Search

• The impact of internal variability and climate change on streamflow is studied in 133 catchments of different sizes. • Internal climate variability is important for masking anthropogenic climate change, especially in the near future. • The size of the catchment used to have a significant impact on the timing of the emergence of the extreme flow. • Smaller catchments experience floods (droughts) at earlier (later) times. • Instead of KS-test and standard error of the mean (SEM), signal-to-noise (SNR) approaches are too conservative to detect the timing of emergence. Uncertainty about the future impacts of climate change represents a significant barrier to implementing adaptation measures. This work explores the impact of internal climate variability on streamflow projections for 133 catchments across the eastern and northeastern United States. Using data from a single model initial-condition large ensemble (SMILE) at high spatial and temporal resolution, this work assesses the magnitude of anthropogenic climate change and internal climate variability on projected future streamflow. The impact of catchment size is studied by grouping catchments into three different size classes (less than500 km2, between 500 and 1000 km2, and greater than 1000 km2). Results show that in a warmer climate, low to middle quantiles of future streamflow will systematically decrease, while the upper quantiles will increase. Increases are largest for more extreme streamflow indices. Using three different approaches, the role of internal variability is studied to estimate the time of emergence (TOE). In this case, results show that the climate change signal of extreme floods and droughts emerges later than that of median flow quantiles, even though the changes for floods as droughts is more significant. There is a clear relationship between catchment size and TOE, with small catchments seeing an earlier TOE for floods, and a later one for droughts. These results provide insight into adaptation times for small to large watersheds. [ABSTRACT FROM AUTHOR]

Highlights • Calibration on full time series is shown to be more robust than split-sample methods. • 30 Bootstrapping tests on 6 cases provide evidence towards this method being optimal. • Verification on 10 independent catchment-model pairs support the conclusions. • Caveats of split-sampling on model performance are demonstrated. • Length of the calibration period is proportional to the parameter set robustness. Abstract This paper investigates the issues related to the use of validation in hydrological model calibration. Traditionally, models are calibrated and then assessed on an independent period (split-sample) to determine their adequacy in simulating streamflow as compared to observations. In this study, two hydrological models and three North American catchments are used to evaluate the effects of using validation to assess the model parameters' robustness on the model's actual simulation capabilities and accuracy in simulating streamflow. The length of the calibration period is increased from 1 to 16 years, and for each case a large number of randomly selected combinations of years are used for calibration and for validation using the Nash-Sutcliffe Efficiency metric. The calibrated model is then run on an independent 8-year test-period to assess the model's actual performance in simulation mode in unknown conditions. The process is bootstrapped 30 times to ensure the robustness of the results. The tests pit the calibration/validation methods on increasing calibration period lengths against a full calibration on the entire available dataset. Results show that the calibration on the full dataset is the optimal strategy as it generates the most robust parameter sets, provides the best model accuracy on an independent testing period and does not require assumption-making on the modeler's part. The calibrated parameter sets for each test-case were evaluated using the relative bias and correlation metrics, which revealed that the method transfers well to these two other metrics. Results also demonstrate the pitfalls of the commonly used split-sampling strategy, where good parameter sets may be discarded due to model performance discrepancies between calibration and validation periods. The conclusions point to the need to use as many years as possible in the calibration step and to entirely disregard the validation aspect under certain conditions. [ABSTRACT FROM AUTHOR]

• Future change in global extreme meteorological droughts. • Two climate model large ensembles were used to assess the change of frequency. • Changes in meteorological droughts are mostly correlated to mean precipitation. • Projected increases in frequency are much more severe for short-term droughts. • Total area affected by a worsening of droughts gets larger with the return periods. Several studies have looked at the evolution of droughts in a changing climate in various locations and have typically found that droughts are expected to become more severe in the future. However, relatively little work has been done on extreme droughts. This study focuses on the evolution of extreme meteorological droughts over near (2036–2065) and far (2071–2100) future time periods. Monthly precipitation from two large multi-member climate model ensembles: the Canadian Earth System Model (CanESM2) 50-member ensemble and the Community Earth System Model (CESM1) 40-member ensemble, was used to calculate the Standard Precipitation Index (SPI) to quantify and explore the evolution of future meteorological droughts. The use of these large ensembles allows for the calculation of the 20- and 100-year droughts return period and offers a unique perspective into those rare events. Both climate models project increases in extreme meteorological drought frequency over many of the world's regions. The spatial patterns of regions with worsening droughts generally match those of projected change in mean annual precipitation, although the former is more extensive, indicating that changes in internal variability will increase drought frequency even in some regions projected to see increased mean annual precipitation. The projected increase in meteorological drought frequency is more significant for short-term JJA droughts, and for the larger return periods. Large increases in frequency are observed in many regions, all the way up to 20 times for the historical100-year JJA drought indicating a return period shift from 100 to 5 year. Considering the significant predicted increases in temperature, it follows that future changes for both agricultural (soil water deficit) and hydrological droughts (low stream levels), which result from the interaction between precipitation deficit and temperature (as well as other variables), will be even more severe. [ABSTRACT FROM AUTHOR]

