Search

[Display omitted] • Riverbed sediment thickness affect the hydraulic connection between river water and groundwater. • Temporal and spatial changes in DO concentration dominated the iron-manganese reduction zone. • Organic matter-bound iron-manganese oxidation/iron-manganese oxides reduction mainly controlled Fe and Mn release. • The predominant microbial species and its abundance were resilient to riverbed sediment thickness. • Iron-manganese migration and transformation mechanism supported significant information for protecting groundwater quality. Riverbed scouring and siltation, affected by variations in river flow and sediment concentration, can cause changes in the sediment thickness, leading to the complex feedback between riverbed permeability and nutrient–reactive transport processes. Currently, the migration and transformation of iron and manganese under changed riverbed sediment thickness have not been taken into account. Based on indoor one-dimensional soil column simulation experiments, along with in-situ monitoring, Rhizon pore water sampling, and 16 s rRNA high-throughput sequencing technologies, this study revealed the migration and transformation patterns of iron and manganese in the river infiltration zone and their contribution rates under different sediment thickness conditions. The results demonstrated that as the riverbed sediment thickness increased, the river infiltration rate and sediment permeability coefficient demonstrated a significant decrease over time. For instance, a 5 cm sediment thickness can decrease the sediment permeability coefficient by 30 % within 32 d of infiltration, expand the disconnection zone to 25 cm, narrow the oxidation–reduction zone, and cause the iron and manganese reduction zone to evolve from a single peak to a double peak pattern. Additionally, as the sediment thickness increased, the contribution of organic matter bound iron-manganese oxidation and iron-manganese oxide reduction to Mn2+ and Fe2+ concentrations in sediment pore water decreased, while the contribution of adsorption and complexation-precipitation increased. This study holds great significance for ensuring the safety of drinking water supply and promoting the sustainable utilization of groundwater resources. [ABSTRACT FROM AUTHOR]

• Nitrate concentrations in groundwater and surface water in Dorset (UK) were studied. • Both seasonal and event-based variations were found in groundwater and surface water. • Nitrate peaks in springs and streams lag groundwater nitrate peaks by days to weeks. • Lag times reflect celerity and velocity of preferential flow in the chalk aquifer. Short-term nitrate variations in streams are common, but the extent of groundwater contributions to such variability has been unclear. Analysis of well, spring and stream nitrate concentrations plus well water levels and stream flow for a chalk aquifer in Dorset (UK) shows that nitrate variation in streams lags discharge peaks by days to weeks. Spring monitoring also shows a short lag time, showing that high nitrate concentrations are rapidly transmitted from aquifers to streams. The lag time reflects the difference between the celerity (pressure wave speed) and the advective velocity of groundwater flow through the fracture network. Results give celerities of ∼500 m/d, advective velocities of 100–200 m/d, and kinematic porosities of ∼0.001. In addition, long-term nitrate data gives time lags of several decades, reflecting the matrix porosity of ∼0.3. Consequently, nitrate can be used an environmental tracer for both preferential flow through fracture networks and for retarded matrix flow and storage in this dual-porosity bedrock aquifer. Dual porosity is common in bedrock aquifers, so it is possible that this methodology may be widely applicable to understand the hydraulic characteristics of bedrock aquifers and to explain short-term nitrate variation in streams. [ABSTRACT FROM AUTHOR]

[Display omitted] • A systematic method for optimizing regional hydrological LSTMs. • An efficient approach to identify optimal network configurations. • Distinct configurations show statistically significant different performance metrics. • Thoughtful configuration selection post-random search in regional hydrology. • Highly accurate hourly streamflow and water level predictions, up to 0.97 NSE/KGE. This paper introduces a novel approach for hyperparameter optimization of long short-term memory networks (LSTMs) to achieve highly accurate hourly streamflow and water level predictions in the realm of regional rainfall-runoff modeling. Leveraging simultaneous systematic hyperparameter optimization of 10 distinct hyperparameters by Random Search, the study achieves high accuracy in terms of predictions across 40 humid flashy catchments in Basque Country, north of Spain. By carefully designing the search space and incorporating domain expertise, the approach quickly converges to optimal and highly accurate network configurations with both efficiency and efficacy. LSTMs ingested precipitation, temperature, and potential evapotranspiration as inputs to predict 2 targets of streamflow and water level, in an hourly timestep. On the test set, the optimized LSTM networks accurately predicted streamflow and water level with Nash-Sutcliffe (NSE) and Kling-Gupta (KGE) efficiencies as high as 0.97, in one of the catchments. Across all 40 studied catchments, the overall average NSE and KGE values for streamflow were 0.89 and 0.87, respectively; water level exhibited average NSE and KGE scores of 0.91 and 0.92. Moreover, statistical analysis reveals significant differences in the performance of the 2 distinct optimized network architectures in different hydrological catchments, underscoring the importance of deliberate network configuration selection post-random search. This selection process is vital for achieving higher performance in as many catchments as possible. The findings highlight opportunities for enhancing the "learning maturity" of regional hydrological deep learning LSTM networks. This research provides valuable insights for researchers and practitioners involved in optimizing regional hydrological deep learning models for a variety of applications and on new datasets. [ABSTRACT FROM AUTHOR]

• We explore spatial scales at which irrigation infrastructure predicts flow intermittency. • Local temperature and precipitation had the strongest correlation with flow intermittency. • Infrastructure and discharge correlations vary across subreach scales. • Factors at <7 km have clearest influence on spatial flow intermittency patterns. • Scale-sensitive models are needed to inform environmental flow management. Fluvial ecosystems are vital for biodiversity and human welfare but face increasing threats from flow intermittency caused by climate change and water development. We analyzed 12 years of spatially explicit daily river drying from the Rio Grande, a historically perennial and regulated river in the North American desert southwest. Using multivariate autoregressive state space models, we identified the spatial scales at which irrigation diversions and return flows are correlated with flow intermittency as well as local precipitation, temperature, and river discharge. Influences of temperature and precipitation were not detected at the whole reach scale (∼154 km). However, they were significant and similar across various subreach scales (7 to 86 km-long subreaches) with temperature's estimated effect on flow intermittency exceeding precipitation by 2.5 times. The influences of irrigation infrastructure and discharge were not detected at the whole reach scale either but had varied effects across subreach models. Differential process variance and covariance across subreach scales indicated factors (possibly geomorphic) at fine (<7 km) spatial scales influenced spatial patterns of flow intermittency. Process variance decreased by 98 % from reach to subreach models, suggesting scale-sensitive models are needed to inform management strategies that focus on reducing flow intermittency. [ABSTRACT FROM AUTHOR]

