Search

This study examined the impact of climate change on streamflow in the Andit Tid watershed using climate models of dynamically downscaled Ethiopia's CORDEX. The Arc SWAT and ArcGIS 10.5 software assessed the spatial and temporal distribution of streamflow, incorporating geospatial data like land use maps, digital elevation models, soil maps, and climate data. The SWAT model was calibrated and validated using SWAT-CUP with the SUFI-2 algorithm. The Canadian Centre for Climate Modeling and Analysis, Canada (CCCma (RCA4) model was selected for future projections after validation. From 1991 to 2021, the average streamflow rate was 0.0374 m3/s (247 mm), with R2 values of 0.83 for calibration and 0.72 for validation. Hotspots with active gullies and slopes over 20% were identified mainly in cultivated lands. Future projections indicated a comparable streamflow rate to current conditions at 0.0322 m3/s (212.6 mm). A decline in streamflow is projected: 7.2% and 30.2% decreases in the near and far future under RCP 4.5, and 32.3% decreases and 5% increases under RCP 8.5 scenarios. These variations were attributed to differences in catchment characteristics and climate variability. Further research is needed to validate these findings by incorporating additional biophysical variables. This study provides insights into hydrological planning and management in the Andit Tid watershed and similar regions facing climate variability. [ABSTRACT FROM AUTHOR]

Estimating river flow is a key parameter for effective water resource management, flood risk prevention, and hydroelectric facilities planning. Yet, traditional gauging methods are not reliable under very high flows or extreme events. Hydrometric network stations are often sparse, and their spatial distribution is not optimal. Therefore, many river sections cannot be monitored using traditional flow measurements and observations. In the last few decades, satellite sensors have been considered as complementary observation sources to traditional water level and flow measurements. This kind of approach has provided a way to maintain and expand the hydrometric observation network. Remote sensing data can be used to estimate flow from rating curves that relate instantaneous flow (Q) to channel cross-section geometry (effective width or depth of the water surface). Yet, remote sensing has limitations, notably its dependence on rating curves. Due to their empirical nature, rating curves are limited to specific river sections (reaches) and cannot be applied to other watercourses. Recently, deep-learning techniques have been successfully applied to hydrology. The primary goal of this study is to develop a deep-learning approach for estimating river flow in the Boreal Shield ecozone of Eastern Canada using RADARSAT-1 and -2 imagery and convolutional neural networks (CNN). Data from 39 hydrographic sites in this region were used in modeling. A new CNN architecture was developed to provide a straightforward estimation of the instantaneous river flow rate. Our results yielded a coefficient of determination (R2) and a Nash–Sutcliffe value of 0.91 and a root mean square error of 33 m3/s. Notably, the model performs exceptionally well for rivers wider than 40 m, reflecting its capability to adapt to varied hydrological contexts. These results underscore the potential of integrating advanced satellite imagery with deep learning to enhance hydrological monitoring across vast and remote areas. [ABSTRACT FROM AUTHOR]

In deltas and estuaries throughout the world, a fluvial‐to‐tidal transition zone (FTTZ) exists where both the river discharge and the tidal motion drive the flow. It is unclear how dune characteristics are impacted by changes in tidal flow strength, and how this is reflected in the hydraulic roughness. To understand dune geometry and variability in the FTTZ and possible impacts on hydraulic roughness, we assess dune variability from multibeam bathymetric surveys, and we use a calibrated 2D hydrodynamic model (Delft3D‐FM) of a sand‐bedded lowland river (Fraser River, Canada). We focus on a period of low river discharge during which tidal impact is strong. We find that the fluvial‐tidal to tidal regime change is not directly reflected in dune height, but local patterns of increasing and decreasing dune height are present. The fluvial‐to‐tidal regime change is reflected in dune shape, where dunes have lower leeside angles and are more symmetrical in the tidal regime. The calibrated model allows to estimate local patterns of dune heights using tidally averaged values of bed shear stress. However, the spatially variable dune morphology hampers local dune height estimation. Changes in dune shape do not significantly impact the reach‐scale roughness, and estimated dune roughness using dune height and length is similar to the dune roughness inferred from model calibration. Hydraulic model performance with a calibrated, constant roughness is not improved by implementing dune‐derived bed roughness. Instead, the data analysis revealed that large‐scale river morphology may explain differences in model roughness and corresponding estimates from dune predictors. Plain Language Summary: Where rivers meet the sea, the flow will often be driven by tides from the sea and by river flow, resulting in a fluvial‐to‐tidal transition zone. The transition can be abrupt or gradual, which might influence the bed of the river, which is covered by dunes. Dune geometry is important in understanding the degree of friction in the river, which in turn determines water levels. It is unclear how bedform characteristics and the related friction are impacted by change in tidal flow strength. This study of the Fraser River in Canada used survey data of the river bed and a computer model of the river flow to study the geometry of dunes and the corresponding friction in this transitional region. We find that dune height and length vary considerably, but they are is unrelated to the fluvial‐tidal regime change. Instead, only the dune leeside, that is, the downstream facing side, was impacted. The difference in leeside angle before and after the regime change, did not result in a different friction produced by the dunes. Using the friction produced by dunes in the model, instead of a constant friction, does not improve model performance. Instead, large‐scale river morphology determines roughness variations. Key Points: Hydraulic roughness in the fluvial‐to‐tidal transition zone estimated from dune geometry agrees with calibrated model roughnessVariation in dune asymmetry and leeside angle across a fluvial‐to‐tidal transition zone has little impact on reach‐scale hydraulic roughnessEstimated spatial bedform patterns from modeled shear stress match measured bedform patterns, but absolute dune heights do not [ABSTRACT FROM AUTHOR]

In addition to having far‐reaching impacts on human health, agriculture, wildfires, ecosystems, and infrastructure, heatwaves control streamflow through the melting of seasonal snow and glacier ice. Despite their importance, there is limited understanding of how heatwaves modify streamflow at regional scales, how these impacts vary by heatwave timing and duration, and how glaciers control the streamflow response. Here, we use a deep learning hydrological model, which has previously been trained, evaluated, and interpreted in southwestern Canada, to simulate the streamflow response to heatwaves at 111 basins in the region. The model, driven by gridded ERA5 reanalysis temperature and precipitation data from 1979 to 2015, is forced by synthetic heatwave conditions that vary in their duration and onset throughout the year. We consider how the streamflow response to heatwaves is sensitive to annual temperatures by adding spatially and temporally uniform warming of 2°C across the study region, under the assumption that the underlying hydrological system behavior remains unchanged. We find that heatwaves, particularly in spring and summer, induce an initial streamflow surplus followed by a streamflow deficit, relative to the non‐heatwave case. In summer, glacier contributions to streamflow partially compensate for streamflow deficits that arise from heatwaves earlier in the melt season. In the scenario with 2°C warmer annual temperatures, heatwaves induce a lesser streamflow response in spring when the seasonal streamflow is most increased due to the advancing freshet. Our findings demonstrate how glaciers buffer the impacts of heatwaves on streamflow, but this buffering effect is expected to diminish as glaciers retreat. Plain Language Summary: Heatwaves are periods of extremely warm temperatures that can have profound consequences on not only human health, agriculture, wildfires, ecosystems, and infrastructure, but also on river flows. We use a model to simulate river flows throughout southwestern Canada under heatwave conditions. We find that 1 week of extreme heat can influence river flows for several months by changing the timing and amount of snow and glacier melt. Excess melt during heatwaves can lead to river flows that initially are higher than they otherwise would be; however, river flows later in the season are then lower than they otherwise would be. Glacierized rivers are unique in that glaciers melt during summer, which can compensate for heatwaves that melt snow earlier in spring. If average temperatures were warmer by 2°C, heatwaves could melt more snow and ice during winter and early spring, but less in late spring and summer. Key Points: We use a deep learning hydrological model to quantify streamflow responses to simulated heatwaves that vary in duration and date of onsetHeatwaves drive initial streamflow surpluses and delayed streamflow deficits that are linked to heatwave timing and glacier coverageWith warmer annual temperatures, spring heatwaves drive lesser streamflow responses when streamflow is most increased by the earlier freshet [ABSTRACT FROM AUTHOR]

The magnitudes of dissolved organic carbon (DOC) exports from boreal peatlands to streams through lateral subsurface flow vary during the ice-free season. Peatland water table depth and the alternation of low and high flow in peat-draining streams are thought to drive this DOC export variability. However, calculation of the specific DOC exports from a peatland can be challenging considering the multiple potential DOC sources within the catchment. A calculation approach based on the hydrological connectivity between the peat and the stream could help to solve this issue, which is the approach used in the present research. This study took place from June 2018 to October 2019 in a boreal catchment in northeastern Canada, with 76.7 % of the catchment being covered by ombrotrophic peatland. The objectives were to (1) establish relationships between DOC exports from a headwater stream and the peatland hydrology; (2) quantify, at the catchment scale, the amount of DOC laterally exported to the draining stream; and (3) define the patterns of DOC mobilization during high-river-flow events. At the peatland headwater stream outlet, the DOC concentrations were monitored at a high frequency (hourly) using a fluorescent dissolved organic matter (fDOM) sensor, a proxy for DOC concentration. Hydrological variables, such as stream outlet discharge and peatland water table depth (WTD), were continuously monitored at hourly intervals for 2 years. Our results highlight the direct and delayed control of subsurface flow from peat to the stream and associated DOC exports. Rain events raised the peatland WTD, which increased hydrological connectivity between the peatland and the stream. This led to increased stream discharge (Q) and a delayed DOC concentration increase, typical of lateral subsurface flow. The magnitude of the WTD increase played a crucial role in influencing the quantity of DOC exported. Based on the observations that the peatland is the most important contributor to DOC exports at the catchment scale and that other DOC sources were negligible during high-flow periods, we propose a new approach to estimate the specific DOC exports attributable to the peatland by distinguishing between the surfaces used for calculation during high-flow and low-flow periods. In 2018–2019, 92.6 % of DOC was exported during flood events despite the fact that these flood events accounted for 59.1 % of the period. In 2019–2020, 93.8 % of DOC was exported during flood events, which represented 44.1 % of the period. Our analysis of individual flood events revealed three types of events and DOC mobilization patterns. The first type is characterized by high rainfall, leading to an important WTD increase that favours the connection between the peatland and the stream and leading to high DOC exports. The second is characterized by a large WTD increase succeeding a previous event that had depleted DOC available to be transferred to the stream, leading to low DOC exports. The third type corresponds to low rainfall events with an insufficient WTD increase to reconnect the peatland and the stream, leading to low DOC exports. Our results suggest that DOC exports are sensitive to hydroclimatic conditions; moreover, flood events, changes in rainfall regime, ice-free season duration, and porewater temperature may affect the exported DOC and, consequently, partially offset the net carbon sequestration potential of peatlands. [ABSTRACT FROM AUTHOR]