Weighting climate models is controversial in climate change impact studies using an ensemble of climate simulations from different climate models. In climate science, there is a general consensus that all climate models should be considered as having equal performance or in other words that all projections are equiprobable. On the other hand, in the impacts and adaptation community, many believe that climate models should be weighted based on their ability to better represent various metrics over a reference period. The debate appears to be partly philosophical in nature as few studies have investigated the impact of using weights in projecting future climate changes. The present study focuses on the impact of assigning weights to climate models for hydrological climate change studies. Five methods are used to determine weights on an ensemble of 28 global climate models (GCMs) adapted from the Coupled Model Intercomparison Project Phase 5 (CMIP5) database. Using a hydrological model, streamflows are computed over a reference (1961–1990) and future (2061–2090) periods, with and without post-processing climate model outputs. The impacts of using different weighting schemes for GCM simulations are then analyzed in terms of ensemble mean and uncertainty. The results show that weighting GCMs has a limited impact on both projected future climate in term of precipitation and temperature changes and hydrology in terms of nine different streamflow criteria. These results apply to both raw and post-processed GCM model outputs, thus supporting the view that climate models should be considered equiprobable. [ABSTRACT FROM AUTHOR]

Reanalyses are important sources of meteorological data. Recent studies have shown that precipitation and temperature data from reanalysis present a strong potential for hydrological modelling, especially in regions with a sparse observational network. The objective of this study is to evaluate the impacts of the combination of three global atmospheric reanalyses – ERA-Interim, CFSR and MERRA – and one gridded observation dataset on the accuracy of hydrological model discharge simulations. Two combination approaches were used. The first one combined reanalyses and the observational database using a weighted average of the precipitation and temperature inputs. The second one consisted in using all meteorological inputs separately and combining the simulated hydrographs. The combinations were performed over 460 Canadian watersheds (representing regions with a low density of weather stations) and 370 US watersheds (representing regions with a higher density of weather stations). Results showed significant improvements in the simulated discharges for 68% and 92% of the Canadian watersheds for the input combinations and output combinations, respectively. Moreover, both approaches led to significant improvements in the simulated discharges for 72% of the US watersheds studied. For all watersheds where simulated discharges using observational data had a Nash Sutcliffe efficiency (NSE) lower than 0.5, the combination with reanalyses resulted in a median NSE increase of 0.3. This indicates that reanalysis can successfully compensate for deficiencies in the surface observation record and provide significantly better hydrological modelling performance. [ABSTRACT FROM AUTHOR]

Summary: Statistical and dynamical downscaling techniques have been proposed to bridge the gaps between coarse-scale and generally biased climate model outputs and the point-scale requirements of impact model inputs. Amongst the various statistical approaches, empirical downscaling methods are the most commonly used due to their ease of implementation. Several empirical downscaling approaches have been proposed and need to be assessed as to which method contributes (or not) to the overall climate change uncertainty. Accordingly, this work aims at assessing the uncertainty of six empirical downscaling methods in quantifying the hydrological impact of climate change over two North American river basins under different climate conditions. The six empirical downscaling methods are grouped into change factor (two methods) and bias correction (four methods) approaches. The uncertainty related to the choice of an empirical downscaling method is compared to that associated with the choice of climate simulation, through the use of two Regional Climate Models (RCMs) driven by three different General Circulation Models (GCMs), totaling four RCM simulations, taken from the NARCCAP inter-comparison project. The future (2041–2065) hydrological regimes simulated with an empirical lumped hydrology model (HSAMI) are compared to the reference period (1971–1995) using a set of hydrology criteria which includes statistics of both mean and extreme values. The results show a large uncertainty envelope associated with the choice of a given empirical downscaling method, as well as for the choice of an RCM simulation. The uncertainty due to empirical downscaling and RCM simulation was more significant in projecting extreme streamflow than in projecting mean flows. Comparing the uncertainty envelope of empirical downscaling methods to the envelope resulting from four RCM simulations indicates that both are similar, even though the latter was slightly larger for some statistics. Finally, the uncertainty linked to the choice of an empirical downscaling approach (change factor vs. bias correction) was much larger than within each type. Overall, this work emphasizes the importance of using several climate projections and empirical downscaling approaches to delineate uncertainty when assessing the climate change impacts on hydrology. [Copyright &y& Elsevier]