• This study aims to propose a postprocessing framework. • Various postprocessors are applied to improve the streamflow predictions. • The COP-BMA integrates outputs of postprocessors for probabilistic prediction. • Both daily and monthly time scales are examined during Hurricane Harvey event. • Results indicate that the proposed method enhances the accuracy of prediction. This study proposes a two-step probabilistic post-processing approach that combines different machine learning-based postprocessors through the Copula-Embedded Bayesian Model Averaging (COP-BMA) method to improve the performance of a hydrological model for streamflow predictions. The proposed approach serves a two-fold purpose: firstly, it aims to enhance the accuracy of streamflow predictions, and secondly, it provides probabilistic results that implicitly address the structural uncertainty inherent in different postprocessing methods. We validate our approach by applying it to the Conceptual Functional Equivalent, a lumped hydrologic model utilized for simulating extreme floods during Hurricane Harvey. The validation is conducted across twelve distinct watersheds in the Southeast Texas region at both daily and monthly scales. The findings indicate that the proposed framework significantly enhances the performance of the hydrologic model across the studied watershed. Specifically, on a daily time scale, there is a 23% and 53% improvement in the NSE and KGE respectively, while on a monthly time scale, the framework enhances NSE by 21% and KGE by 25%. Additionally, the MAE (cms) was notably reduced from 4.64 to 2.23 on the daily scale, and from 2.8 to 1.65 on the monthly scale. [ABSTRACT FROM AUTHOR]

• Most of the surface runoff of major rivers significantly decreased in Changbai Mountain. • The climate sensitivity of seasonal surface runoff was examined. • Extreme runoff exhibited a sensitive response to extreme precipitation. • Precipitation was the main factor affecting runoff, especially in snowless period. Due to the sensitivity of runoff to changes in seasonal snowpack, the changes in surface runoff within high-latitude and high-elevation areas are more complex than in other areas. However, this sensitivity is often overlooked when correlating changes in runoff with meteorological variables. Here, we analyze changes in surface runoff during snowmelt and snowless periods and their sensitivity to climate change from 1960 to 2019 in the upstream watersheds of the Songhua, Yalu, and Tumen Rivers in Changbai Mountain. Most runoff indices, except for low flow in the Yalu River, exhibited decreasing trends during snowmelt and snowless periods. Precipitation was the primary variable responsible for surface runoff during snowless and snowmelt periods. Notably, rainfall replenished all runoff indices during snowless periods, particularly under high flow conditions. The average and high flows were also primarily supplemented by rainfall during snowmelt periods. However, low flow was positively correlated with snowmelt in the Songhua and Tumen Rivers and with temperature in the Songhua, Yalu and Tumen Rivers during snowmelt periods. Additionally, extreme runoff events were closely associated with extreme precipitation. These findings demonstrate the seasonal and index-dependent climate sensitivity of surface runoff within Changbai Mountain, offering insights into the surface runoff response mechanism to climate change in high-altitude mountainous areas. Under future climate change conditions, frequent and intense extreme precipitation events may increase the risk of flooding, owing to the high sensitivity of high flows to precipitation, posing a serious threat to regional ecological security. [ABSTRACT FROM AUTHOR]

• Temporal heterogeneity poses challenges to runoff prediction in semi-arid basins. • Our data segmentation approach enhances period-specific rainfall-runoff modeling. • Attention-based deep learning models excel in extreme flood forecasting. • TCN effectively captures temporal hydrological relationships in semi-arid basins. The increasing threats of flash floods and water scarcity in semi-arid regions necessitate high-quality streamflow forecasting with process-based or data-driven models. However, temporal heterogeneity of hydrological patterns and infrequent flood events present challenges for accurate streamflow forecasting. To address this issue, we purpose an effective modeling approach that combines a novel data segmentation method, BPX, with an attention-based deep learning (DL) model. BPX integrates Bai-Perron analysis and XGBoost to accurately identify dry and wet periods from hydro-meteorological variables both in the past and future times. The DL model introduces an attention mechanism to temporal convolutional network (TCN) to enhance its ability in handling heterogeneous time-series data. We demonstrate the effectiveness of our approach using nearly 20 years of hydro-meteorological data from the Wei River basin in China. Various modeling methods are compared, including two process-based models (Xinanjiang and GR4J) and four DL models (classical RNN, GRU, LSTM and TCN) with or without the attention module. After systematic benchmarking, it is found that both the BPX segmentation and the attention mechanism can improve streamflow forecasting with DL. Among the various modeling approaches, BPX-TCN exhibits the best overall performance throughout the prediction period, achieving the highest Kling-Gupta efficiency of 0.91 (compared to 0.45–0.88 for the other models), while BPX-TCN-attention provides more accurate predictions of floods and multi-step predictions. In similar applications in semi-arid regions, our approach may serve as a valuable reference where temporal heterogeneity is significant and poses challenges to traditional modeling methods. [ABSTRACT FROM AUTHOR]

Integrated hydrological modeling is gaining popularity due to its mechanistic representation of the surface and subsurface processes. However, estimating the parameters of such process-based models can be computationally expensive if careful consideration is not given to the length of streamflow observations used during model calibration. Here we evaluate the influence of the calibration period, the role of streamflow information content, and the gage location in parameter learning and calibration of a fully integrated hydrological model, the Advanced Terrestrial Simulator (ATS). We conducted the study at the Upper Neversink River Watershed within the Delaware River Basin, where streamflow observations are available at 11 gages with varying record lengths. We leveraged a recently proposed knowledge-informed deep learning technique for parameter estimation. To assess the impact of observation period and gage location, model parameters were learned on scenarios using different chunks of streamflow observations, including (1) using only one year or consecutive multiple years of streamflow observations at the watershed outlet and (2) using one shared year of observations at each of the sub-catchment gages, with the period from 1991-10-01 through 1999-09-30 as the overall calibration period. Using the estimated parameters, ATS was rerun for each scenario and evaluated on a subsequent period from 1999-10-01 through 2002-09-30. Results show that the basin outlet discharge prediction is mostly improved when using at least four years of observations for parameter estimation. Further, the performance of the calibrated ATS run correlates with the information content of the observed streamflows, suggesting that the information-theoretic metrics could be indicators for selecting the observation period for parameter estimation. Finally, we find that observations from an informative gage can be used in learning parameters to predict the streamflow at a nearby gage, which would potentially lower the computational expense by reducing the watershed domain used in calibration. Our success underscores the potential of using information theory to achieve robust model parameter estimation on a reduced computational budget. • We studied the impact of the observation length and information on integrated hydrological model calibration. • Information-theoretic metrics unveil to what extent observed streamflow can be used in learning model parameters. • Informative observations can be used to constrain the model to predict the streamflows at a nearby gage. [ABSTRACT FROM AUTHOR]