Arctic and Subarctic environments are among the most vulnerable regions to climate change. Increases in liquid precipitation and changes in snowmelt onset are cited as the main drivers of change in streamflow and water temperature patterns in some of the largest rivers of the Canadian Arctic. However, in spite of this evidence, there is still a lack of research on water temperature, particularly in the eastern Canadian Arctic. In this paper, we use the CEQUEAU hydrological‐water temperature model to derive consistent long‐term daily flow and stream temperature time series in Aux Mélèzes River, a non‐regulated basin (41 297 km2) in the eastern Canadian subarctic. The model was forced using reanalysis data from the fifth‐generation ECMWF atmospheric reanalyses (ERA5) from 1979 to 2020. We used water temperature derived from thermal infrared (TIR) images as reference data to calibrate CEQUEAU's water temperature model, with calibration performed using single‐site, multi‐site, and upscaling factors approaches. Our results indicate that the CEQUEAU model can simulate streamflow patterns in the river and shows excellent spatiotemporal performance with Kling‐Gupta Efficiency (KGE) metric >0.8. Using the best‐performing flow simulation as one of the inputs allowed us to produce synthetic daily water temperature time series throughout the basin, with the multi‐site calibration approach being the most accurate with root mean square errors (RMSE) <2.0°C. The validation of the water temperature simulations with a three‐year in situ data logger dataset yielded an RMSE = 1.38°C for the summer temperatures, highlighting the robustness of the calibrated parameters and the chosen calibration strategy. This research demonstrates the reliability of TIR imagery and ERA5 as sources of model calibration data in data‐sparse environments and underlines the CEQUEAU model as an assessment tool, opening the door to its use to assess climate change impact on the arctic regions of Canada. [ABSTRACT FROM AUTHOR]

Reliable, long-term records of glacier mass change are invaluable to the glaciological and climate-change communities and used to assess the importance of glacier wastage on streamflow. Here we evaluate the in-situ observations of glacier mass change for Place (1982–2020) and Peyto glaciers (1983–2020) in western Canada. We use geodetic mass balance to calibrate a physically-based mass-balance model coupled with an ice dynamics routine. We find large discrepancies between the glaciological and geodetic records for the periods 1987–1993 (Place) and 2001–2006 (Peyto). Over the period of observations, the exclusion of ice dynamics in the model increased simulated cumulative mass change by ~10.6 (24%) and 7.1 (21%) m w.e. for Place and Peyto glacier, respectively. Cumulative mass loss using geodetic, modelled and glaciological approaches are respectively − 30.5 ± 4.5, − 32.0 ± 3.6, − 29.7 ± 3.6 m w.e. for Peyto Glacier (1982–2017) and − 45.9 ± 5.2, − 43.1 ± 3.1, − 38.4 ± 5.1 m w.e. for Place Glacier (1981–2019). Based on discrepancies noted in the mass-balance records for certain decades (e.g. 1990s), we caution the community if these data are to be used for hydrological model development. [ABSTRACT FROM AUTHOR]

Data‐driven techniques are used extensively for hydrologic time‐series prediction. We created various data‐driven models (DDMs) based on machine learning: long short‐term memory (LSTM), support vector regression (SVR), extreme learning machines, and an artificial neural network with backpropagation, to define the optimal approach to predicting streamflow time series in the Carson River (California, USA) and Montmorency (Canada) catchments. The moderate resolution imaging spectroradiometer (MODIS) snow‐coverage dataset was applied to improve the streamflow estimate. In addition to the DDMs, the conceptual snowmelt runoff model was applied to simulate and forecast daily streamflow. The four main predictor variables, namely snow‐coverage (S‐C), precipitation (P), maximum temperature (Tmax), and minimum temperature (Tmin), and their corresponding values for each river basin, were obtained from National Climatic Data Center and National Snow and Ice Data Center to develop the model. The most relevant predictor variable was chosen using the support vector machine‐recursive feature elimination feature selection approach. The results show that incorporating the MODIS snow‐coverage dataset improves the models' prediction accuracies in the snowmelt‐dominated basin. SVR and LSTM exhibited the best performances (root mean square error = 8.63 and 9.80) using monthly and daily snowmelt time series, respectively. In summary, machine learning is a reliable method to forecast runoff as it can be employed in global climate forecasts that require high‐volume data processing. [ABSTRACT FROM AUTHOR]

Having a complete hydrological time series is crucial for water-resources management and modeling. However, this can pose a challenge in data-scarce environments where data gaps are widespread. In such situations, recurring data gaps can lead to unfavorable outcomes such as loss of critical information, ineffective model calibration, inaccurate timing of peak flows, and biased statistical analysis in various applications. Despite its importance, predicting monthly streamflow can be a complex task due to its connection to random dynamics and uncertain phenomena, posing significant challenges. This study introduces an ensemble machine-learning regression framework for modeling and predicting monthly streamflow time series with a high degree of accuracy. The framework utilizes historical data from multiple monthly streamflow datasets in the same region to predict missing monthly streamflow data. The framework selects the best features from all available gap-free monthly streamflow time-series combinations and identifies the optimal model from a pool of 12 machine-learning models, including random forest regression, gradient boosting regression, and extra trees regressor, among others. The model selection is based on cross-validation train-and-test set scores, as well as the coefficient of determination. We conducted modeling on 26 monthly streamflow time series and found that the gradient boosting regressor with bagging regressor produced the highest accuracy in 7 of the 26 instances. Across all instances, the models using this method exhibited an overall accuracy range of 0.9737 to 0.9968. Additionally, the use of either a bagging regressor or an AdaBoost regressor improved both the tree-based and gradient-based models, resulting in these methods accounting for nearly 80% of the best models. Between January 1960 and December 2021, an average of 40% of the monthly streamflow data was missing for each of the 26 stations. Notably, two crucial stations located in the economically significant lower Athabasca Basin River in Alberta province, Canada, had approximately 70% of their monthly streamflow data missing. To address this issue, we employed our framework to accurately extend the missing data for all 26 stations. These accurate extensions also allow for further analysis, including grouping stations with similar monthly streamflow behavior using Pearson correlation. [ABSTRACT FROM AUTHOR]

Soil moisture is an essential hydrological variable for a suite of hydrological applications. Its spatio-temporal variability can be estimated using satellite remote sensing (e.g., SMOS and SMAP) and in-situ measurements. However, both have their own strengths and limitations. For example, remote sensing has the strength of maintaining the spatial variability of near-surface soil moisture, while in-situ measurements are accurate and preserve the dynamics range of soil moisture at both surface and larger depths. Hence, this study is aimed at (1) merging the strength of SMAP with in-situ measurements and (2) exploring the effectiveness of merged SMAP/in-situ soil moisture in improving ensemble streamflow forecasts. The conditional merging technique was adopted to merge the SMAP-enhanced soil moisture (9 km) and its downscaled version (1 km) separately with the in-situ soil moisture collected over the au Saumon watershed, a 1025 km2 watershed located in Eastern Canada. The random forest machine learning technique was used for downscaling of the near-surface SMAP-enhanced soil moisture to 1 km resolution, whereas the exponential filter was used for vertical extrapolation of the SMAP near-surface soil moisture. A simple data assimilation technique known as direct insertion was used to update the topsoil layer of a physically-based distributed hydrological model with four soil moisture products: (1) the merged SMAP/in-situ soil moisture at 9 and 1 km resolutions; (2) the original SMAP-enhanced (9 km), (3) downscaled SMAP-enhanced (1 km), and (4) interpolated in-situ surface soil moisture. In addition, the vertically extrapolated merged SMAP/in-situ soil moisture and subsurface (rootzone) in-situ soil moisture were used to update the intermediate layer of the model. Results indicate that downscaling of the SMAP-enhanced soil moisture to 1 km resolution improved the spatial variability of soil moisture while maintaining the spatial pattern of its original counterpart. Similarly, merging of the SMAP with in- situ soil moisture preserved the dynamic range of in-situ soil moisture and maintained the spatial heterogeneity of SMAP soil moisture. Updating of the top layer of the model with the 1 km merged SMAP/in-situ soil moisture improved the ensemble streamflow forecast compared to the model updated with either the SMAP-enhanced or in-situ soil moisture alone. On the other hand, updating the top and intermediate layers of the model with surface and vertically extrapolated SMAP/in-situ soil moisture, respectively, did not further improve the accuracy of the ensemble streamflow forecast. Overall, this study demonstrated the potential of merging the SMAP and in-situ soil moisture for streamflow forecast. [ABSTRACT FROM AUTHOR]

Hydrological processes within the alpine tundra of the Taiga Cordillera ecozone in northwestern Canada are poorly understood, yet these areas receive more precipitation per unit area than lowlands and sustain late summer and winter flow in large river systems when contributions from other areas are reduced. The objective of this study was to quantify the spatial and temporal variability in streamflow and groundwater recharge within an alpine tundra basin with discontinuous permafrost and explore the potential impacts of climate change on the timing and intensity of these hydrological processes. Hydrometric and remote sensing methods were used to complete a water balance assessment of the study basin and compare spatial and seasonal differences in inputs, outputs, runoff ratio, and runoff–recharge partitioning during the 2019 open water season. During the freshet, the basin received large daily melt volumes and responded with highly efficient runoff. Evapotranspiration became the primary means of water loss in the early summer but declined as the summer progressed. During the summer, groundwater discharge exceeded precipitation inputs and sustained headwater subbasin streamflow. Groundwater recharge occurred primarily via glaciofluvial upland infiltration during the freshet and channel bed infiltration during the summer. The partitioning of basin outputs between runoff and groundwater recharge was highly seasonal, with a freshet ratio favoring runoff (0.83) while the early and late summer favored recharge (0.28 and 0.17, respectively). As climate change continues, higher air temperatures and greater precipitation are expected for the study basin. Longer open water seasons and declining permafrost extent within the study basin will result in a greater proportion of input water routed to storage and/or groundwater recharge instead of runoff. Shrubification and treeline expansion may also increase evaporative losses from alpine tundra areas, reducing both rapid runoff and delayed aquifer recharge contributions important for larger rivers at lower elevation. [ABSTRACT FROM AUTHOR]