Summary: This paper investigates the effects of model structure and parameter equifinality on the uncertainty related to hydrological modelling in climate change impact studies. The study is conducted on a snow-dominated watershed located in the southern part of the province of Quebec (Canada). Hydrological model structure uncertainty is examined through the use of two very different simulation tools, one lumped conceptual model and one spatially-distributed physically-based model. Parameter equifinality is examined by performing multiple automatic calibrations with both hydrological models. The analysis is first carried out under recent past climate and then under modified climate conditions following two contrasted projections that are analysed separately. The delta change approach is used to build the two climate projections. Overall, this study reveals that the impact of hydrological model structure uncertainty is more significant than the effect of parameter uncertainty, under recent past climate as well as future climate conditions. Ultimately, the use of hydrological models with different levels of complexity should be considered as part of the global uncertainty related to hydrological model structure. [Copyright &y& Elsevier]

Summary: Uncertainty estimation of climate change impacts has been given a lot of attention in the recent literature. It is generally assumed that the major sources of uncertainty are linked to General Circulation Models (GCMs) and Greenhouse Gases Emissions Scenarios (GGES). However, other sources of uncertainty such as the choice of a downscaling method have been given less attention. This paper focuses on this issue by comparing six downscaling methods to investigate the uncertainties in quantifying the impacts of climate change on the hydrology of a Canadian (Quebec province) river basin. The downscaling methods regroup dynamical and statistical approaches, including the change factor method and a weather generator-based approach. Future (2070–2099, 2085 horizon) hydrological regimes simulated with a hydrological model are compared to the reference period (1970–1999) using the average hydrograph, annual mean discharge, peak discharge and time to peak discharge as criteria. The results show that all downscaling methods suggest temperature increases over the basin for the 2085 horizon. The regression-based statistical methods predict a larger increase in autumn and winter temperatures. Predicted changes in precipitation are not as unequivocal as those of temperatures, they vary depending on the downscaling methods and seasons. There is a general increase in winter discharge (November–April) while decreases in summer discharge are predicted by most methods. Consistently with the large predicted increases in autumn and winter temperature, regression-based statistical methods show severe increases in winter flows and considerable reductions in peak discharge. Across all variables, a large uncertainty envelope was found to be associated with the choice of a downscaling method. This envelope was compared to the envelope originating from the choice of 28 climate change projections from a combination of seven GCMs and three GGES. Both uncertainty envelopes were similar, although the latter was slightly larger. The regression-based statistical downscaling methods contributed significantly to the uncertainty envelope. Overall, results indicate that climate change impact studies based on only one downscaling method should be interpreted with caution. [Copyright &y& Elsevier]

Summary: Weather generators are computer models that produce time series of meteorological data that have similar statistical properties as that of observed data. The past decade has seen a sharp and renewed increase in interest in weather generators, linked to their potential use in climate change studies. One appealing property of weather generators is their ability to rapidly produce time series of unlimited length, thus permitting impact studies of rare occurrences of meteorological variables. However, one problem with daily weather generators is that they underestimate monthly and inter-annual variances because they do not take into account the low-frequency component of climate variability. This research aims to present an approach for correcting the low-frequency variability of weather variables for weather generator and to assess its ability to reproduce key statistical parameters at the daily, monthly and yearly scales. The approach is applied to precipitation which is usually the variable displaying the largest inter-annual variability. The daily stochastic precipitation model is a Richardson-based weather generator that uses a first-order two-state Markov chain for precipitation occurrence and a gamma distribution for precipitation amounts. Low-frequency variability was modeled based on observed power spectra of monthly and annual time series.Generation of synthetic monthly and yearly precipitation data was achieved by assigning random phases for each spectral component.This preserved the power spectra, variances and the autocorrelation functions at the monthly and annual scales. The link to daily parameters was established through linear functions.The quality of these corrections was assessed through direct and indirect validation tests, with the direct validation focusing on comparing the means, standard deviations and autocorrelations of different weather series. The results showed that standard deviations of both monthly and annual precipitations were produced almost exactly. The proposed method also preserved the autocorrelation of annual precipitation. The indirect validation involved modelling the discharge of a river basin using a hydrological model driven by different precipitation series. The results showed that the corrected weather series significantly improved the variability of simulated flow discharges at the monthly and annual scales compared to those simulated using the data generated by the standard weather generator. [Copyright &y& Elsevier]