• Two eco-hydrological models with PET or ET inputs are developed and compared. • Analysis of climate and vegetation change in energy- and water-limited catchments. • Daily streamflow and leaf area index are simulated and projected simultaneously. • Highlights the importance of the high-accuracy evapotranspiration inputs. Evapotranspiration (ET) is an essential variable linking hydrological and ecological processes and is typically modeled as a function of potential evapotranspiration and soil moisture in traditional hydrological models. However, commonly used empirical ET models typically do not recognize the underlying vegetation dynamics. This can have implications when models are extrapolated under future climate change. In this study, the traditional HYMOD-BVM (THV) eco-hydrological model is adopted as a benchmark model. A modified HYMOD-BVM (MHV) is developed using an actual ET substitute for the traditional PET input to consider vegetation dynamics under climate change. The THV and MHV models are compared to evaluate how streamflow (Q) and leaf area index (LAI) vary under future climate in the Florentine River (FR, energy-limited) catchment and Murray River (MR, water-limited) catchment in Australia. Six global climate models (GCMs) from the latest Coupled Model Intercomparison Project (CMIP6) are bias-corrected and employed in the calibrated THV and MHV models, and ensemble mean results are analyzed. Results suggest that the streamflow projected by the MHV model tends to be higher than that from the THV model, while LAI presents an opposite trend. The energy-limited catchment appears more susceptible to climate change whereas larger vegetation effects are seen in the water-limited catchment. Overall, our research highlights the effects of evapotranspiration on future climate change scenario analysis and prompts the need for the development of spatial and temporal continuously accurate ET models for both current and future climates. [ABSTRACT FROM AUTHOR]

• The paired-catchment approach was improved and successfully utilized to detect reservoir impacts on flow regime alteration. • Reservoirs primarily homogenize temporal runoff variations, reshape flow duration curve, and reduce runoff sensitivity to precipitation. • The extent of flow regime alteration caused by reservoirs varies depending on their functionalities and regulatory capacities. The 20th century witnessed a substantial surge in global reservoir construction, representing a significant milestone and resulting in a profound anthropogenic alteration of flow regimes. Despite this surge, a comprehensive large-sample quantification of reservoir impacts on flow regime alteration has been impeded by limitations in statistical and modeling methodologies. Here we present an improved paired-catchment method tailored to discern the impact of reservoirs on flow regime within the context of climate change. Employing this method, we identified and analyzed 177 paired catchments from a selection of over 12,000 global catchments. Our findings highlight the effectiveness of the improved method, demonstrating that reservoirs predominantly alleviate high flows while enhancing low flows. This serves to moderate the annual periodicity of streamflow signals while amplifying their low-frequency components. Furthermore, predetermined water releases play a significant role in diminishing the seasonality of low flows and reducing the sensitivity of runoff to precipitation. Additionally, the extent of flow regime alteration caused by reservoirs is closely linked to their functionalities and regulatory capacity. For example, reservoirs designed for irrigation and water supply show minimal enhancement of low flow values, while the magnitude of reservoir-induced changes rises with higher regulatory capacity. The improved method can be further employed in the future to investigate the impact of reservoirs on extreme hydrological events such as droughts and floods. It emerges as a crucial scientific tool for effective water resource management and the mitigation of water-related disasters in the Anthropocene era. [ABSTRACT FROM AUTHOR]

[Display omitted] • First time calibration of digital filter with natural tracers in Mediterranean region. • Efficient calibrated digital filter for long-term functional baseflow analysis. • Mountainous Mediterranean rivers are seasonally and interannually baseflow-dominated. • Larger baseflow contribution sustaining Mediterranean rivers during severe droughts. • Baseflow and groundwater can buffer impacts of climate change on the short-term. Baseflow defined as the contribution of groundwater and delayed subsurface flow, is generally described by indices such as the baseflow index (BFI), the long-term ratio of baseflow to streamflow. Characterizing baseflow is challenging because it is not directly accessible, particularly in sensitive areas to climate change like the mountainous Mediterranean regions with hard-rock aquifers. In these areas, little is known about the temporal and spatial contribution of baseflow to streamflow. This study aimed to address this scientific gap by using the first calibrated and validated recursive digital filter with natural tracers for such regions. Two watersheds of Corsica (France), the Tavignanu and the Fiumaltu, were chosen as application models for calibration and validation, respectively, due to their representative characteristics of these regions. Various methods exist for baseflow analysis, including empirical, analytical and physical methods. Empirical and analytical methods, such as digital filters, are easy to use but lack physical significance. Conversely, physical methods using natural tracers, such as isotopes and ions, are more physically representative but are difficult to implement. By calibrating and validating analytical methods with tracers in this specific context, we were able to leverage the strengths of both approaches for effective long-term baseflow analysis. This innovative approach confirmed that baseflow was the main contributor to river flow during dry months, with a contribution of 88–90 % for both watersheds. Notably, baseflow remained a significant contributor during periods of high flow, averaging 54 % for the Tavignanu and 60 % for the Fiumaltu. Over a 25-year period, the mean annual contribution was substantial, at 73 % for the Tavignanu and 77 % for the Fiumaltu. In a particularly dry year, which are expected to become more common with climate change, baseflow contribution was even higher, confirming its origin from deep groundwater and delayed subsurface components. Our results demonstrate that, in the context of climate change, groundwater and delayed subsurface flow will play an increasingly important role in sustaining rivers, connected ecosystems, as well as many human activities in the mountainous Mediterranean regions. [ABSTRACT FROM AUTHOR]

• Inundation areas are divided into three areas for different flood drivers. • The water tracer model is used to simulate the process of urban river flow. • The impact of urban inland rivers on flooding in the coastal city is analyzed. • Flood regulation measures based on inland rivers are proposed. Urban inland rivers play a significant role in mitigating flooding. However, the influence of their discharge on urban flooding affected by different flood drivers remains to be fully explored. Taking a catchment area in Haikou City, Hainan Province, China as an example, in light of the urban flooding model coupled with the water tracer method, this study investigates the degree of inundation in each flood zone under different inland river flows. The flood zones are determined by the flooding simulation results of three cases, and then are verified by the zonation based on the tide tracer approach. Finally, the tracer method is used to study the impact of urban inland rivers on flooding and the connection of hydraulic and water volume between rivers and pipelines. The results indicate that the inundation area decreased by 3.71 % as the river flow increased from 15 to 25 m3/s. In the process of river flow rising from 5 to 45 m3/s, The average flooding depth in the rainfall-affected area climbs slowly from 0.53 m to 0.62 m; In the tidal-affected area, the average flooding depth drops from 0.52 m to 0.12 m when the inland river flow increases from 5 to 25 m3/s, and rises from 0.12 m to 0.54 m when the inland river flow continues from 25 to 45 m3/s; The percentage of pipelines grows from 18.5 % to 28.0 %. This study provides valuable scientific insights and support for flood prevention planning and flood control strategies in coastal cities. [ABSTRACT FROM AUTHOR]

