our genetic sampling of collected slugs, 92% belonged to A. vulgaris. Conclusions Our study showed that citizen science (i) is a feasible approach to record species occurrence in restricted areas across a wide geographical range and (ii) could be more widely employed in order to identify underlying environmental factors of species occurrence.

Compared to the alpine regions of Austria, the eastern part of the country is overall warmer and drier with some tree species growing at the dry limit within their ecological range. This suggests that tree rings may be a valuable proxy for hydroclimatic variability. In this study, we develop ring-width, earlywood-width and latewood-width chronologies obtained from drought sensitive living trees and historical timber from one of the driest and warmest regions of Austria, the Weinviertel. For this, samples of four different tree species − fir ( Abies alba Mill.), spruce ( Picea abies (L.) Karst.), oak ( Quercus petraea Liebl., Q. robur L., Q. cerris L.), and pine ( Pinus sylvestris L., P. nigra Arnold) were collected from 88 sites (23 forest stands and 65 historical constructions). 1589 samples (oak: 592 samples, pine: 577, spruce: 212, fir: 208) were dated. Dendro-provenancing was required to ensure that only regional material has entered the chronologies. According to the analyses, historical pine and oak wood were nearly 100% regional, while spruce and fir wood were imported to a great extent with only 44% spruce and 35% fir confirmed to be regional. Because site conditions of historical wood samples are unknown but can have a significant influence on tree growth, changes to site replication over time were important in the assessment of chronology reliability. We also controlled for potential artificial increases in sample depth when more than one construction element is made from a single trunk. We assessed the pine (1584-2011 AD) and oak (1244-2011 AD) chronologies’ potential for reconstructing past hydroclimatic variability by means of response functions in a 17-months window from previous June to current October with climate data from three weather stations (Vienna, Retz, Brno) from 1897 to 2010 AD. The highest seasonal response coefficients are for oak and pine ring width (∼0.43) with respect to aggregated March to July precipitation totals and Thornthwaite climatic water balance. These chronologies reveal a high potential for estimating past changes in regional-scale moisture availability during the earlier growing season.

The assessment of road infrastructure exposure to extreme weather events is of major importance for scientists and practitioners alike. In this study, we compare the different extreme value approaches and fitting methods with respect to their value for assessing the exposure of transport networks to extreme precipitation and temperature impacts. Based on an Austrian data set from 25 meteorological stations representing diverse meteorological conditions, we assess the added value of partial duration series over the standardly used annual maxima series in order to give recommendations for performing extreme value statistics of meteorological hazards. Results show the merits of the robust L-moment estimation, which yielded better results than maximum likelihood estimation in 62 % of all cases. At the same time, results question the general assumption of the threshold excess approach (employing partial duration series, PDS) being superior to the block maxima approach (employing annual maxima series, AMS) due to information gain. For low return periods (non-extreme events) the PDS approach tends to overestimate return levels as compared to the AMS approach, whereas an opposite behaviour was found for high return levels (extreme events). In extreme cases, an inappropriate threshold was shown to lead to considerable biases that may outperform the possible gain of information from including additional extreme events by far. This effect was neither visible from the square-root criterion, nor from standardly used graphical diagnosis (mean residual life plot), but from a direct comparison of AMS and PDS in synoptic quantile plots. We therefore recommend performing AMS and PDS approaches simultaneously in order to select the best suited approach. This will make the analyses more robust, in cases where threshold selection and dependency introduces biases to the PDS approach, but also in cases where the AMS contains non-extreme events that may introduce similar biases. For assessing the performance of extreme events we recommend conditional performance measures that focus on rare events only in addition to standardly used unconditional indicators. The findings of the study directly address road and traffic management, but can be transferred to a range of other environmental variables including meteorological and hydrological quantities.

Zu den wichtigsten Umweltdaten zahlen Klima-, Luft- und Bodenparameter, Landbedeckung, Wasserstand und -qualitat. Um Umweltinformationssystemen aufbauen zu konnen, braucht es ein Verstandnis dafur, welche Daten und Indikatoren mit welcher Methode fur welche Gebiete und in welcher Regelmasigkeit zu erheben sind. Eine Fulle an Umweltdaten wird heute mittels Satelliten- und Lasertechnik bereitgestellt. Um diese Daten systematisch erfassen und entsprechend verarbeiten zu konnen, benotigt man in vielen Handlungsfeldern des UBRM Methoden der Fernerkundung und Geoinformation. Umweltinformationssysteme konnen Daten auch fur eine grosere Gruppe an Interessierten, z.B. uber Social Media, bereitstellen. Je nach Zielgruppe aufbereitet, geben sie Auskunft uber die naturliche, gebaute und soziale Umwelt und deren Wechselwirkungen. Diese Umweltdaten konnen fur hoheitliche Aufgaben wie Raum- und Umweltplanung, fur private Unternehmen, etwa in der Logistikbranche, und fur die Gesellschaft, etwa bei Hochwasserwarnungen, als Entscheidungsgrundlage herangezogen werden.