Summary: The impact of climate change on the hydrology of the Chute-du-Diable watershed (Quebec, Canada) is studied by comparing statistics on current and projected future discharge resulting from a wide range of climate change scenarios. The use of 10 equally weighted climate projections from a combination of 5 general circulation models (GCMs) and 2 greenhouse gas emission scenarios (GHGES) allows for the definition of an uncertainty envelope of future hydrologic variables. GCM data is downscaled using the change factor approach for 30-year time slices centered around years 2020, 2050 and 2080. To estimate natural variability, synthetic time series are then computed for each horizon and for each climate change scenario, using a stochastic weather generator (30 series of 30 years), and are entered into a hydrology model. Future hydrological regimes are then compared to the control period (1961–1990) using the annual and seasonal mean discharge, peak discharge and timing of peak discharge criteria. Results indicate a 1–14°C increase in seasonal temperature and a −9 to +55% change in seasonal precipitation. The largest increases in both temperature and precipitation are observed in the winter and spring seasons. The main hydrologic impact observed is a spring flood appearing 1–5 weeks earlier than usual and a modification of its amplitude from −40 to +25%. Most scenarios suggest increases in the winter, spring and fall discharge, whereas summer is expected to see a decrease in discharge. While there is still a large scatter in projected values, the uncertainty analysis projects a better view of the most probable future hydrologic behaviour of the watershed. Of all sources of uncertainty considered in this study, the largest comes from the choice of a GCM. Accordingly, all impact studies based on results from only one GCM should be interpreted with caution. [Copyright &y& Elsevier]

• Stochastic weather generator is used a main computational tool for taking climate non-stationarity into account. • Parameters of stochastic weather generator are conditioned on the most relevant subset of large-scale climate indices. • Long-term streamflow forecast can be improved, particularly around spring flood. This paper aims at improving long-term streamflow forecasts by implementing a novel technique based on conditioning the parameters of a stochastic weather generator on large-scale climate indices, with varying lengths of training periods during the establishment of correlations. The most important climate indices are identified by looking at yearly correlations between a set of 40 indices and meteorological data (precipitation and temperature) at the watershed scale. A linear model is then constructed to identify precipitation and temperature anomalies to induce perturbations in the stochastic weather generator. Time windows of 5, 10, 15, 20 and 30 years are used in determining the optimal linear model. The performance of the proposed approach is assessed against that of a resampling of past climatology and using the same stochastic weather generator unconditioned on climate indices. Each member of the ensemble weather forecast is then fed to a hydrological model to create the Ensemble Streamflow Forecasts (ESF) with a one-year forecasting horizon. The three approaches are tested in hindcast mode over a 30-year period at 12 forecast dates. Results show that temperatures are significantly correlated with large-scale climate indices, whereas precipitation is only weakly related to the same indices. The length of the time window has a considerable impact on the prediction ability of the linear models. The precipitation models based on short duration time windows performed better than those based on longer windows, while the reverse was found for the temperature models. A comparison between all three Ensemble Streamflow Forecast approaches is assessed using the Continuous Ranked Probability Score (CRPS) metric. Results show that the proposed method improves long-term streamflow forecasting, particularly the volumetric bias and the peak flows during the spring flood. [ABSTRACT FROM AUTHOR]

• Future change in extreme droughts over 5797 North American watersheds was studied. • Two climate model large ensembles were used to assess the change of frequency. • Changes in meteorological droughts are mostly correlated to mean precipitation. • Large increases in the frequency of hydrological droughts are observed over NA. • Projected changes in future extreme droughts get worse for longer return periods. This paper describes projected frequency changes in extreme summer meteorological and hydrological droughts over North American catchments. It uses two large ensemble climate models (50-member CanESM2 and 40-member CESM1) under the RCP 8.5 scenario to robustly assess frequency changes up to the 100-year drought, relative to the 1980–2009 reference period. Meteorological droughts, linked solely to precipitation deficit, are represented by the 1-month Standardized Precipitation Index (SPI), whereas hydrological droughts are characterized by the 1-month Streamflow Drought Index (SDI), based on hydrological simulation using a lumped hydrological model. Results show widely different patterns for future changes in extreme meteorological versus hydrological droughts. Future meteorological droughts show patterns of increasing and decreasing frequency that roughly match future expected changes of mean summer precipitation, although some regions are nonetheless projected to see more frequent extreme meteorological droughts despite increases in mean summer precipitation. The frequency changes for the 100-year meteorological droughts can be quite severe, with up to a 20-fold increase observed over some watersheds. On the other hand, hydrological droughts, which combine the effect of precipitation and temperature changes, show a mostly uniform pattern of large to very large increases in drought frequency. This shows that the projected temperature increase is a main driver of future extreme hydrological droughts, enough to overcome the projected increase in mean summer precipitation for many North American catchments. Projected changes in both meteorological and hydrological droughts get consistently worse for the longer considered return periods. In other words, frequency changes for the 100-year droughts are more significant than those expected for the 2- and 20-year droughts. This gradual worsening toward larger extremes has potentially large societal and economic impacts. The large projected increases in the frequency of extreme hydrological drought frequency (up to 27 times) are likely to severely stress water management systems across North America. [ABSTRACT FROM AUTHOR]