• Four tension water storage capacity distribution curves are investigated. • The topography-based GTHYMOD performs best over 460 Australian catchments. • Improved streamflow performance and spatial characterization ability are achieved. • High KGE is more likely to be obtained in humid regions. • The tension water storage capacity is significantly related to elevation. Accurately characterizing the spatial variability of tension water storage capacity (TWC) within a catchment is challenging due to limited in-situ hydrologic data availability. Conventional conceptual rainfall-runoff models typically rely on an empirically specified TWC distribution. However, this empirical distribution lacks a physical foundation and fails to effectively redistribute critical hydrologic components, such as local capacity and contributing area, to real-world contexts. To overcome this limitation, the topographic wetness index (TWI) and its generalized form (GTWI) are introduced to bridge local topographic information with hydrologic components. Four TWC distribution curves are contrived based on the empirical parabolic distribution, empirical linear distribution, TWI, and GTWI, respectively. The effects of these alternate TWC distributions on streamflow are investigated within the framework of the HYdrologic MODel (HYMOD) across 460 Australian catchments. The results illustrate that the GTWI-based HYMOD (GTHYMOD) outperforms other models in terms of daily streamflow, with high Kling-Gupta Efficiency (KGE) attained in 74.8% of the study catchments during the validation period. The eastern coast of Australian catchments presents a superior streamflow performance compared to that in the western coast. GTHYMOD demonstrates its superiority in characterizing spatial variability, an aspect HYMOD lacked. This study has the potential to refine the empirical TWC distribution from a physical perspective and advance our comprehension of underlying hydrologic behaviors. [ABSTRACT FROM AUTHOR]

• A novel self-contained hydrodynamic observation platform system was deployed to monitor flow and sediment with ultra-high-resolution in the Yarlung Tsangpo River. • Temporal-spectral analysis with ultra-high-resolution revealed the asynchronous variations of flow and sediment over nearly two months. • The varying behaviors of flow and sediment in different flood events was distinguished using ultra-high-resolution data. Traditional monitoring techniques of flow and sediment are inadequate for the sparsely populated, the rivers in the Qinghai-Tibetan Plateau region which are experiencing amplified climate change. Inspired by the Euchiloglanis fish, this study developed a unique self-contained hydrodynamic observation platform to withstand the high velocities of mountain rivers typically encountered in mountain rivers. The platform, equipped with sensors, was successfully deployed in the Yarlung Tsangpo River, collecting ultra-high-resolution data on flow and sediment over a period of nearly two months. Performance assessments showed that the median value of discharge per unit width measured and suspended sediment concentration by this system were basically consistent with those daily measurements from the hydrological station. Meanwhile, the superiority of the system was reflected in the more significant daily average change, especially at the flooding period. The data unveiled significant minute-level hydrological fluctuations and a complex interplay of factors affecting the river's flow and sediment transport. A wavelet coherence analysis revealed asynchrony changes in short-period flow and sediment variables. Persistent short-period and coherent long-period signals are significantly displayed during flood events. The platform promises future applications, such as automated benthic observations and replacing traditional 'lead-line' methods. Despite acknowledging system limitations of underwater attitude, burying tendency, and real-time data transmission capabilities, the study highlighted the crucial role of this system in understanding and managing river dynamics, particularly amid escalating environmental threats. [ABSTRACT FROM AUTHOR]

• Streamflow exhibits diurnal fluctuation throughout the year, reaching up to 36% of the daily mean. • Choosing proper timing for daily discharge observation can reduce bias due to diurnal fluctuation. • As the drought increases, the impact of diurnal fluctuation grows more significant. • Groundwater levels in riparian zones and hillslopes fluctuate in varied phase during rainy seasons. • Evapotranspiration rate in the riparian zone predominates diurnal fluctuation in rainy seasons. Diurnal fluctuation in streamflow during rainless periods provides valuable signals for the calibration of hydrological models and comprehension of hydrological processes. Previous research has primarily focused on this phenomenon in the context of single-factor analysis in the Mediterranean, Maritime, and Alpine climates of Middle Europe and Western US. This study is based on the Xitaizi Experimental Watershed (XEW), a typical mountainous headwater watershed situated in the semi-humid monsoon North China Plain. Seven years of high-frequency observation data were used to investigate the characteristics of streamflow diurnal fluctuation, its driving mechanisms, and the propagation pathways of these fluctuation signals through hillslope-riparian-stream. The results are as follows: (1) Streamflow exhibits diurnal fluctuation during rainless periods throughout the year, with variations up to 36% of the daily mean and distinct patterns in rainy (peak at night) and dry seasons (peak in the afternoon). Choosing the proper timing for daily discharge observation can reduce bias due to fluctuations. (2) In rainy seasons, the impact of discharge fluctuation is significant, particularly in years with lower discharge. Meanwhile, the soil moisture in both the hillslope and the riparian zone, and the groundwater level in the riparian zone exhibit a daytime decline and a nighttime rise. The groundwater level in the hillslope fluctuates with a lag time of approximately 6 to 9 hours. (3) Numerical simulations and cross-correlation analysis demonstrate that the dominant driving mechanism of diurnal fluctuations in rainy seasons is the diurnal evapotranspiration variation in the riparian zone, while the influence of aquifer hydraulic conductivity is insignificant. [ABSTRACT FROM AUTHOR]