Within the SMIRES COST Action (Science and Management of Intermittent Rivers & Ephemeral Streams, www.smires.eu), examples of intermittent rivers across Europe were collected, including gauged catchments with both natural and highly influenced river flow regimes. The examples will be published in a catalogue to give a good overview of the variety of intermittent streams in Europe. Information on the gauged intermittent rivers is summarised in a two-page template: The first page of the template describes the main characteristics of the catchments (land-use, geology, climate, etc.) and the river flow regime. Two panels display hydrographs and flow durations curves and a table gives metrics specific to river flow intermittence relevant for ecology. These hydrological indices have been computed by the SMIRES R package (https://homepage.boku.ac.at/h0540352/smires/framework.html) applied to daily discharge data. The second page of the template is dedicated to the description and reasons for intermittence. A short description about the spatio-temporal pattern of zero-flow events (encompassing seasonal behaviour, observed long-term trends, locations with frequently observed zero-flow events along the river network, etc.). The monitoring network, including gauging stations and other types of observations (e.g. visual inspection of the flow states at different locations along the river) in the catchment, is also described. This presentation gives examples of the pages in the catalogue and a preliminary classification of intermittent rivers at the European scale.

The objective of this paper is to present a new strategy for assessing climate impacts on low flows and droughts. The strategy is termed a three-pillar approach as it combines different sources of information. The first pillar, trend extrapolation, exploits the temporal patterns of observed low flows and extends them into the future. The second pillar, rainfall–runoff projections uses precipitation and temperature scenarios from climate models as an input to rainfall–runoff models to project future low flows. The third pillar, stochastic projections, exploits the temporal patterns of observed precipitation and air temperature and extends them into the future to drive rainfall–runoff projections. These pieces of information are combined by expert judgement based on a synoptic view of data and model outputs, taking the respective uncertainties of the methods into account. The viability of the approach is demonstrated for four example catchments from Austria that represent typical climate conditions in Central Europe. The projections differ in terms of their signs and magnitudes. The degree to which the methods agree depends on the regional climate and the dominant low flow seasonality. In the Alpine region where winter low flows dominate, trend projections and climate scenarios yield consistent projections of increasing low flows, although of different magnitudes. In the region north of the Alps, consistently small changes are projected by all methods. In the regions in the South and Southeast, more pronounced and mostly decreasing trends are projected but there is disagreement in the magnitudes of the projected changes. These results suggest that conclusions drawn from only one pillar of information would be highly uncertain. The three-pillar approach offers a systematic framework of combining different sources of information aiming at more robust projections than obtained from each pillar alone.

The main objective of the paper is to understand the contributions to the uncertainty in low-flow projections resulting from hydrological model uncertainty and climate projection uncertainty. Model uncertainty is quantified by different parameterisations of a conceptual semi-distributed hydrologic model (TUWmodel) using 11 objective functions in three different decades (1976–1986, 1987–1997, 1998–2008), which allows for disentangling the effect of the objective function-related uncertainty and temporal stability of model parameters. Climate projection uncertainty is quantified by four future climate scenarios (ECHAM5-A1B, A2, B1 and HADCM3-A1B) using a delta change approach. The approach is tested for 262 basins in Austria. The results indicate that the seasonality of the low-flow regime is an important factor affecting the performance of model calibration in the reference period and the uncertainty of Q95 low-flow projections in the future period. In Austria, the range of simulated Q95 in the reference period is larger in basins with a summer low-flow regime than in basins with a winter low-flow regime. The accuracy of simulated Q95 may result in a range of up to 60 % depending on the decade used for calibration. The low-flow projections of Q95 show an increase of low flows in the Alps, typically in the range of 10–30 % and a decrease in the south-eastern part of Austria mostly in the range −5 to −20 % for the climate change projected for the future period 2021–2050, relative the reference period 1978–2007. The change in seasonality varies between scenarios, but there is a tendency for earlier low flows in the northern Alps and later low flows in eastern Austria. The total uncertainty of Q95 projections is the largest in basins with a winter low-flow regime and, in some basins the range of Q95 projections exceeds 60 %. In basins with summer low flows, the total uncertainty is mostly less than 20 %. The ANOVA assessment of the relative contribution of the three main variance components (i.e. climate scenario, decade used for model calibration and calibration variant representing different objective function) to the low-flow projection uncertainty shows that in basins with summer low flows climate scenarios contribute more than 75 % to the total projection uncertainty. In basins with a winter low-flow regime, the median contribution of climate scenario, decade and objective function is 29, 13 and 13 %, respectively. The implications of the uncertainties identified in this paper for water resource management are discussed.

International audience; It is generally accepted that drought is one of the most costly weather-related natural hazards. In 2015, a long-lasting drought hit Europe, particularly affecting central and eastern Europe. In some regions it was the driest (North Slovakia) and in others (Czech Republic and Poland) it was the second driest summer of the last 50 years (following 2003). Key questions are: (i) how extreme are these events, not only in terms of hydro-meteorological characteristics but also impacts? and (ii) how are these impacts managed?

The objective of this paper is to assess the performance of methods that predict low flows and flood runoff in ungauged catchments. The aim is to learn from the similarities and differences between catchments in different places, and to interpret the differences in performance in terms of the underlying climate-landscape controls. The assessment is performed at two levels. The Level 1 assessment is a meta-analysis of 14 low flow prediction studies reported in the literature involving 3112 catchments, and 20 flood prediction studies involving 3023 catchments. The Level 2 assessment consists of a more focused and detailed analysis of individual basins from selected studies from Level 1 in terms of how the leave-one-out cross-validation performance depends on climate and catchment characteristics as well as on the regionalisation method. The results indicate that both flood and low flow predictions in ungauged catchments tend to be less accurate in arid than in humid climates and more accurate in large than in small catchments. There is also a tendency towards a somewhat lower performance of regressions than other methods in those studies that apply different methods in the same region, while geostatistical methods tend to perform better than other methods. Of the various flood regionalisation approaches, index methods show significantly lower performances in arid catchments than regression methods or geostatistical methods. For low flow regionalisation, regional regressions are generally better than global regressions.