Bias correction is usually implemented prior to using climate model outputs for impact studies. However, bias correction methods that are commonly used treat climate variables independently and often ignore inter-variable dependencies. The effects of ignoring such dependencies on impact studies need to be investigated. This study aims to assess the impacts of correcting the inter-variable correlation of climate model outputs on hydrological modeling. To this end, a joint bias correction (JBC) method which corrects the joint distribution of two variables as a whole is compared with an independent bias correction (IBC) method; this is considered in terms of correcting simulations of precipitation and temperature from 26 climate models for hydrological modeling over 12 watersheds located in various climate regimes. The results show that the simulated precipitation and temperature are considerably biased not only in the individual distributions, but also in their correlations, which in turn result in biased hydrological simulations. In addition to reducing the biases of the individual characteristics of precipitation and temperature, the JBC method can also reduce the bias in precipitation-temperature (P-T) correlations. In terms of hydrological modeling, the JBC method performs significantly better than the IBC method for 11 out of the 12 watersheds over the calibration period. For the validation period, the advantages of the JBC method are greatly reduced as the performance becomes dependent on the watershed, GCM and hydrological metric considered. For arid/tropical and snowfall-rainfall-mixed watersheds, JBC performs better than IBC. For snowfall- or rainfall-dominated watersheds, however, the two methods behave similarly, with IBC performing somewhat better than JBC. Overall, the results emphasize the advantages of correcting the P-T correlation when using climate model-simulated precipitation and temperature to assess the impact of climate change on watershed hydrology. However, a thorough validation and a comparison with other methods are recommended before using the JBC method, since it may perform worse than the IBC method for some cases due to bias nonstationarity of climate model outputs. [ABSTRACT FROM AUTHOR]

Summary Climate data measured by weather stations are crucially important and regularly used in hydrologic modeling. However, they are not always available due to the low spatial density and short record history of many station networks. To overcome these limitations, gridded datasets have become increasingly available. They have excellent continuous spatial coverage and no missing data. However, these datasets are usually interpolated using station data, with little new information besides elevation. Furthermore, minimal validation has been done on most of these datasets. This study compares three such datasets covering the continental United States to evaluate their differences and their impact on lumped hydrological modeling. Three daily time step gridded datasets with resolutions varying between 0.25° and 1 km were used in this study – Santa-Clara, Daymet and CPC. The hydrological modeling evaluation of these datasets was performed over 424 basins from the MOPEX database. Results show that there are significant differences between the datasets, even though they were essentially all interpolated from almost the same climate databases. Despite those differences, the hydrological model used in this study was able to perform equally well after a specific calibration to each dataset. While there were a few exceptions, by and large, Nash–Sutcliffe efficiency metrics obtained in validation were not statistically different from one database to the other for most basins. It appears that there are no reasons to favor one dataset versus another for lumped hydrological modeling, and that these datasets perform just as well as using the original station data. [ABSTRACT FROM AUTHOR]

This paper evaluated the effects of parametric reduction of a hydrological model on five regionalization methods and 267 catchments in the province of Quebec, Canada. The Sobol’ variance-based sensitivity analysis was used to rank the model parameters by their influence on the model results and sequential parameter fixing was performed. The reduction in parameter correlations improved parameter identifiability, however this improvement was found to be minimal and was not transposed in the regionalization mode. It was shown that 11 of the HSAMI models’ 23 parameters could be fixed with little or no loss in regionalization skill. The main conclusions were that (1) the conceptual lumped models used in this study did not represent physical processes sufficiently well to warrant parameter reduction for physics-based regionalization methods for the Canadian basins examined and (2) catchment descriptors did not adequately represent the relevant hydrological processes, namely snow accumulation and melt. [ABSTRACT FROM AUTHOR]