• We investigate the hydraulic complexity metrics based on laboratory and field data. • We study the kinetic energy and coherent structures related to aquatic organisms. • We quantify biologically flow patterns occurring at different spatial scales. • We find that flow circulations are potential feeding locations. In non-uniform river flows, hydrodynamic features, such as gradients of velocity in flow directions, are of particular importance for explaining the abundance of aquatic habitats. Hydraulic complexity metrics referred to as M 1 , M 2 , and M 3 play an important role when it comes to the analysis of habitat metrics on a 3D spatial level. Parameter M 1 is proportional to the drag force experienced by an organism, parameter M 2 represents how much more energy an organism must expend if it moves from the lower velocity to the higher velocity location, and parameters M 3 illustrates the circulation in flow. The specific aim of the present study is to apply those parameters to characterize, under different flow conditions, the cross-sectional distribution of kinetic energy and coherent structures which are both relevant for many aquatic organisms. To this aim, laboratory data as well as field observations along Tiber River, in central Italy, were considered and the hydraulic complexity metrics were investigated in dimensionless form. On the laboratory-scale, the dimensionless parameters M 1 *, M 2 * and M 3 * identify the velocity gradient related to the high/low cross-sectional velocity and the high/low vorticity areas in selected cross-sections along the flume. Then, the evaluation of the parameter M 2 * in the horizontal plane allows to verify the relation between the localization of the high/low kinetic energy areas in the longitudinal direction. For the field-scale, the parameters were investigated, under high, moderate and low flow conditions, in a gauged site in the Tiber River. The results indicate that an aquatic organism should spend more energy where M 3 *, which is related to flow circulation, assumes high values. Furthermore, significant values of M 1 * and M 2 * are observed, which are linked to gradients in the cross-sectional distribution of velocity. These values are predominantly found at the river centerline for M 1 * and at the banks for M 2 *. In terms of echo-hydraulics, the results based on both laboratory and field data indicate that they are complementary, showing that for the larger magnitudes of M 1 and M 3 , which are related to the kinetic energy and flow circulation, respectively, an aquatic organism should spend more energy in these zones. Overall, the results based on both laboratory and field studies suggest that parameters M 1 and M 2 are inversely linked, i.e., M 1 decreased with increasing M 2 , while, there is no relationship between parameters M 3 and M 1. The findings of the present research would be of particular interest in quantifying biologically important flow patterns occurring at different spatial scales within different streams and under different flow conditions. [ABSTRACT FROM AUTHOR]

• SAC-SMA model was evaluated using machine learning-based limits-of-acceptability over streamflow. • Streamflow-based signatures were also used to identify physically unrealistic simulations. • All the 1-million parameter sets tested in this study were rejected as unfit-for-purpose. • Low flows were critical in identifying behavioral and non-behavioral models. Hydrological models are evaluated by comparisons with observed hydrological quantities such as streamflow. A model evaluation procedure should account for dominantly epistemic errors in hydrological data such as model input precipitation and streamflow and avoid type-2 errors (rejecting a good model). This study uses quantile random forest (QRF) to develop limits-of-acceptability (LoA) over streamflows that account for uncertainties in precipitation and streamflow values. A significant advantage of this method is that it can be used to evaluate models even at ungauged basins. This method was used to evaluate a hydrological model –Sacramento Soil Moisture Accounting (SAC-SMA) – over the St. Joseph River Watershed (SJRW) for both gauged and hypothetical ungauged scenarios. QRF defined wide LoAs that yielded a large number of models as behavioral, suggesting the need for additional measures to develop a more discriminating inference procedure. The paper discusses why the LoAs defined by QRF were wide, along with some ways to define more discriminating LoAs. To further constrain the model, five streamflow-based signatures (i.e., autocorrelation function, Hurst exponent, baseflow index, flow duration curve, and long-term runoff coefficient) were used. The combination of LoAs over streamflow and streamflow-based signatures helped constrain the set of behavioral models in both the gauged and the ungauged scenarios. Among the signatures used in this study, the Hurst exponent and baseflow index were the most useful ones. All the 1-million models evaluated in this study were eventually rejected as unfit-for-purpose. [ABSTRACT FROM AUTHOR]

This paper introduces and applies a novel methodology that integrates watershed modeling with the traditional regional flood frequency analysis. The methodology introduces a novel site-similarity measure that relies on hydrologic simulations and accounts for the effect of land depressions on streamflow generation. The new measure is tested along with other traditional measures for regional flood frequency analysis in the Canadian prairies. The case study is chosen carefully to critically test the new methodology. An application of 30 combinations of the new and traditional site similarity measures is assessed for pooling 109 sites. The homogeneity of the clustered groups is evaluated, and different probability distributions are applied to describe at-site and regional annual maximum flows. The results present enhanced groups' homogeneity when the new measure is employed due to a better representation of the hydrologic similarity between the pooled sites. Furthermore, the regionally estimated quantiles are found susceptible to the chosen site similarity measures in the pooling process, which highlights the importance of considering the proposed measure that describes a key hydrologic aspect when land depressions exist. [ABSTRACT FROM AUTHOR]

• A novel model can capture hydrological heteroscedasticity and non-Gaussianity. • Our model generates conditional mixture distributions with time-varying parameters. • Number of mixture components is identified with Akaike information criterion. Improving probabilistic streamflow prediction requires proper characterization of residual errors in hydrological models. Previous studies often adopt data transformation techniques to stabilize the variability of residual errors to make them more Gaussian. Here, we introduce a novel residual error model that directly captures the intrinsic properties of residual errors, thereby avoiding potential pitfalls such as misinterpretation and conflict assumptions during data transformation. The proposed model, based on a Mixture Density Network (MDN), combines a neural network with a mixture density model to generate time-varying mixture distributions conditioned on the magnitude of streamflow and other variables. The model was examined using both synthetic data as well as data from a real catchment in the source of the Yellow River. Two model selection criteria, the Akaike information criterion (AIC) and the Bayesian information criterion (BIC), were adopted to choose the optimal number of mixture components. Compared to the benchmark model, which assumes homoscedastic Gaussian residual errors, our proposed model yields uncertainty intervals over 20% narrower and provides more accurate uncertainty coverage, particularly enhancing the capture of low and high flows. Besides, it effectively characterized heteroscedasticity and non-Gaussianity in residual errors following model assumptions checking. Further, we discovered that the optimal number of mixture components was better identified using AIC rather than BIC as the criterion. [ABSTRACT FROM AUTHOR]