This study investigates the capability of sequence-to-sequence machine learning (ML) architectures in an effort to develop streamflow forecasting tools for Canadian watersheds. Such tools are useful to inform local and region-specific water management and flood forecasting related activities. Two powerful deep-learning variants of the Recurrent Neural Network were investigated, namely the standard and attention-based encoder-decoder long short-term memory (LSTM) models. Both models were forced with past hydro-meteorological states and daily meteorological data with a look-back time window of several days. These models were tested for 10 different watersheds from the Ottawa River watershed, located within the Great Lakes Saint-Lawrence region of Canada, an economic powerhouse of the country. The results of training and testing phases suggest that both models are able to simulate overall hydrograph patterns well when compared to observational records. Between the two models, the attention model significantly outperforms the standard model in all watersheds, suggesting the importance and usefulness of the attention mechanism in ML architectures, not well explored for hydrological applications. The mean performance accuracy of the attention model on unseen data, when assessed in terms of mean Nash–Sutcliffe Efficiency and Kling-Gupta Efficiency is, respectively, found to be 0.985 and 0.954 for these watersheds. Streamflow forecasts with lead times of up to 5 days with the attention model demonstrate overall skillful performance with well above the benchmark accuracy of 70%. The results of the study suggest that the encoder–decoder LSTM, with attention mechanism, is a powerful modelling choice for developing streamflow forecasting systems for Canadian watersheds. [ABSTRACT FROM AUTHOR]

This study investigates the impacts of climate change on the hydrology and soil thermal regime of 10 sub‐arctic watersheds (northern Manitoba, Canada) using the Variable Infiltration Capacity (VIC) model. We utilize statistically downscaled and bias‐corrected forcing datasets based on 17 general circulation model (GCM) ‐ representative concentration pathways (RCPs) scenarios from phase 5 of the Coupled Model Intercomparison Project (CMIP5) to run the VIC model for three 30‐year periods: a historical baseline (1981–2010: 1990s), and future projections (2021–2050: 2030s and 2041–2070: 2050s), under RCPs 4.5 and 8.5. Future warming increases the average soil column temperature by ~2.2°C in the 2050s and further analyses of soil temperature trends at three different depths show the most pronounced warming in the top soil layer (1.6°C 30‐year−1 in the 2050s). Trend estimates of mean annual frozen soil moisture fraction in the soil column show considerable changes from 0.02 30‐year−1 (1990s) to −0.11 30‐year−1 (2050s) across the study area. Soil column water residence time decreases significantly (by 5 years) during the 2050s when compared with the 1990s as soil thawing intensifies the infiltration process thereby contributing to faster conversion to baseflow. Future warming results in 40%–50% more baseflow by the 2050s, where it increases substantially by 19.7% and 46.3% during the 2030s and 2050s, respectively. These results provide crucial information on the potential future impacts of warming soil temperatures on the hydrology of sub‐arctic watersheds in north‐central Canada and similar hydro‐climatic regimes. [ABSTRACT FROM AUTHOR]

A small gradient in the densities (Δρ) of two rivers in Canada was recently shown to develop coherent streamwise orientated vortices (SOVs) in their confluence. Here we use eddy‐resolved numerical modeling to examine how the magnitude and direction of Δρ affect these secondary flow structures. At equal density, lone anticlockwise SOVs are predicted near the surface, a possibility supported herein by recent aerial observations of such SOVs at the confluence. When a Δρ is considered, a front from the denser channel always slides underneath the lighter channel independent of whether the dense front pushes into the fast (Coaticook) or slow (Massawippi) tributary. When the fast Coaticook is denser, coherent clockwise rotating SOVs tend to form. However, when the slow Massawippi is denser, interfacial instabilities are generated as the fast flow of the Coaticook shears overtop the dense front. Thus, the tributary's velocity opposing the dense front's propagation modifies secondary flow characteristics. Importantly, both these cases would have the same densimetric Froude number (FD) if defined using an average velocity of the confluence (e.g., that of the downstream channel), yet this single FD value cannot account for the differences observed between both. This ambiguity inherent to current conventions for calculating FD places into question its use as an adequate predictor of density‐driven secondary flow at confluences. Key Points: The direction of the density gradient decides the sense of rotation of streamwise orientated vorticity in the mixing interfaceThe velocity of the light channel largely determines whether coherent density streamwise orientated vortices or interfacial instabilities formFD based on an averaged velocity of both tributaries isn't fully accurate for predicting density effects [ABSTRACT FROM AUTHOR]

Anthropogenic climate change--induced snowpack loss is affecting streamflow predictability, as it becomes less dependent on the initial snowpack conditions and more dependent on meteorological forecasts. We assess future changes to seasonal streamflow predictability over two large river basins, Liard and Athabasca in western Canada, by approximating streamflow response from the Variable Infiltration Capacity (VIC) hydrologic model with the Bayesian regularized neutral network (BRNN) machine learning emulator. We employ the BRNN emulator in a testbed ensemble streamflow prediction system by treating VIC-simulated snow water equivalent (SWE) as a known predictor and precipitation and temperature from GCMs as ensemble forecasts, thereby isolating the effect of SWE on streamflow predictability. We assess warm-season mean and maximum flow predictability over 2041--70 and 2071--2100 future periods against the1981--2010 historical period. The results indicate contrasting patterns of change, with the predictive skills for mean flow generally declining for the two basins, and marginally increasing or decreasing for the headwater subbasins. The predictive skill for maximum flow declines for the relatively warmer Athabasca basin and improves for the colder Liard basin and headwater subbasins. While the decreasing skill for the Athabasca is attributable to substantial loss in SWE, the improvement for the Liard and headwaters can be attributed to an earlier maximum flow timing that reduces the forecast horizon and offsets the effect of SWE loss. Overall, while the future change in SWE does affect the streamflow prediction skill, the loss of SWE alone is not a sufficient condition for the reduction in streamflow predictability. [ABSTRACT FROM AUTHOR]

First‐order streams flowing through peatlands receive, carry and transform large amounts of organic carbon, methane (CH4) and carbon dioxide (CO2) but remain poorly documented. The objectives of this study were to (a) identify the origins of CO2 and CH4 (thereafter C‐GHG for carbon greenhouse gases) in a peatland headwater stream, (b) determine the environmental factors driving C‐GHG export and emissions, and (c) quantify C‐GHG losses from this system and discuss its implications. Data were collected from eight sampling sites along a 3 km boreal peatland headwater stream (Eastern Canada) over the growing seasons 2019 and 2020. The studied stream was oversaturated in pCO2 [min: 2,044; max: 23,306 μatm] and pCH4 [14; 17,614 μatm]. A mass balance model showed that ∼81% of in‐stream CO2 originated from porewater seepage while the remaining 17% and 2% originated from in‐stream productivity and methane oxidation, respectively. Porewater seepage was concluded to be the primary source of CH4. Seasonal dissolved C‐GHG concentrations were negatively correlated with the peatland water table depth, suggesting an active release of carbon‐rich peat porewater during the base flow. Nevertheless, greater C‐GHG losses occurred during stormflow periods which acted as pulses with most of the C‐GHG being shunted downstream. The sum of C‐GHG export and emissions at our site was 8.08 gC m−2 y−1 with 86% being released to the atmosphere and 14% being exported downstream. Our study demonstrates that peatland headwater streams act as large sources of C‐GHG and that precipitation events and topography control the magnitude of the fluxes. Plain Language Summary: Headwater streams are the inceptive fragment of the river network where a substantial quantity of organic carbon is released from the surrounding substrate. This terrestrial organic carbon can also be transformed into greenhouse gases (GHGs) such as methane and carbon dioxide. In the stream, those GHG can be either consumed, emitted to the atmosphere, or exported downstream. However, the respective contribution of those pathways is unknown. We sampled a stream draining a rain‐fed peatland in Eastern Canada to describe and quantify its carbon GHG dynamics. We found higher concentrations in this stream than in most headwater streams but the total carbon emissions and export were comparable. Peatland streams typically have a low velocity and gentle slope which minimizes atmospheric gas release. We also found that concentrations were greater during the low flow period, potentially because carbon‐rich porewater was released from the peat. However, greater export was reported during the few days of intense rain. The significance of this work is that GHG in headwater streams are spatially and temporally variable and controlled by the surrounding ecosystems, the intensity of rain events and the stream morphology. The inclusion of this variability will improve regional and global carbon budget estimates. Key Points: Porewater seepage accounted for 81% of the carbon dioxide (CO2) present in the stream, in‐stream production and methane (CH4) oxidation contributed to 17% and 2%86.5% of the dissolved CO2 and CH4 was emitted to the atmosphere rather than exported downstream due to gases oversaturation and slow flow46% of the CO2 and CH4 exported downstream occurred during the 10% period when water discharge rates were maximum, suggesting a pulse effect [ABSTRACT FROM AUTHOR]