• Flood volume evolution under climate change is evaluated on 1403 catchments. • Multiple climate models, hydrological models and bias correction methods are used. • Uncertainty analysis showed climate and hydrological models as key contributors. • Approximately 80% of basins will see increased extreme flood volumes in the future. • Changes seem to follow a pattern that can help predict flood volume evolution. Climate change is expected to increase the frequency and intensity of extreme events, such as droughts and floods. Assessing the impacts of climate change on flood volumes is crucial to provide better management of flooding disasters given the devastating consequences they can cause. Using flood volume instead of flood peak is critical because the latter focuses on the highest possible discharge observed during a flood event, while flood volume also considers flow duration, which is an important factor in terms of the hazard caused to the surrounding environment. This study aims to evaluate the overall impact of climate change on floods caused by long-duration flows exceeding synthetic flooding thresholds. These flows are used to compute flood storage with various flow thresholds increasing from the 50th to 95th percentile of annual maximum observed discharge over a large sample of 1403 North American catchments. This study also aims to evaluate the contribution of each uncertainty source of the ensemble approach (climate models, bias-correction methods, and hydrological models) on future flood volumes. The results show that flood volumes are expected to decrease in western mountainous areas, the Great Lakes region, and the Maritimes, while increases are expected over most of eastern North America. The study finds that climate models contribute the most to the variance of flood volume uncertainty, followed by hydrological models. Overall, this study provides projected flood volume changes for North American catchments from a comprehensive ensemble that includes eleven climate models driven by two RCP scenarios and four hydrological models of varying complexity. This leads to a large-sample assessment of future flood volumes that could be useful to policymakers for making better-informed decisions in flood risk management. [ABSTRACT FROM AUTHOR]

• A method to quantify the maximum theoretical skill of ESP forecasts is proposed. • Removing ensemble members in ESP can lead to improvements in forecasted inflow volume. • Multiple ESP member selection methods are compared and evaluated. • Properties of desirable/undesirable forecast members are analyzed to improve predictors. Streamflow forecasting is a crucial task for hydropower operations, flood forecasting and water resource use optimization. Ensemble Streamflow Prediction (ESP) is commonly used in the case of long-term lead times (a few months or seasons ahead). ESP covers the use of historical meteorological scenarios in driving a hydrological model to generate an ensemble of possible future streamflow. Many studies have evaluated methods for selecting optimal subsets of scenarios to improve forecasting skill, and indeed, this is still an ongoing area of research. In this study, we propose a procedure that calculates the maximum potential skill of a classic ESP forecast. A genetic algorithm (GA) is used to determine the best possible set of climatological scenarios given any ensemble size. Along with providing a direct estimate of the ESP forecasting potential in hindcast experiments, the method can be used as a reference for comparing other methods to ESP. The procedure is also used to compare classical ESP, a well-established forecasting method, with two new methods, namely, the Analogue method and the Contingency Table (CT) approach. A discriminant analysis is finally implemented to attempt to identify key features of ESP members that performed well as compared to their counterparts using historic climatology and climate indices. It is shown while there exists a potential for improvement, a lot of research must still be realized to exploit this potential. The procedure was tested over two basins in Canada. In general, results showed that for any forecast date, decreasing the ensemble size led to a higher potential for better forecasting skills. However, the method does not yet allow identifying the subset of the entire climatology to be used to maximize the ESP forecast performance. [ABSTRACT FROM AUTHOR]

• A Long Short Term Memory neural network is applied to forecast streamflow. • The LSTM is compared to a distributed hydrological model using data assimilation. • The forecast skill of the LSTM is better for lead-times up to 7 days. • The LSTM does not require data assimilation when fed recent observed flow data. Streamflow forecasting is crucial in water planning and management. Physically-based hydrological models have been used for a long time in these fields, but improving forecast quality is still an active area of research. Recently, some artificial neural networks have been found to be effective in simulating and predicting short-term streamflow. In this study, we examine the reliability of Long Short-Term Memory (LSTM) deep learning model in predicting streamflow for lead times of up to ten days over a Canadian catchment. The performance of the LSTM model is compared to that of a process-based distributed hydrological model, with both models using the same weather ensemble forecasts. Furthermore, the LSTM's ability to integrate observed streamflow on the forecast issue date is compared to the data assimilation process required for the hydrological model to reduce initial state biases. Results indicate that the LSTM model forecasted streamflows are more reliable and accurate for lead-times up to 7 and 9 days, respectively. Additionally, it is shown that the LSTM model using recent observed flows as a predictor can forecast flows with smaller errors in the first forecasting days without requiring an explicit data assimilation step, with the LSTM model generating a median value of mean absolute error (MAE) for the first day of lead-time across all forecast issue dates of 25 m3/s compared to 115 m3/s for the assimilated hydrological model. [ABSTRACT FROM AUTHOR]