[Display omitted] • Introduce an integrated approach of morpho-hydrologic and Nash-GIUH-based sub-basin (SB) prioritization. • Provide valuable insights to reduce quick peak flows (q p) in ungauged SBs. • In q p analysis, geomorphology plays a more significant role than the climatic factor. • AHP stands out as the superior model for SB prioritization due to its best alignment with the shape of GIUH avg. • The parameters of T c , R r and HI have considerable potential in estimating q p. In ungauged sub-basins (SBs), estimation of peak flow (q p) is inevitable in designing both structural and non-structural flood mitigating measures to control outflow into the Master River. However, the inadequacy of stream flow data is the pivotal challenge in estimating q p and design flood in ungauged SBs. Therefore, an alternative basin prioritization approach is adopted for flood mitigation and sustainable development of ten tributary SBs of the Shilabati River which is one of the most flood-prone tropical plateau fringe river basins in eastern India. Significant morphometric and hydrologic parameters are used for SB prioritization in terms of five different flood generation potentiality (FGP) models, namely compound factor, hazard degree, principal component analysis, weighted sum analysis and analytical hierarchy process (AHP). Results of the FGP models are compared with the q p of the Geomorphologic Instantaneous Unit hydrograph (GIUH) model linked with the Nash two-parameter gamma distribution function. Among a total of five Nash-GIUH models, the GIUH avg model (Nash-GIUH based on average dynamic velocity) is used to evaluate the FGP models. Results imply that AHP-FGP is the most efficient model for SB prioritization. According to AHP scores, the highest priority for flood mitigation is given to SB-6 due to its high FGP value followed by SB-3, and SB-4. Similarly, the highest q p (0.106 hr-1) and shortest time to peak (t p) (4.718 hrs) of GIUH avg are observed in SB-6, representing the hazardous nature of the SB. Morpho-hydrologic parameters such as time of concentration (T c), relative relief (R r) and hypsometric integral (HI) are identified as significant for the AHP-FGP model. In addition, moderate-strong correlations of T c , R r and HI with the q p of GIUH avg exhibit the association between geomorphologic characteristics and hydrologic behaviour of SBs. Therefore, this study presents an integrated approach of FGP and GIUH-based flood hazard assessment for sub-basin prioritization that will help to develop valuable insights to reduce the formation of quick q p in ungauged SBs that can trigger flooding in the Master River. [ABSTRACT FROM AUTHOR]

Determining the respective attribution proportions of climate change and land use change to streamflow variations in river systems is of increasing interest to researchers and practitioners tasked with managing river basins. This paper proposes an extension to established techniques of attributing the relative proportions of climate change (CC) and land use change (LUC) drivers to streamflow change by instead considering these proportions as distributed through a probability density function (pdf), rather than as a point value. The novel method is demonstrated for the River Tweed in the UK. Results are determined by the flow, temperature and precipitation data, and upon the algorithms used to identify change points in these vectors. The ratio of the LUC and CC attribution proportions (Land Use and Climate Change Attribution Proportions, LCAP) is more appropriately expressed as a vector of values, each associated with its own probability value within a probability density function. The paper demonstrates that the LCAP ratio pdf can vary considerably over time and that it is possible to track physical changes in the catchment in the evolution of the probability density function. Results show that the land-use change/climate change attribution proportion (LCAP) ratio varies over time and can be expressed as probabilistic estimate. It can be concluded, with a high degree of confidence, that for the Tweed catchment the LUC is a significant driver of streamflow change. Hence this finding may have implications for future catchment flood management utilising nature based solutions (NbS) to reverse landscape degradation and mitigate effects of climate change, provided that the economic and social costs are outweighed by the benefits. [ABSTRACT FROM AUTHOR]

• The hydrological response of a small Mediterranean catchment was investigated. • Soil moisture, groundwater, and streamflow response varied seasonally. • Streamflow response was controlled by antecedent wetness and precipitation depth. • Event water fractions showed a seasonal response across multiple spatial scales. • Time lags in peak response between multiple spatial scales showed a complex pattern. Understanding hydrological processes during dry periods in Mediterranean mountain catchments is critical due to the increasing frequency of drought episodes. In this work, we aimed at characterizing the effect of the seasonal variability of meteorological forcing on the hydrological response of a small mountain forested catchment in the Mediterranean region. We analyzed the hydrological response and its timing based on hydrometric and electrical conductivity (EC) data for a year in the nested (0.31–2 km2) Re della Pietra catchment, in Central Italy. We used a soil moisture-based metric to distinguish between wet and dry periods and performed EC-based hydrograph separations during these two periods. The results revealed the important role of seasonality as a meteorological forcing affecting soil moisture, groundwater, streamflow response, and stream event water fractions. Wet and dry periods were distinctly characterized by different streamflow, soil moisture, and groundwater responses. Event water fractions in streamflow also highlight the relevance of the seasonality in the meteorological forcing on runoff generation. Particularly, at the rainfall-runoff event scale, the combination of antecedent soil moisture and precipitation depth controlled the non-linear response of streamflow, groundwater, and different event water fractions in the wet and dry periods. Stream stages and event water fractions also varied across nested spatial scales. Antecedent moisture conditions triggered a faster streamflow response due to higher connectivity along the hillslope in the wet period, with higher event water fractions in the upper sub-catchments (25 %) compared to the lower sub-catchments (15 %). Conversely, in the dry period, higher event water fractions were registered at the outlet (11 %) than at the headwaters (7 %). Time lags between peak flows observed across the nested catchment showed a complex pattern, suggesting the interaction of multiple factors controlling the timing of streamflow peaks in the study area. These findings contributed to improve our mechanistic insights into the elusive seasonal hydrological patterns observed in Mediterranean mountain forested catchments. [ABSTRACT FROM AUTHOR]

• A novel data-driven regionalization modeling framework for PUB is proposed. • The DTW-KMeans technique is suitable for catchment classification. • Matching model complexity and number of training catchments can improve performance. • Catchment similarity plays a crucial role on regionalization modeling. Streamflow prediction in ungauged catchments is a challenging task in hydrological studies. Recently, data-driven models have demonstrated their superiority over traditional hydrological models in predicting streamflow in ungauged catchments. However, previous studies have overlooked the similarities between the training and the target catchments. Therefore, this study explores the role of catchment similarity in regionalization modeling using the publicly available CAMELS dataset. We employed the dynamic time warping-based KMeans (DTW-KMeans) time-series clustering technique to cluster the streamflow data from gauged catchments. We utilized the long short-term memory (LSTM) neural network to construct regional models for different classes of gauged catchment. Additionally, the mapping relationship between gauged catchment classes and static attributes was established using the random forest (RF). By combining the trained RF model with the static attributes of an ungauged catchment, we determined its class and used the corresponding regional LSTM to predict streamflow. To evaluate the effectiveness of the framework, we applied the classification-based regionalization modeling (CRM) and non-classification-based regionalization modeling (NRM) approach for comparison. The results indicate that: (1) The DTW-KMeans-based catchment classification method is generally accurate and reasonable; (2) the complexity of the LSTM model and the number of training catchments should be appropriately matched to improve streamflow prediction; and (3) catchment similarity plays a crucial role in regionalization modeling, the proportion of training catchments with high similarity to ungauged catchments significantly affects prediction results. [ABSTRACT FROM AUTHOR]