Macro‐scale distributed hydrologic modeling advanced hydro‐system understandings and scientized relevant human activities. However, it faces challenges of hydrometeorological heterogeneities, parametric interactions, data uncertainty, computational expensiveness, and other complexities, especially over cold regions with intense climatic changes. As an effort to address them, a multifactorial principal‐monotonicity inference (MFPMI) method is developed through integrating, extending, or improving climate classification, hydrologic modeling, and sensitivity analysis. MFPMI is applied to an undammed macro‐scale high‐latitude cold‐region watershed, Athabasca River Basin (ARB) in Canada. MFPMI mitigates the underestimation of climatic impacts on streamflows in process‐based models, hydrologic classification, large‐scale hydroclimatic data deconstruction, parametric‐interaction neglection, and climatic homogenization; its superiority is particularly evident for highly heterogeneous climates. Dominant climatic impacts on ARB streamflows of various regimes increase from tributaries to the mainstem and decrease from up‐ through down‐ to mid‐stream catchments, possibly due to the offset effects of non‐climatic factors (e.g., vegetation and soil). The impacts also decline with streamflow magnitudes, and vary with seasons, spatial scales and lead months rather than temporal resolutions. Streamflow magnitudes, catchments and metrics differ compositions of the climate conditions explaining cross‐scale uppermost discharge variations. In spite of this, streamflow increases with temperature and precipitation, and headwater climates play part of dominant roles in forcing discharges throughout ARB. Accuracy metrics differentiate parameters and accordingly structures of MFPMI models, and hydroclimatic data uncertainty is high for high flows, fine temporal scales or low climatic impacts, increasing uncertainty in hydrologic simulations and climatic‐impact estimates. This study helps advance modeling and understandings of macro‐scale cold‐region hydrology. Key Points: Develop a method for macro‐scale hydrologic modeling under hydrometeorological heterogeneities, parametric interactions, data uncertainty, etcAvoid underestimates of dominant cold‐region climatic impacts in climate homogenization, process‐based modeling & hydrologic classificationReveal impacts' variabilities, metrics & data uncertainties, and climate conditions explaining cross‐scale uppermost discharge variations [ABSTRACT FROM AUTHOR]

[Display omitted] • Uncertainty assessments were estimated in 141 basins for 13 hydrologic model predictions. • Quality of estimates was loosely correlated to model accuracy. • LSTM models were consistently best performing in terms of KGE and CRPS. Probabilistic streamflow predictions are valuable tools for predictive uncertainty estimation, hydrologic risk management, and support for decision-making in water resources. Usually, predictive uncertainty quantification is developed and assessed using only a single hydrological model, making it difficult to generalize to other model configurations. To tackle this issue, we assess changes in the model performance ranking of diverse streamflow models by applying a residual error model post-processing approach to multiple basins and multiple models. This assessment employed 141 basins from the Great Lakes watershed covering the USA and Canada, and 13 diverse streamflow models, which are evaluated using deterministic and probabilistic performance metrics. As the first study to implement probabilistic methods to diverse streamflow models applied to a multitude of basins, the analysis here examines the dependence of probabilistic streamflow estimation quality on model quality. Our findings show that streamflow model choice influences the robustness of probabilistic predictions. It was found that moving from deterministic to probabilistic predictions using a post-processing approach does not change the streamflow model performance ranking for the best and worst deterministic models, but models of intermediate rank in deterministic evaluation do not have consistent ranking when evaluated in probabilistic mode. Post-processing residual errors of long short-term memory (LSTM) network models are consistently the best-performing model in terms of deterministic and probabilistic metrics. This study highlights the significance of combining deterministic streamflow model predictions with residual error models for improving the quality and increasing the value of hydrological predictions, quantifying uncertainty, and facilitating decision-making in operational water management. It also clarifies the degree to which probabilistic predictions depend upon good model performance and can compensate for poor model performance. [ABSTRACT FROM AUTHOR]

[Display omitted] • The proposed framework characterizes the successive wet-dry and dry-wet extremes across climate zones using streamflow data. • We investigate the teleconnections of ENSO and NAO with successive events based on Bayesian quantile regression. • Coastal regions in lower latitudes of Canada show a higher frequency of successive dry-wet and wet-dry extremes. • At higher quantiles, successive events show stronger relationships with low-frequency variability modes. • Results show relatively strong associations between ENSO and NAO with the successive events in Western Canada. Successive hydro-hazards, particularly dry and wet weather whiplash, can severely impact societies and infrastructure undermining the resilience measures developed based on the traditional univariate approaches. Understanding the spatial and temporal characteristics of such lagged compound events is important to improve the efficacy of disaster risk measures (DRMs). This study analyzes the successive dry-wet and wet-dry (SDW/SWD) spells based on the run theory across Canada and proposes a compound magnitude index to identify the corresponding hotspots across 268 watersheds for the historical period of 1963–2012. The frequency, transition time, and magnitude of SDW/SWD events are assessed considering the variable monthly 20th percentile (dry spell) and a fixed 90th percentile (wet spell) of natural streamflow values as thresholds. The potential influence of low-frequency variability modes such as Multivariate El Niño–Southern Oscillation (ENSO) index (MEI) and the North Atlantic Oscillation (NAO) index on the regional hydro-hazards, with a transition time of less than a month, is assessed based on the Bayesian quantile regression approach. Results indicate that the coastal areas, including the west and east coast and the Great Lakes region, are the hotspots for such abrupt transitions. Furthermore, the spatial and temporal characteristics of SDW/SWD events are largely variable and different from those of individual events. The magnitude of these events is influenced by low frequency variability modes particularly at higher quantiles, suggesting that strong phases of MEI and NAO can potentially lead to SDW and SWD events. The findings of this study support the future development of robust DRMs and effective water resource management strategies. [ABSTRACT FROM AUTHOR]

Developing reliable streamflow forecasting models is critical for hydrological tasks such as improving water resource management, analyzing river patterns, and flood forecasting. In this research, for the first time, an emerging multi-level TOPSIS (technique for order preference by similarity to the ideal solution) -based hybridization comprised of the Boruta classification and regression tree (Boruta-CART) feature selection, multivariate variational mode decomposition (MVMD), and a hybrid Convolutional Neural Network (CNN) Bidirectional Gated Recurrent Unit (CNN-BiGRU) deep learning was adopted to multi-temporal (one and three days ahead) forecast the daily streamflow in the Rivers of Prince Edward Island, Canada. For this aim, in the first step, the Boruta-CART feature selection technique determines the most effective lagged components among all the antecedent two-day information (i.e., t-1 and t-2) of hydro-meteorological features (from 2015 to 2020), including the water level, mean air temperature, heat degree days, total precipitation, dew point temperature, and relative humidity in the Bear and Winter Rivers of Prince Edward Island, Canada. Afterwards, a multivariate variational mode decomposition (MVMD) decomposes the input time series to decrease the complexity and non-linearity of the non-stationary ones before feeding the deep learning (DL) models. Here, the CNN-GRU was employed as the primary DL model, along with the kernel extreme machine method (KELM), random variational function link (RVFL), and hybrid CNN bidirectional recurrent neural network (CNN-BiRNN) as the comparative models. A TOPSIS scheme applying several performance measures like the correlation coefficient (R), root mean square error (RMSE), and reliability was designed for the robustness assessment of the hybrid (MVM-CNN-BiGRU, MVM-CNN-BiRNN, MVM-RVFL, and MVM-KELM) and standalone models. The computational outcomes revealed that in the Bear River, the MVM-CNN-BiGRU, owing to its best forecasting performance (one day ahead: TOPSIS score 1, R = 0.960, RMSE = 0.098, and reliability = 65.082; three days ahead: TOPSIS score = 0.999, R = 0.924, and RMSE = 0.33) outperformed the other hybrid models, followed by the MVM-CNN-BiRNN, MVM-RVFL, and MVM-KELM, respectively. Moreover, in the Winter River, the MVM-CNN-BiGRU in terms of (one-day ahead: TOPSIS score = 0.890, R = 0.955, RMSE = 0.274, and reliability = 34.004; three-days ahead: TOPSIS score = 0.686, R = 0.924, and RMSE = 0.330) was superior to the other models. The provided expert system could be vital in the local flood decision-making process, in the absence of streamflow information as input modeling, during the flood seasons to reduce flood damage in residential areas. [Display omitted] • A modern intelligent system to multi-temporal forecast streamflow in Atlantic Canada. • A TOPSIS-based hybridization of Boruta-CART, MVMD, and CNN-BiGRU deep learning scheme. • Accurate capturing the short-term scale streamflow rivers of PEI province of Canada. • Comparative assessment of CNN-BiGRU, CNN-BiRNN, RVFL, and KELM against CNN-BiGRU. [ABSTRACT FROM AUTHOR]

Deep learning has emerged as a useful tool across geoscience disciplines; however, there remain outstanding questions regarding the suitability of unexplored model architectures and how to interpret model learning for regional-scale hydrological modelling. Here we use a convolutional long short-term memory network, a deep learning approach for learning both spatial and temporal patterns, to predict streamflow at 226 stream gauges across southwestern Canada. The model is forced by gridded climate reanalysis data and trained to predict observed daily streamflow between 1980 and 2015. To interpret the model's learning of both spatial and temporal patterns, we introduce a set of experiments with evaluation metrics to track the model's response to perturbations in the input data. The model performs well in simulating daily streamflow over the testing period, with a median Nash–Sutcliffe efficiency (NSE) of 0.68 and 35 % of stations having NSE>0.8. When predicting streamflow, the model is most sensitive to perturbations in the input data prescribed near and within the basins being predicted, demonstrating that the model is automatically learning to focus on physically realistic areas. When uniformly perturbing input temperature time series to obtain relatively warmer and colder input data, the modelled indicator of freshet timing and peak flow changes in accordance with the transition timing from below- to above-freezing temperatures. We also demonstrate that modelled August streamflow in partially glacierized basins is sensitive to perturbations in August temperature, and that this sensitivity increases with glacier cover. The results demonstrate the suitability of a convolutional long short-term memory network architecture for spatiotemporal hydrological modelling, making progress towards interpretable deep learning hydrological models. [ABSTRACT FROM AUTHOR]