• Reducing initial forecasting errors using the EnKF data assimilation technique. • Improving real-time streamflow forecasts by an optimal EnKF implementation. • A seasonal adjustment of the EnKF hyperparameters improves forecast accuracy. • The vadose zone state variable is the most impactful and should be prioritized. • Data assimilation is most sensitive to temperature uncertainty for forecasting. Data Assimilation (DA) is an important step for improving prediction accuracy and real-time correction of hydrological models for operational forecasting purposes. The Ensemble Kalman filter (EnKF) is one of the most popular techniques used to address the issues of updating the model states and parameters by creating a novel set of initial conditions in real-time. This study aims at identifying optimal seasonal EnKF parametrizations to reduce the uncertainty associated with the initial conditions in a semi-distributed hydrological model of a snow-dominated catchment in Canada. Sensitivity analysis is performed to evaluate the effects of the EnKF individual hyperparameters (temperature, precipitation and inflow uncertainties) and the updating of three state variables (vadose zone, saturated zone and snowpack) on the skill of short-term (up to 9 days lead-time) forecasts. Results show that the performance of the forecasts is sensitive to the individual hyperparameters and particularly so to the temperature uncertainty, which varies between seasons. Additionally, the forecast skill is related to the choice of the state variables to be updated depending on the season. The vadose zone state variable displays higher importance and sensitivity than the other states, and the results indicate that all state variables should not be systematically updated on this catchment and with the implemented hydrological model. Finally, combining the best hyperparameter values with the optimal combination of state variables to update, insight is provided on the success of the data assimilation scheme that is evaluated on a rolling horizon using a set of seasonal rules to gain forecast performance over the span of multiple years. [ABSTRACT FROM AUTHOR]

This paper presents the PAVICS-Hydro platform, the result of long-term strategic considerations and compromises between functionality and ease-of-use, which is developed to provide users with direct access to a comprehensive hydroclimatic modeling and analysis framework over North America. This includes a Python hydrological modelling package, efficient access to meteorological data, a large array of post-processed hydroclimatic data, numerical weather forecasts and bias-corrected CMIP6 climate projections at the scale of 5797 catchments, thus providing the basic components necessary for hydrological modeling and forecasting, as well as climate change impact studies. This paper presents the principles on which the PAVICS-Hydro platform was developed while explaining the choices behind its key features, with a specific focus on lessons learned for creators of the next generations' scientific research platforms. A case study using PAVICS-Hydro is given to illustrate how this platform can be used to support different needs in the field of water resources. • We present the PAVICS-Hydro platform for hydrological modelling and research. • The platform can be used for simulation, forecasting and climate change studies. • It helps automate repetitive tasks such as model building and data processing. • Access to multiple large-scale hydrometeorological and climate datasets is provided. • The platform is built on open-source software and free to use. [ABSTRACT FROM AUTHOR]

Projected climate change effects on snow hydrology are investigated for the 2041–2060 horizon following the SRES A2 emissions scenario over three snowmelt-dominated catchments in Quebec, Canada. A 16-member ensemble of eight snow models (SM) simulations, based on the high-resolution Canadian Regional Climate Model (CRCM-15 km) simulations driven by two realizations of the Canadian Global Climate Model (CGCM3), is established per catchment. This study aims to compare a range of SMs in their ability at simulating snow processes under current climate, and to evaluate how they affect the assessment of the climate change-induced snow impacts at the catchment scale. The variability of snowpack response caused by the use of different models within two different SM approaches (degree-day (DD) versus mixed degree-day/energy balance (DD/EB)) is also evaluated, as well as the uncertainty of natural climate variability. The simulations cover 1961–1990 in the present period and 2041–2060 in the future period. There is a general convergence in the ensemble spread of the climate change signals on snow water equivalent at the catchment scale, with an earlier peak and a decreased magnitude in all basins. The results of four snow indicators show that most of the uncertainty arises from natural climate variability (inter-member variability of the CRCM) followed by the snow model. Both the DD and DD/EB models provide comparable assessments of the impacts of climate change on snow hydrology at the catchment scale. [ABSTRACT FROM AUTHOR]