• Pre-event water concentration has a major role on hydrograph separation. • Event water concentration has little influence on hydrograph separation results. • Pre-event water concentration should be determined for each rainfall-runoff event. • Simple and complex hydrograph separation models give similar results. • Electrical conductivity yields similar results than Oxygen-18. Hydrograph separation assessment is crucial to understand stormflow generation at catchments worldwide. Tracer-based methods provide robust estimations of event (or new) and pre-event (or old) water fractions as they account for external and internal catchment hydrological behavior. While models of different mathematical and computational complexity are often used in tracer-based hydrograph separation studies, direct comparisons between those models are limited. Here, we compare hydrograph separation results yielded by the simplest Two-Component Mixing Model (TCMM) and a Tracer-based Streamflow Partitioning ANalysis model (TraSPAN) assumed to provide robust results as it combines conceptual rainfall-runoff modelling with tracers' mass balance. We carried out the analysis using high temporal frequency (sub-daily to sub-hourly) data of two tracers, Oxygen-18 and Electrical Conductivity (EC), monitored during 37 rainfall-runoff events with different hydrometeorological conditions in a high-Andean páramo catchment located at the Zhurucay Ecohydrological Observatory in southern Ecuador. Both approaches yield similar estimations of event and pre-event water fractions regardless of the tracer used as long as appropriate concentrations of event (C e) and pre-event (C p) water for the TCMM are determined. Although the estimate of C e has little influence with one rainfall sample collected during the event being sufficient to obtain reliable results, results hinge heavily on the estimate of C p. We found that the TCMM yields similar results than TraSPAN when C p is represented by the stream water concentration corresponding to a sample collected prior to the beginning of each of the events. We conclude that the combination of a simple framework (TCMM) with sub-hourly EC measurements provides reliable hydrograph separation results when representative C p samples are used. These findings will allow to lower the logistical and economical resources needed to adequately assess hydrograph separation and to carry out quasi-continuous assessments of flow partitioning with high accuracy in high-Andean páramo catchments. [ABSTRACT FROM AUTHOR]

• The performance of triple collocation-based merged ET in model calibration is evaluated. • A multi-variable calibration scheme that utilizes ET and streamflow features is proposed. • Merged ET exhibits higher accuracy in model calibration than raw remote sensing ET. • Multi-variable calibration scheme outperforms ET-based single-variable calibration. • The multi-variable calibrations show more pronounced improvements in arid catchments. Estimating model parameters for accurate hydrological simulations poses a challenge in ungauged catchments. Evapotranspiration (ET) data derived from remote sensing (RS) holds promise for parameter calibration in ungauged catchments. Nonetheless, uncertainties inherent in raw RS-based ET data can pose challenges in the calibration process. This study evaluates two calibration methods mitigating the impact of this uncertainty. The first, a single-variable calibration, utilizes merged ET data based on the triple collocation technique, while the second incorporates streamflow features alongside ET data for multi-variable calibration. Applying the variable infiltration capacity model to 49 Chinese catchments revealed the superiority of the merged ET-based calibration scheme over the three raw RS ET-based calibrations, with improvements noted in the Nash-Sutcliffe efficiency (NSE) and Kling-Gupta efficiency (KGE) coefficients for simulated streamflow. On average, the NSE and KGE for the simulated streamflow demonstrate improvements ranging from 0.023 to 0.080 and 0.026 to 0.061, respectively. The multi-variable calibration schemes exhibit a significant enhancement in the reliability of streamflow simulations compared to the ET-based single-variable calibration schemes. On average, the NSE and KGE for the simulated streamflow exhibit improvements of 0.054 to 0.133 and 0.062 to 0.122, respectively. The performance of these calibrations was generally higher in humid catchments but showed more pronounced improvements in arid catchments when compared to raw RS ET-based calibrations. In conclusion, the proposed calibrations using streamflow features and/or merged ET data offer viable solutions for parameter calibration in ungauged catchments. [ABSTRACT FROM AUTHOR]

[Display omitted] • We studied wildfire effects on streamflow, runoff and ET ratios in western Oregon. • Runoff increased the most in catchments with higher burned severity and area burned. • ET decreased 6.1–40.2% and was related to burn severity and area burned. • Slope, elevation, geology, aspect, and pre-fire vegetation influenced streamflow. • Our study provides insights useful to water and forest management. Wildfires can dramatically alter vegetation cover and soil properties across large scales, resulting in substantial shifts in runoff generation, streamflow, and water quality. In September 2020, extensive and high-severity wildfires burned more than 490,000 ha of forest land on the westside of the Cascade Mountain Range in the Pacific Northwest. Much of the area impacted by these fires is critical for the provision of water for downstream aquatic ecosystems, agriculture, hydropower, recreation, and municipal drinking water. We undertook a study to evaluate the effects of four of the large high severity wildfires from 2020 (Riverside, Beachie Creek, Lionshead, and Holiday Farm) on streamflow in nine burned catchments in western Oregon. We also included four unburned, reference catchments in our analysis to enable us to assess post-fire streamflow changes in the burned catchments. To quantify the effects of wildfire on the catchment water balance we used publicly available streamflow data and estimated precipitation, potential evapotranspiration (PET), and actual evapotranspiration (ET), using satellite-based meteorological data. We quantified catchment area burned and burn severity with the average differenced normalized burn ratio (dNBR). We compared hydrologic conditions for the pre-fire (2001–2020) and post-fire (2021–2022) periods by analyzing catchment runoff ratios, ET ratios (evaporative index: quotient of ET divided by precipitation, referred to as EI hereafter), and Budyko curves. We also used random forest models to explore factors influencing the variability in EI. During the post-fire period, we observed decreases in EI and increases in runoff ratio in the burned catchments. Post-fire declines in EI were positively related to burn severity (R 2 = 0.70 in 2021; 0.76 in 2022) and area burned (R 2 = 0.91 in 2021; 0.95 in 2022), and were primarily driven by decreases in ET. Declines in ET were highly variable, ranging from 10.7–40.2 % in the first year after the fires and 6.1–32.0 % in the second year after the fires, and were generally related to catchment burn severity and area burned. The greatest increases in runoff (16.1 % in 2021 and 19.8 % in 2022) occurred in the same catchment. These results were reinforced by the random forest analysis, which illustrated the importance of burn severity as a predictor of EI. Interestingly, the variability in changes in EI during the post-fire period was also associated with other geomorphic factors such as catchment slope, elevation, geology, aspect, and pre-fire vegetation type. Since the duration and seasonality of post-fire impacts on hydrology remain uncertain, our findings bring new insights and guide future studies into the post-fire responses on hydrology that are crucial for water and forest management. [ABSTRACT FROM AUTHOR]