Setting e-flows (instream, environmental or ecological flow regimes) for existing or new hydroelectric and other projects is a key worldwide consideration. A Canadian perspective and experience in arriving at e-flow regimes for ice-free or ice-covered rivers, with emphasis on small and large hydroelectric projects, is presented through general concepts and modeling approaches. Rather than a single standard approach, several methodologies are used throughout Canada, although in some regions preferred techniques may be favoured. Methods range from relatively simple desktop calculations based primarily on hydrological data to sophisticated hydrodynamic and habitat modeling followed by time series analyses. Method complexity usually depends on impact severity, particularly on fish and fish habitat. Multi-jurisdictional, legislative, climatic (e.g. ice regimes) and geographic landscapes present unique challenges for methods, data availability, and tools used to set e-flow regimes. Nature-mimicking approaches are also employed to enhance the ecological integrity of estimated flow regimes by imitating key elements of natural hydrographs and geomorphic features. More holistic approaches and scientific reviews may supplement studies, particularly when complex technical and societal issues are involved. Plans for new or modified hydroelectric projects strive to balance power generation with social and economic considerations, as well as river flow regimes for ecological needs, including fish and fish habitat. [ABSTRACT FROM AUTHOR]

This study examines centennial‐scale hydrological and sedimentological effects of floodplain inundation by avulsion and its upstream and downstream controls. The 1870s avulsion in Cumberland Marshes diverted the Saskatchewan River flow towards Cumberland Lake, a local base level. It invaded a poorly drained sub‐basin of Cumberland Marshes floodplain linked to the parent Saskatchewan River by two small outlets in the resistant substrate. The rapid increase in inflow (~5× on average) during the earlier stages of the avulsion resulted in the base‐level rise and floodplain inundation by the avulsion lake. Since the early 20th century, the forced regression of the avulsion lake occurred, caused by ~5× outlet channel enlargement by 'hungry‐water' outflows, whereas the mean lake inflows experienced little change. The avulsion lake served as an effective sediment trap and was filled by predominantly progradational sandy and silty avulsion deposits up to 3–4 m thick, covering about 700 km2. Elsewhere, fluviodeltaic settings with 'negative relief' and limited hydrologic connectivity with the rest of the floodplain may be prone to avulsion lakes that form if the rates of inflow increased by avulsion exceed the rates of outflow. Avulsion lakes can last for ~100 years or more before they drain and/or become filled by overbank sediments. On well‐drained floodplains, inundations by avulsions are expected to be short‐term and result in little progradational deposition. This study demonstrates that in some local hydrographic basins, base level becomes a variable of an evolving avulsion rather than its fixed external control. Although avulsion‐induced base‐level changes are short‐lived, they affect 102–103 km2 of a floodplain and occur rapidly, accompanied by high aggradation rates. [ABSTRACT FROM AUTHOR]

Sub-daily and weekly flow cycles termed 'hydropeaking' are common features in regulated rivers worldwide. Weekly flow periodicity arises from fluctuating electricity demand and production tied to socioeconomic activity, typically with higher consumption during weekdays followed by reductions on weekends. Here, we propose a weekly hydropeaking index to quantify the 1920–2019 intensity and prevalence of weekly hydropeaking cycles at 500 sites across the United States of America and Canada. A robust weekly hydropeaking signal exists at 1.8% of sites starting in 1920, peaking at 18.9% in 1963, and diminishing to 3.1% in 2019, marking a 21st century decline in weekly hydropeaking intensity. We propose this decline may be tied to recent, above-average precipitation, socioeconomic shifts, alternative energy production, and legislative and policy changes impacting water management in regulated systems. Vanishing weekly hydropeaking cycles may offset some of the prior deleterious ecohydrological impacts from hydropeaking in highly regulated rivers. Rapid sub-daily and weekly fluctuations in streamflow are known as hydropeaking. Here, the authors apply a new "weekly hydropeaking index" (WHI) to streamflow data from 500 gauges across the USA and Canada and find that the strength of WHI peaked in the mid-20th century and has been declining through the 2010s. [ABSTRACT FROM AUTHOR]

Groundwater recharge (GWR) is a strategic hydrologic variable, and its estimate is necessary to implement sustainable groundwater management. This is especially true in a global warming context that highly impacts key winter conditions in cold and humid climates. For this reason, long-term simulations are particularly useful for understanding past changes in GWR associated with changing climatic conditions. However, GWR simulation at the regional scale and for long-term conditions is challenging, especially due to the limited availability of spatially distributed calibration data and due to generally short observed time series. The objective of this study is to demonstrate the relevance of using a water budget model to understand long-term transient and regional-scale GWR in cold and humid climates where groundwater observations are scarce. The HydroBudget model was specifically developed for regional-scale simulations in cold and humid climate conditions. The model uses commonly available data such as runoff curve numbers to describe the study area, precipitation and temperature time series to run the model, and river flow rates and baseflow estimates for its automatic calibration. A typical case study is presented for the southern portion of the Province of Quebec (Canada, 36 000 km 2). With the model simultaneously calibrated on 51 gauging stations, the first GWR estimate for the region was simulated between 1961 and 2017 with very little uncertainty (≤ 10 mm/yr). The simulated water budget was divided into 41 % runoff (444 mm/yr), 47 % evapotranspiration (501 mm/yr), and 12 % GWR (139 mm/yr), with preferential GWR periods during spring and winter (44 % and 32 % of the annual GWR, respectively), values that are typical of other cold and humid climates. Snowpack evolution and soil frost were shown to be a key feature for GWR simulation in these environments. One of the contributions of the study was to show that the model sensitivity to its parameters was correlated with the average air temperature, with colder watersheds more sensitive to snow-related parameters than warmer watersheds. Interestingly, the results showed that the significant increase in precipitation and temperature since the early 1960s did not lead to significant changes in the annual GWR but resulted in increased runoff and evapotranspiration. In contrast to previous studies of past GWR trends in cold and humid climates, this work has shown that changes in past climatic conditions have not yet produced significant changes in annual GWR. Because of their relative ease of use, water budget models are a useful approach for scientists, modelers, and stakeholders alike to understand regional-scale groundwater renewal rates in cold and humid climates, especially if they can be easily adapted to specific study needs and environments. [ABSTRACT FROM AUTHOR]

Floods are the most frequently occurring natural hazard in Canada. An in‐depth understanding of flood seasonality and its drivers at a national scale is essential. Here, a circular, statistics‐based approach is implemented to understand the seasonality of annual‐maximum floods (streamflow) and to identify their responsible drivers across Canada. Nearly 80% and 70% of flood events were found to occur during spring and summer in eastern and western watersheds across Canada, respectively. Flooding in the eastern and western watersheds was primarily driven by snowmelt and extreme precipitation, respectively. This observation suggests that increases in temperature have led to early spring snowmelt‐induced floods throughout eastern Canada. Our results indicate that precipitation (snowmelt) variability can exert large controls on the magnitude of flood peaks in western (eastern) watersheds in Canada. Further, the nonstationarity of flood peaks is modelled to account for impact of the dynamic behaviour of the identified flood drivers on extreme‐flood magnitude by using a cluster of 74 generalized additive models for location scale and shape models, which can capture both the linear and nonlinear characteristics of flood‐peak changes and can model its dependence on external covariates. Using nonstationary frequency analysis, we find that increasing precipitation and snowmelt magnitudes directly resulted in a significant increase in 50‐year streamflow. Our results highlight an east–west asymmetry in flood seasonality, indicating the existence of a climate signal in flood observations. The understating of flood seasonality and flood responses under the dynamic characteristics of precipitation and snowmelt extremes may facilitate the predictability of such events, which can aid in predicting and managing their impacts. [ABSTRACT FROM AUTHOR]

Hydrologic functions of catchments are intrinsically diverse and defined as the ways catchments partition, store, and drain rainfall and snowmelt. Large‐sample hydrology (LSH) uses existing datasets of catchments to derive generalizable conclusions on hydrologic behaviors. LSH has the potential to synthesize the diversity of catchment hydrologic functions, allowing a robust extrapolation of streamflow generation mechanisms to poorly gauged regions. However, the descriptors of hydrologic functions, required to synthesize and extrapolate, have not been developed in LSH methodologies. This has potentially resulted in unexpectedly small associations between catchments' physical features and streamflow characteristics as well as poor predictability of the shape of streamflow hydrographs, as shown in recent LSH studies. Here, we propose three dimensionless indices—which directly quantify hydrologic functions—based on how interactions between a catchment's climatic and physical attributes construct catchment functions. Using climatic and physical data as well as long‐term streamflow observations at hundreds of gauged catchments across the United States and Canada, our results depict that the use of interactive functional indices as catchment descriptors improves the performance of LSH methodologies (hierarchical clustering, multivariate regression, and random forests) in identifying hydrologic similarities in shape‐based streamflow signatures among catchments, and in predicting shape‐based streamflow signatures in poorly gauged regions. This research highlights the importance of physical interpretability of LSH models and showcases the development of parsimonious and process‐based catchment‐scale frameworks, allowing the analysis of globally available catchment data to progress the generalizable understanding of catchment hydrologic functions and streamflow generation mechanisms. Key Points: Inclusion of interactive functional indices, developed independent of streamflow data, improves performance of hierarchical clusteringFunctional indices improve regression and random forests accuracy in extrapolating the predictions of shape‐based streamflow signaturesClimatic information could sufficiently explain (dis)similarity in magnitude‐based, but not in shape‐based, streamflow signatures [ABSTRACT FROM AUTHOR]

Climate change affects natural streamflow regimes globally. To assess alterations in streamflow regimes, typically temporal variations in one or a few streamflow characteristics are taken into account. This approach, however, cannot see simultaneous changes in multiple streamflow characteristics, does not utilize all the available information contained in a streamflow hydrograph, and cannot describe how and to what extent streamflow regimes evolve from one to another. To address these gaps, we conceptualize streamflow regimes as intersecting spectrums that are formed by multiple streamflow characteristics. Accordingly, the changes in a streamflow regime should be diagnosed through gradual, yet continuous changes in an ensemble of streamflow characteristics. To incorporate these key considerations, we propose a generic algorithm to first classify streams into a finite set of intersecting fuzzy clusters. Accordingly, by analyzing how the degrees of membership to each cluster change in a given stream, we quantify shifts from one regime to another. We apply this approach to the data, obtained from 105 natural Canadian streams, during the period of 1966 to 2010. We show that natural streamflow in Canada can be categorized into six regime types, with clear hydrological and geographical distinctions. Analyses of trends in membership values show that alterations in natural streamflow regimes vary among different regions. Having said that, we show that in more than 80 % of considered streams, there is a dominant regime shift that can be attributed to simultaneous changes in streamflow characteristics, some of which have remained previously unknown. Our study not only introduces a new globally relevant algorithm for identifying changing streamflow regimes but also provides a fresh look at streamflow alterations in Canada, highlighting complex and multifaceted impacts of climate change on streamflow regimes in cold regions. [ABSTRACT FROM AUTHOR]