Summary This study aims to test whether a weighted combination of several hydrological models can simulate flows more accurately than the models taken individually. In addition, the project attempts to identify the most efficient model averaging method and the optimal number of models to include in the weighting scheme. In order to address the first objective, streamflow was simulated using four lumped hydrological models (HSAMI, HMETS, MOHYSE and GR4J-6), each of which were calibrated with three different objective functions on 429 watersheds. The resulting 12 hydrographs (4 models × 3 metrics) were weighted and combined with the help of 9 averaging methods which are the simple arithmetic mean (SAM), Akaike information criterion (AICA), Bates–Granger (BGA), Bayes information criterion (BICA), Bayesian model averaging (BMA), Granger–Ramanathan average variant A, B and C (GRA, GRB and GRC) and the average by SCE-UA optimization (SCA). The same weights were then applied to the hydrographs in validation mode, and the Nash–Sutcliffe Efficiency metric was measured between the averaged and observed hydrographs. Statistical analyses were performed to compare the accuracy of weighted methods to that of individual models. A Kruskal–Wallis test and a multi-objective optimization algorithm were then used to identify the most efficient weighted method and the optimal number of models to integrate. Results suggest that the GRA, GRB, GRC and SCA weighted methods perform better than the individual members. Model averaging from these four methods were superior to the best of the individual members in 76% of the cases. Optimal combinations on all watersheds included at least one of each of the four hydrological models. None of the optimal combinations included all members of the ensemble of 12 hydrographs. The Granger–Ramanathan average variant C (GRC) is recommended as the best compromise between accuracy, speed of execution, and simplicity. [ABSTRACT FROM AUTHOR]

• The impacts of hydrological model components on performance and uncertainty of streamflow simulations are investigated. • A multi-model approach over 698 North America catchments is used. • No clear relationship between the structural and parametric complexity of the hydrological components and model performance is found. • No single model structure performs best in all climate zones and for all streamflow metrics. The impacts of hydrological model components on performance and uncertainty of streamflow simulations are investigated using a multi-model approach over 698 catchments in North America. An ensemble of 180 different hydrological model structures is obtained by combining five snow models (SM), four potential evapotranspiration formulations (PET), three vertical (VFS) and three horizontal (HFS) flow rounting schemes. This study aims to first compare the multiple component combinations of SM, PET, VFS and HFS in terms of their ability at simulating streamflows for ten Köppen climate zones. A secondary goal is to evaluate the contribution of each model component to the total variance of the streamflow response according to the model structure for three streamflow metrics representing mean, high and low flows. Results indicate that there is no clear relationship between the structural and parametric complexity of the hydrological components and model performance for all three streamflow metrics. There is no single model structure which performs best in all climate zones and for all streamflow metrics. Streamflow uncertainty generated by the 180 synthetic hydrological model structures depends primarily on the metrics under study, but also on the climate zone. The evapotranspiration formulation is only critical for the mean flow metric. For the low-flow criterion, the VFS generates the most uncertainty and is therefore the most critical, whereas for the high-flow metric, performance is mostly related to the HFS component and SM for Nordic catchments. A better understanding of how hydrological model components generate uncertainty for a given climate and streamflow metric will lead to more robust model structures and therefore reduce the uncertainty of hydrological simulations. [ABSTRACT FROM AUTHOR]

Summary Projected climate change effects on streamflow are investigated for the 2041–2070 horizon following the SRES A2 emissions scenario over two snowmelt-dominated catchments in Canada. A 16-member ensemble of SWAT hydrological model (HM) simulations, based on a comprehensive ensemble of the Canadian Regional Climate Model (CRCM) simulations driven by two global climate models (GCMs), with five realizations of the Canadian CGCM3 and three realizations of the German ECHAM5 is established per catchment. This study aims to evaluate, once model bias has been removed by statistical post-processing (SP), how the RCM-simulated climate changes differ from those of the parent GCMs, and how they affect the assessment of climate change-induced hydrological impacts at the catchment scale. The variability of streamflow caused by the use of different SP methods (mean-based versus distribution-based) within each statistical post-processing pathway of climate model outputs (bias correction versus perturbation) is also evaluated, as well as the uncertainty of natural climate variability. The simulations cover 1971–2000 in the reference period and 2041–2070 in the future period. For a set of criteria, results based on raw and statistically post-processed model outputs for the reference climate are compared with observations. This process demonstrates that SP is important not only for GCMs outputs, but also for CRCM outputs. SP leads to a high level of agreement between the CRCM and the driving GCMs in reproducing patterns of observed climate. The ensemble spread of the climate change signal on streamflow is large and varies with catchments and hydrological periods (winter/summer flows). The results of various hydrological indicators show that most of the uncertainty arises from the natural climate variability followed by the statistical post-processing. The uncertainty linked to the choice of statistical pathway is much larger than that associated with the choice of the method in quantifying the hydrological impacts. We find that the incorporation of dynamical downscaling of global models through the CRCM as an intermediate step in the GCM–RCM–SP–HM model chain does not lead to considerable differences in the assessment of the climate change impacts on streamflow for the study catchments. [ABSTRACT FROM AUTHOR]