This study generated a water yield dataset for Canada for 1979–2016 by subtracting the land surface evapotranspiration (ET) and water surface evaporation (E0) from precipitation (P). The dataset was validated in Budyko space and compared with streamflow (Q) before the spatial variability and trends were analysed. Results indicate (1) uncertainties of the dataset are generally small; (2) despite the asynchronous inter-annual change, annual water yield varies in a similar temporal pattern to Q; (3) annual water yield varies dramatically across Canada, ranging from about zero on the Canadian Prairies to over 2500 mm on the West Coast; and (4) annual water yield shows no significant changes over the study period in the vast majority (82.4%) of Canada's landmass. The most significant increasing trend appears in South Central Canada, attributed to increasing P. The most significant decreasing trend appears in Northeast Canada and the Southern Montane Cordillera, attributed to decreasing P and increasing ET. [ABSTRACT FROM AUTHOR]

Warm and dry summer days can lead to low streamflow due to a lack of rainfall and increased evaporation. In glacierized catchments, however, such periods can lead to a very different hydrological response as glaciers can supply an increased amount of meltwater, thereby compensating for the rainfall deficits. Here, we analyzed glacier-fed streamflow responses to warm and dry (WD) periods in long-term streamflow observations (>50 years). WD events during summer (June–September) were analyzed for catchments with varying glacier cover in western Canada, southwestern Norway, and the European Alps. WD events were defined by days with temperatures above a daily varying threshold, based on the 80th percentile of the respective long-term temperature data for that day in the year, and daily precipitation sums below a fixed threshold (<2 mm d -1) for a minimum duration of 7 d. Streamflow responses to these WD events were expressed as level of compensation (C) and were calculated as the event streamflow relative to the long-term streamflow regime. C≥100 % indicates that increased melt and other catchment storages could compensate, or even overcompensate, the rainfall deficit and increased evaporation. Results showed a wide range of compensation levels, both between catchments and between different WD events in a particular catchment. C was, in general, higher than 100 % for catchments with a relative glacier cover higher than 5 %–15 %, depending on region and month. June was the month with highest compensation levels, but this was likely more influenced by snowmelt than by glacier melt. For WD events in September, C was still higher than 100 % in many catchments, which likely indicates the importance of glacier melt as a streamflow contributor in late summer. There was a considerable range in C of different WD events for groups of catchments with similar glacier cover. This could be partly explained by antecedent conditions, such as the amount of snow fallen in the previous winter and the streamflow conditions 30 d before the WD event. Some decreasing trends in C were evident, especially for catchments in western Canada and the European Alps. Overall, our results suggest that glaciers do not compensate straightforwardly, and the range in compensation levels is large. The different streamflow components – glacier, snow and rain – and their variations are important for the buffering capacity and the compensating effect of glaciers in these high mountain water systems. [ABSTRACT FROM AUTHOR]

Sparse precipitation information can result in uncertainties in hydrological modeling practices. Precipitation observation network augmentation is one way to reduce the uncertainty. Meanwhile, in basins with snowpack-dominated hydrology, in the absence of a high-density precipitation observation network, assimilation of in situ and remotely sensed measurements of snowpack state variables can also provide the possibility to reduce flow estimation uncertainty. Similarly, assimilation of existing precipitation observations into gridded numerical precipitation products can alleviate the adverse effects of missing information in poorly instrumented basins. In Canada, the Regional Deterministic Precipitation Analysis (RDPA) data from the Canadian Precipitation Analysis (CaPA) system have been increasingly applied for flow estimation in sparsely gauged Nordic basins. Moreover, CaPA-RDPA data have also been applied to establish observational priorities for augmenting precipitation observation networks. However, the accuracy of the assimilated data should be validated before being applicable in observation network assessment. The assimilation of snowpack state variables has proven to significantly improve streamflow estimates, and therefore, it can provide the benchmark against which the impact of assimilated precipitation data on streamflow simulation can be compared. Therefore, this study introduces a parsimonious framework for performing a proxy validation of the precipitation-assimilated products through the application of snow assimilation in physically based hydrologic models. This framework is demonstrated to assess the observation networks in three boreal basins in Yukon, Canada. The results indicate that in most basins, the gridded analysis products generally enjoyed the level of accuracy required for accurate flow simulation and therefore were applied in the meteorological network assessment in those cases. [ABSTRACT FROM AUTHOR]

• New high air temperature records were established over a large domain over a few days. • Changes to streamflow coincided with the air temperature extremes. • The hydrologic regime determined the type of response to the heatwave. • Hydrological responses were conditioned by the existing snowpack. The 2021 Western North America heatwave resulted in record high air temperatures over a large area of Canada west of 100°. The heatwave persisted from mid-June into July, depending upon the threshold used to define the heatwave. The heatwave was a weather event that was short lived but had a widespread transient impact on hydrology in this single year. These impacts were more evident in nival systems because of the prominent role of snowmelt. In normal years, the timing of hydrological processes, particularly snowmelt, is a function of latitude and elevation; in 2021 the heatwave resulted in simultaneous high rates of snowmelt across rivers where a snowpack existed at the time of the event, and strong diurnal melt signal was a diagnostic. In 2021, rivers throughout Canada west of 100°W responded strongly to the heatwave, but the response depended on three factors: location relative to the heat dome, the hydrologic regime, and the amount of snow present at the time. The melt signal was strongest in those basins that had not reached their annual peak often having discharges in the highest 5 % of historical observations for that day of the year. Discharges rapidly declined in basins already in recession indicating the rapid depletion of the snowpack. In basins with little or no remaining snow, streamflow often declined during the same period often to levels in the lowest 5 % of historical observations for that day of the year. [ABSTRACT FROM AUTHOR]

• Generative Deep learning models were developed for streamflow forecasting. • A specific generative model, conditional variational auto-encoder was selected. • Conditional variational auto-encoder models were tested on 75 basins in Canada. • Generative models outperformed long short-term memory networks. • Examining more than 9000 models proved models produce reliable forecasts. Probabilistic hydrological forecasting has gained increasing importance in recent years, as it offers essential information for risk-based decision-making and flood management. Traditional hydrological models often produce deterministic forecasts, which do not account for the inherent uncertainties in hydrological systems. Although many previous studies have investigated using deep learning (DL) models for hydrological prediction, the development of probabilistic DL models (especially, generative models) has not yet been thoroughly examined for hydrological forecasting. The present study investigates the efficacy of a generative DL model, namely, conditional variational auto-encoder (CVAE). CVAE is applied for one-seven day(s) ahead probabilistic streamflow forecasting in 75 basins from the Canadian model parameter experiment (CANOPEX) database. The CVAE forecast model, which outputs forecasts in the form of a probability distribution, was benchmarked against two state-of-the-art quantile-based DL models: quantile-based encoder-decoder (ED) and quantile-based CVAE (QCVAE). The latter outputs forecasts for specific quantiles of a probability distribution (here, q = 0.05, 0.5, 0.95). More than 9000 models were developed based on different basins, input variable sets, and model structures. The models were evaluated in terms of point forecast accuracy and forecast reliability. The results indicate that the CVAE model generally outperforms the benchmark models in terms of reliability at a 90 % confidence level (median reliability of 92.49 % compared to 87.35 % and 84.59 % for ED and QCVAE, respectively). However, the quantile-based forecast models produce slightly more accurate point forecasts than the CVAE (median Kling-Gupta efficiency (KGE) of 0.88 compared to 0.90 for both ED and QCVAE). Notably, the CVAE model exhibits superior probabilistic forecasts in basins with poor point forecast accuracy, highlighting its usefulness over benchmark methods in difficult-to-forecast basins. Overall, CVAE is a promising probabilistic DL model for streamflow forecasting, and it can be readily applied for forecasting other hydrological variables (evapotranspiration, water level, etc.). The findings of this study provide a basis for future research on probabilistic forecasting of hydrological variables using generative DL models. [ABSTRACT FROM AUTHOR]

Copyright of Canadian Water Resources Journal / Revue Canadienne des Ressources Hydriques is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The Little Bow River (LBR) in western Canada naturally displayed an intermittent flow regime; the channel dried up most summers, excluding the development of riparian woodlands within this semi‐arid ecoregion. Around 1900, the Little Bow Canal was excavated to divert water from the adjacent Highwood River and with flow augmentation the LBR flow became continuous through the growth season. We hypothesized that the continuous regime enabled riparian woodland establishment and assessed conditions with sequential aerial photographs and field observations. Supporting the hypothesis, a few woodland groves established near the Highwood River where balsam poplars (Populus balsamifera) provided an abundant seed source. To investigate the basis for woodland development, we analysed historical hydrology of the LBR and assessed the four larger woodland groves, which included mature poplars, trembling aspen (P. tremuloides) and willow shrubs (Salix bebbiana and S. exigua). Each location had some bank excavation with channelization or gravel mining, and tree ageing through ring counts indicated gradual colonization and pulses of establishment after floods in 1920 and 1942. Thus, the conversion from an intermittent to continuous flow regime enabled woodland development which also benefited from excavations that created barren colonization sites. The study revealed four requirements for riparian woodland colonization in a dry region: (a) seeds, (b) barren sites, (c) bank saturation with higher river flow, and (d) sufficient river flows for tree and shrub survival and growth. While water withdrawal commonly degrades riverine ecosystems, flow augmentation can provide the opposite outcome, enhancing the river and riparian environments. [ABSTRACT FROM AUTHOR]

Traditionally, hydrological models are only calibrated to reproduce streamflow regime without considering other hydrological state variables, such as soil moisture and evapotranspiration. Limited studies have been performed on constraining the model parameters, despite the fact that the presence of a large number of parameters may provide large degree of freedom, resulting in equifinality and poor model performance. In this study, a multi-objective optimization approach is adopted, and both streamflow and soil moisture data are calibrated simultaneously for an experimental study basin in the Saskatchewan Prairies in western Canada. The results of this study show that the multi-objective calibration improves model fidelity compared to the single objective calibration. Moreover, the study demonstrates that single objective calibration performed against only streamflow can fairly mimic the streamflow hydrograph but does not yield realistic estimation of other fluxes such as evapotranspiration and soil moisture (especially in deeper soil layers). [ABSTRACT FROM AUTHOR]

During spring 2011, an extreme flood occurred along the Richelieu River located in southern Quebec, Canada. The Richelieu River is the last section of the complex Richelieu basin, which is composed of the large Lake Champlain located in a valley between two large mountains. Previous attempts in reproducing the Richelieu River flow relied on the use of simplified lumped models and showed mixed results. In order to prepare a tool to assess accurately the change of flood recurrences in the future, a state‐of‐the‐art distributed hydrological model was applied over the Richelieu basin. The model setup comprises several novel methods and data sets such as a very high resolution river network, a modern calibration technique considering the net basin supply of Lake Champlain, a new optimization algorithm, and the use of an up‐to‐date meteorological data set to force the model. The results show that the hydrological model is able to satisfactorily reproduce the multiyear mean annual hydrograph and the 2011 flow time series when compared with the observed river flow and an estimation of the Lake Champlain net basin supply. Many factors, such as the quality of the meteorological forcing data, that are affected by the low density of the station network, the steep terrain, and the lake storage effect challenged the simulation of the river flow. Overall, the satisfactory validation of the hydrological model allows to move to the next step, which consists in assessing the impacts of climate change on the recurrence of Richelieu River floods. Plain Language Summary: In order to study the 2011 Richelieu flood and prepare a tool capable of estimating the effects of climate change on the recurrence of floods, a hydrological model is applied over the Richelieu basin. The application of a distributed hydrological model is useful to simulate the flow of all the tributaries of the Richelieu basin. This new model setup stands out from past models due to its distribution in several hydrological units, its high‐resolution river network, the calibration technique, and the high‐resolution weather forcing data set used to drive the model. The model successfully reproduced the 2011 Richelieu River flood and the annual hydrograph. The simulation of the Richelieu flow was challenging due to the contrasted elevation of the Richelieu basin and the presence of the large Lake Champlain that acts as a reservoir and attenuates short‐term fluctuations. Overall, the application was deemed satisfactory, and the tool is ready to assess the impacts of climate change on the recurrence of Richelieu River floods. Key Points: An advanced high‐resolution distributed hydrological model is applied over a U.S.‐Canada transboundary basinThe simulated net basin supply of Lake Champlain and the Richelieu River discharge are in good agreement with observations of the 2011 floodThe flow simulation is challenging due to the topographic and meteorological complexities of the basin and uncertainties in the observations [ABSTRACT FROM AUTHOR]

The magnitude and timing of water delivery in two large northern basins are analysed to clarify where runoff is generated and how their rivers acquire comparable regimes (or seasonal rhythms) of flow. These two rivers, the Mackenzie in Canada and the Yenisei in Russia, traverse similar latitudes, physiographic provinces, vegetation zones and climatic regions. Within the basins, mountainous terrain and high-precipitation sections usually yield large runoff, but low runoff comes from the plains, low plateaus and areas of aridity. Winter runoff is commonly low and snowmelt is responsible for annual peak runoff in most parts of these basins, though rainfall is a prominent runoff source in southern Yenisei. Many rivers in the drainage networks display a seasonal pattern that suggests the dominance of snowmelt to produce a spring freshet followed by a general decline in summer that diminishes to winter low flows. Regulation of reservoir outflow greatly distorts the natural flow regime. Yet, along the main river downstream of the reservoirs, the influx of tributary discharge can dilute such human influence. To truly understand how water is produced and transferred in large northern rivers, the spatial and temporal complexity of flow-generation mechanisms and storage effects need to be unravelled. [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]

• Artificial intelligence and non-linear time series models were integrated. • Hybrid models were developed to improve the monthly river flow prediction. • Time series models showed superiority over the artificial intelligence methods. • Hybrid models outperformed the stand-alone time series and artificial intelligence models. Prediction of river flow as a fundamental source of hydrological information plays a crucial role in various fields of water projects. In this study, at first, the capabilities of two time series analysis approaches, namely self-exciting threshold autoregressive (SETAR) and generalized autoregressive conditional heteroscedasticity (GARCH) models, then three artificial intelligence approaches including artificial neural networks (ANN), multivariate adaptive regression splines (MARS), and random forests (RF) models were investigated to predict monthly river flow. For this purpose, monthly river flow data of Brantford and Galt stations on Grand River, Canada, for the period from October 1948 to September 2017 were used and their performances were evaluated based on various evaluation criteria. The SETAR model showed better performance than the GARCH one in prediction of river flows at the stations of study. Additionally, the stand-alone MARS and RF models performed slightly better than the ANN. Next, hybrid models were developed by coupling the used ANN, MARS, and RF models with SETAR and GARCH models as the non-linear time series models. The performance of various models presented in this study indicated that the new hybrid models demonstrated a much better performance compared with the stand-alone ones at both stations. Among the developed hybrid models, the RF-SETAR models generally had the best accuracy to improve the river flows modeling. As a result, it can be concluded that the presented methodology can be used to predict hydrological time series such as river flow with a high level of accuracy. [ABSTRACT FROM AUTHOR]

• Multiscale and multivariate analyses of ENSO, NAO and PDO on precipitation and flow. • Wavelet-based models were used to perform multiscale and multivariate analyses. • Positive and negative phases caused drier and wetter conditions, respectively. • Annual periodicity characterized precipitation and streamflow variability. • Three-factor combinations were needed to explain the precipitation and flow variability. A number of past studies have investigated the influence of various teleconnections – such as El Niño-Southern Oscillation (ENSO), the North Atlantic Oscillation (NAO) and the Pacific Decadal Oscillation (PDO) – on precipitation and streamflow. These studies, however, have not focused on analyzing the combined influence of the different phases of these teleconnections and the simultaneous influence of these teleconnections at differing time-frequency scales. The present study addresses this issue by exploring the use of wavelet-based methods in combination with non-parametric approaches to analyze individual and combined influences involving ENSO, NAO, and PDO on monthly precipitation and streamflow data from watersheds in Alberta, Ontario, and Newfoundland, in Canada. This study is the first time that multiscale and multivariate analyses of ENSO, NAO, and PDO, along with their different phases, are used to explain the variability of streamflow and precipitation in a watershed. Generally, the positive and negative phases (particularly of ENSO and NAO) were respectively associated with lower and higher precipitation/streamflow, while the neutral phase showed similar behavior to that of the negative phase. The results of the bivariate and multivariate wavelet coherences revealed that there were consistent increases in the average wavelet coherence (AWC) and the percentage of significant coherence (PoSP) for all watersheds, from using only one factor to two and three teleconnection factors. The ranges of AWC for one, two and three factors were 0.31–0.40, 0.56–0.66, and 0.77–0.81, respectively. The ranges of PoSP for one, two and three factors were 3.54–14.3, 28.38–47.29, and 69.28–76.71, respectively. This implies that three-factor combinations (ENSO-NAO-PDO) were needed to explain the variability of precipitation and streamflow in all watersheds. [ABSTRACT FROM AUTHOR]

Dams represent one of the major forms of river alteration. As these structures reach the end of their lifespan, they often require extensive refurbishments or removal. A small‐scale water supply dam in Banff National Park (Alberta, Canada) was partially removed, creating a breach that allowed water to scour a new passage resembling a nature‐like fishway. We investigated the permeability of the partially removed dam as a means of validating the conservation benefits of the partial dam removal. We quantified the proportion of bull trout (Salvelinus confluentus), a threatened species in Canada, that approached and passed the fishway using radio telemetry receiver stations. The proportion of bull trout that approached the fishway was low (37.0%; N = 27 of 73), but was consistent with upstream reference sites (33%; N = 20 of 60). For those that did approach, the proportion of bull trout that passed yielded a high passage efficiency (77.8%; N = 21 of 27 that approached). The probability that a fish passed the fishway was related to water depth and time of day. Bull trout were more likely to pass when water depths were high (>0.40 m), and at night. Passage duration ranged from 5‐mins to 13‐days, suggesting that this resident species used the fishway for a variety of purposes (e.g., station holding and foraging) and not just transiting. Some individuals underwent large‐scale movements 2‐km upstream (15.1%; N = 11 of 73), or 2‐km downstream (2.7%; N = 2 of 73) following a successful passage event. This study provides new insight on how, in some instances, a breach in a dam can function as a nature‐like fishway, accommodating year‐round stream flows and providing hydraulic conditions suitable for fish passage without costly engineering or construction. [ABSTRACT FROM AUTHOR]

River flow data are usually subject to several sources of discontinuity and inhomogeneity. An example is seasonality, because climatic oscillations occurring on inter-annual timescale have an obvious impact on the river flow. Further sources of alteration can be caused by changes in reservoir management, instrumentation or even unexpected shifts in climatic conditions. When such changes are ignored the results of a statistical analysis can be strongly misleading, so flexible procedures are needed for building the appropriate models, which may be very complex. This paper develops an automatic procedure to estimate the number and locations of changepoints in Periodic AutoRegressive (PAR) models, which have been extensively used to account for seasonality in hydrology. We aim at filling the literature gap on multiple changepoint detection by allowing several time segments to be detected, inside of which a different PAR structure is specified, with the resulting model being employed to successfully capture the discontinuities of river flow data. The model estimation is performed by optimization of an objective function based on an information criterion using genetic algorithms. The proposed methodology is evaluated by means of simulation studies and it is then employed in the analysis of two river flows: the South Saskatchewan, measured at Saskatoon, Canada, and the Colorado, measured at Lees Ferry, Arizona. For these river flows we build changepoint models, discussing the possible events that caused discontinuity, and evaluate their forecasting accuracy. Comparisons with the literature on river flow analysis and on existing methods for changepoint detection confirm the efficiency of our proposal. [ABSTRACT FROM AUTHOR]

Long‐term weather (1990–2013), streamflow (1990–2013; excluding 7 years with no or poor data), and water chemistry (2009–2013) datasets for hydrological modelling applications were developed using simple methods for the upper La Salle River watershed, in Canada, to address the lack of such datasets in the northern Red River Basin. Weather variables consist of temperature, relative humidity, wind speed, solar radiation, and precipitation disaggregated to an hourly time‐step. The only hydrometric variable included in the dataset is stream discharge in a daily time‐step. Water chemistry data consisted of total nitrogen (TN), total dissolved nitrogen (TDN), total phosphorus (TP), and total dissolved phosphorus (TDP). Samples were collected weekly during the open water season at the same site as the hydrometric gauging station from August 2009 to October 2012 with some gaps (i.e. Fall 2011, Spring 2012, September 2012). In 2013 the sampling frequency was increased to daily or sub‐daily during high stream discharge and weekly during low stream discharge. A data overview indicates values within ranges reported for the area. Mean annual, winter, and summer temperatures were 3.5, −10.7, and 17.2°C, respectively. Annual relative humidity averaged 73.1% but was higher and more homogenous in cold seasons. Wind speed was similar over the year with annual average of 4.3 m/s. Solar radiation followed the typical curve reported for western Canada, with peak daily average values around 250 W/m2 in July. Precipitation records were mostly comprised of dry hours with 75.3% of the events being equal or less than 2 mm/h. Stream discharge was typical of the Canadian Prairies; the average peak discharge over the entire period was larger in April (2.3 m3/s) due to large amounts of snowmelt runoff. Average concentrations of TN, TDN, TP, and TDP of 1.54, 1.35, 0.56, and 0.49 mg/L, respectively, were in agreement with values found in previous studies at the site. Open Practices: This article has earned an Open Data badge for making publicly available the digitally‐shareable data necessary to reproduce the reported results. The data is available at https://doi.org/10.23684/odi-2017-00957. Learn more about the Open Practices badges from the Center for Open Science: https://osf.io/tvyxz/wiki. [ABSTRACT FROM AUTHOR]

Reliable and precise prediction of the rivers flow is a major concern in hydrologic and water resources analysis. In this study, multi-linear regression (MLR) as a statistical method, artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) as non-linear ones and K-nearest neighbors (KNN) as a non-parametric regression method are applied to predict the monthly flow in the St. Clair River between the US and Canada. In the developed methods, six scenarios for input combinations are defined in order to study the effect of different input data on the outcomes. Performances of the models are evaluated using statistical indices as the performance criteria. Results obtained show that adding lag times of flow, temperature and precipitation to the inputs improve the accuracy of the predictions significantly. For a further investigation, the aforementioned models are coupled with wavelet transform. Using the wavelet transform improves the values of Nash-Sutcliff coefficient to 0.907, 0.930, 0.923, and 0.847 from 0.340, 0.404, 0.376 and 0.419 respectively, by coupling it with MLR, ANN, ANFIS, and KNN models. [ABSTRACT FROM AUTHOR]

This paper presents an analysis of the effects of biased extended streamflow prediction (ESP) forecasts on three deterministic optimization techniques implemented in a simulated operational context with a rolling horizon test bed for managing a cascade of hydroelectric reservoirs and generating stations in Québec, Canada. The observed weather data were fed to the hydrological model, and the synthetic streamflow subsequently generated was considered to be a proxy for the observed inflow. A traditional, climatology-based ESP forecast approach was used to generate ensemble streamflow scenarios, which were used by three reservoir management optimization approaches. Both positive and negative biases were then forced into the ensembles by multiplying the streamflow values by constant factors. The optimization method's response to those biases was measured through the evaluation of the average annual energy generation in a forward-rolling simulation test bed in which the entire system is precisely and accurately modelled. The ensemble climate data forecasts, the hydrological modelling and ESP forecast generation, optimization model, and decision-making process are all integrated, as is the simulation model that updates reservoir levels and computes generation at each time step. The study focussed on one hydropower system both with and without minimum baseload constraints. This study finds that the tested deterministic optimization algorithms lack the capacity to compensate for uncertainty in future inflows and therefore place the reservoir levels at greater risk to maximize short-term profit. It is shown that for this particular system, an increase in ESP forecast inflows of approximately 5 % allows managing the reservoirs at optimal levels and producing the most energy on average, effectively negating the deterministic model's tendency to underestimate the risk of spilling. Finally, it is shown that implementing minimum load constraints serves as a de facto control on deterministic bias by forcing the system to draw more water from the reservoirs than what the models consider to be optimal trajectories. [ABSTRACT FROM AUTHOR]

Permafrost strongly controls hydrological processes in cold regions. Our understanding of how changes in seasonal and perennial frozen ground disposition and linked storage dynamics affect runoff generation processes remains limited. Storage dynamics and water redistribution are influenced by the seasonal variability and spatial heterogeneity of frozen ground, snow accumulation and melt. Stable isotopes are potentially useful for quantifying the dynamics of water sources, flow paths and ages, yet few studies have employed isotope data in permafrost-influenced catchments. Here, we applied the conceptual model STARR (the Spatially distributed Tracer-Aided Rainfall–Runoff model), which facilitates fully distributed simulations of hydrological storage dynamics and runoff processes, isotopic composition and water ages. We adapted this model for a subarctic catchment in Yukon Territory, Canada, with a time-variable implementation of field capacity to include the influence of thaw dynamics. A multi-criteria calibration based on stream flow, snow water equivalent and isotopes was applied to 3 years of data. The integration of isotope data in the spatially distributed model provided the basis for quantifying spatio-temporal dynamics of water storage and ages, emphasizing the importance of thaw layer dynamics in mixing and damping the melt signal. By using the model conceptualization of spatially and temporally variable storage, this study demonstrates the ability of tracer-aided modelling to capture thaw layer dynamics that cause mixing and damping of the isotopic melt signal. [ABSTRACT FROM AUTHOR]

Hydrologic studies in northern landscapes indicate there is a critical need to explore how arctic stream discharge patterns and water budgets may be shifting in response to climate warming. The focus of this study was to: (1) assess the pattern of runoff out of Polar Bear Pass, a low-gradient watershed (75400N, 98300W), during two contrasting spring/summer seasons: 2012 (warm, early melt) versus 2013 (cool, late melt); (2) quantify the seasonal water budgets; and (3) place these results in the context of other arctic basin studies. The end-of-winter snowpack was quantified using a terrain-based approach. A physically based snowmelt model using local weather station data provided daily melt estimates. Streamflow at the eastern outlet was estimated using the mid-section velocity approach. Snow water equivalent (SWE) was higher in 2013 while snowmelt began and ended earlier in 2012. Stream hydrographs showed a rapid rise in flow driven by meltwater from the northern part of the Pass in 2012. This was followed by a series of secondary peaks, melt contributions from the southern end. In 2013, the largest runoff peaks came from the southern sector. Runoff ratios and water budgets varied between the two years, and runoff in 2013 was similar to High Arctic watersheds in the early 1970s. [ABSTRACT FROM AUTHOR]

• Most candidate models have difficulties in replicating Prairie streamflow. • Models often lack accurate representation of Prairie non-contributing area dynamics. • Models often compromise AET simulation to achieve good streamflow simulation. • KGE streamflow performance is expected to be approximately 0.17 over the Prairies. • Snow dynamics are well represented in most candidate models. Intercomparison studies play an important, but limited role in understanding the usefulness and limitations of currently available hydrological models. Comparison studies are often limited to well-behaved hydrological regimes, where rainfall-runoff processes dominate the hydrological response. These efforts have not covered western Canada due to the difficulty in simulating that region's complex cold region hydrology with varying spatiotemporal contributing areas. This intercomparison study is the first of a series of studies under the intercomparison project of the international and interprovincial transboundary Nelson-Churchill River Basin (NCRB) in North America (Nelson-MIP), which encompasses different ecozones with major areas of the non-contributing Prairie potholes, forests, glaciers, mountains, and permafrost. The performance of eight hydrological and land surface models is compared at different unregulated watersheds within the NCRB. This is done to assess the models' streamflow performance and overall fidelity without and with calibration, to capture the underlying physics of the region and to better understand why models struggle to accurately simulate its hydrology. Results show that some of the participating models have difficulties in simulating streamflow and/or internal hydrological variables (e.g., evapotranspiration) over Prairie watersheds but most models performed well elsewhere. This stems from model structural deficiencies, despite the various models being well calibrated to observed streamflow. Some model structural changes are identified for the participating models for future improvement. The outcomes of this study offer guidance for practitioners for the accurate prediction of NCRB streamflow, and for increasing confidence in future projections of water resources supply and management. [ABSTRACT FROM AUTHOR]

With drainage from the Waterton‐Glacier International Peace Park, the Waterton River was dammed in 1964 to trap spring flow and permit offstream diversion for irrigation. Field observations in the 1980s indicated some decrepit riparian woodlands suggesting drought stress of the black and narrowleaf cottonwoods (Populus trichocarpa, P. angustifolia) due to insufficient in‐stream flows. Subsequently, an environmental flow regime commenced in 1991 and provided “functional flows,” deliberately regulating in‐stream flow components intended to restore ecological processes and particularly (1) an increase of the minimum flow from 0.93 to 2.27 m3/s (mean discharge 21.9 m3/s) and (2) flow ramping, gradual recession after the spring peak. This study investigated the historic flow patterns and the growth, population age structure, and spatial distributions of riparian cottonwoods along the free‐flowing upstream and regulated downstream reaches over four dam operations intervals: the free‐flowing pre‐dam condition; the initial dammed interval to the mid‐1970s; a post‐dam and drought interval in the 1980s; and with the environmental flow regime. Analyses of sapling, shrub‐, and tree‐sized cottonwoods included tree ring analyses to determine ages and growth patterns, and distributions were assessed relative to streamside elevations and sediment textures. These indicated that there has been progressive cottonwood colonization after damming but the colonization band dropped in elevation with the reduced flow regime and the future woodlands could become narrower. The tree ring analyses indicated that the growth of established trees benefited from the functional flows and the increase in minimum flow was probably particularly beneficial to the riparian cottonwoods. [ABSTRACT FROM AUTHOR]