• A direct two-way coupling of hydrologic and 1D-2D hydrodynamic models is proposed. • The refined river links are used to couple hydrologic and 1D hydrodynamic models. • The proposed model allows runoff discharge into 2D inundation areas and 1D rivers. • The model demonstrates superior physical representation of the flood process. A direct two-way coupling of hydrologic and 1D-2D hydrodynamic models (DCM2D) for watershed flood simulation was proposed. This coupling included three models (i.e., fully distributed hydrologic, 1D hydrodynamic, and 2D hydrodynamic models) and three coupling strategies: the bidirectionally coupled hydrologic-2D hydrodynamic module, two-way coupled hydrologic-1D hydrodynamic module (HH1D), and two-way coupled 1D-2D hydrodynamic module (HH2D). The refined river links (RLs) were used to couple the hydrologic and 1D hydrodynamic models, and the exchange discharge between these models was determined using either the weir flow equations or hydrologic model, depending on the flow transitions. Four cases were used to evaluate the proposed DCM2D. The results indicated that the discharge obtained from the proposed HH1D exhibited a high level of consistency with analytical solutions, with NSE of 0.986. This finding supported the validity of the proposed coupling schemes for hydrologic and 1D hydrodynamic models, particularly in scenarios where there were no distinct flow transitions. The DCM2D accurately represented the physical flood process by allowing rainfall runoff discharge into both 2D low-lying inundation regions and 1D rivers. This contrasted with the indirect coupling models (ICM2D), which required runoff to enter the 1D rivers before flowing into the 2D inundation regions. It exhibited reasonable flow directions and achieved satisfactory simulation accuracy, with higher NSE compared to the ICM2D. The proposed DCM2D model can effectively adapt to the real-life flood evolution process and offer practical and reliable solutions for simulating floods in watersheds. [ABSTRACT FROM AUTHOR]

The distribution of flowmeter data and basin characteristic information exhibits substantial disparities, with most flow observations being recorded at a limited number of well-monitored locations. The perennial challenge of achieving reliable and robust hydrological modeling in ungauged catchments through regionalization has persisted. The increasing availability of large-scale hydrological datasets, coupled with recent advancements in machine learning techniques, offers new opportunities to explore patterns of association between basin attributes and hydrological parameters to enhance streamflow predictions. A novel parameter cross-regional transfer approach based on interpretable machine learning (XGBoost) is proposed to accurately predict runoff processes in ungauged regions by leveraging well-trained models across numerous basins within climate zones. We validate the effectiveness of this framework across 5,764 basins in a large sample dataset (Caravan), employing Nash-Sutcliffe Efficiency (NSE), RMSE and bias to assess performance. And a comparison is made with deep transfer learning based on LSTM and Transformer. Results indicate that the proposed method achieves NSE values exceeding 0.2 for 75 % of the ungauged basins, demonstrating superior performance and more stable accuracy compared to pure deep learning models, owing to its incorporation of physical constraints. Furthermore, the response of parameters to basin attributes within different climatic zones in the large-sample context is elucidated through SHAP values, enriching the understanding of hydrological features through data-driven inverse inference. These findings underscore the capability of interpretable machine learning to leverage hydro-physical regularities extracted from abundant basin features, thereby enhancing the accuracy of runoff predictions in ungauged regions. [ABSTRACT FROM AUTHOR]

• LSTM predicts daily SSC well using basin-scale atmospheric forcings, attributes, and observed or modeled streamflow. • Whole-CONUS LSTM can provide reliable predictions for ungauged stations without SSC observations. • Local and Whole-CONUS models are more successful in the humid Eastern US (with high SSC-streamflow correlation, low SSC) • We hypothesize low SSC-streamflow correlation is due to seasonality, basin size, land cover, and rainfall heterogeneity. Suspended sediment concentration (SSC) is a crucial indicator for aquatic ecosystems and reservoir management but is challenging to predict at large scales. It is unclear whether SSC is predictable using macroscopic environmental attributes and forcings. This study tested the feasibility of deep-network-based models to predict daily SSC at basin outlets given only basin-averaged forcings, readily-available physiographic attributes, and streamflow (from observation or model). We trained long short-term memory (LSTM) deep networks both separately for each of the 377 sites across the conterminous United States (CONUS) (termed "local models"), and on all the sites collectively (Whole-CONUS). The Whole-CONUS and local models presented median coefficient of determination (R2) values of 0.63 and 0.52, respectively. This comparison agrees with previously acknowledged "data synergy" effects for LSTM models, where more data from more sites can help improve predictions overall. Furthermore, the continental-scale analysis provided a wealth of insights about SSC patterns. Both local and Whole-CONUS models tended to be more successful where SSC-streamflow correlations (R s - q ) were high − typically in the humid Eastern US − and with lower SSC. Low R s - q basins were often found in the arid Southwest with higher SSC. The highly-nonlinear SSC-streamflow relationships seem related to seasonality, basin size, and heterogeneity of land cover and rainfall within the basins, suggesting these basins need to be simulated at higher spatial resolution and may require additional inputs related to SSC induced by seasonality. The Whole-CONUS model also performed well for spatial extrapolation (basins not included in the training dataset, median R2 = 0.55), supporting large-scale modeling efforts. These state-of-the-art results using only minimal inputs suggest data-driven approaches can exploit the natural coevolution of sediment processes and the environment to support sediment modeling. [ABSTRACT FROM AUTHOR]

• Proposed the novel PROFINN method for daily streamflow imputation. • Integrates probabilistic fusion principles with deep neural networks. • Proposed method maintains precision and allows quantification of uncertainty. • PROFINN performs remarkably well on comparing with RF, K-NN and PMM. • Performs consistently well under different types of data gaps and flow states. The present-day accessibility of streamflow data, particularly in the developing countries, is often marked by a multitude of data shortfalls or distortions. This study investigates the estimate of missing streamflow data using machine learning approaches, including K-nearest neighbour (KNN), Predictive Mean Matching (PMM), Random Forest (RF) and a novel technique of Probabilistic Fusion Imputer with Neural Networks (PROFINN). This study tackles the issue of data insufficiency in such time series considering the inclusion of numerous hydrological parameters by means of RF feature selector, to determine the most significant ones among them. The study explores the efficacy of selected models on various data gaps under different hydrological scenarios, including diverse flow characteristics (mean, high and low flows) and gap lengths (long and short gaps, continuous and discontinuous gaps) and presents a pattern of ranking system that assesses the level of suitability of each technique for various data gaps. This study underscore PROFINN's remarkable performance across all scenarios, yielding an average Root Mean Square Error (RMSE) of 0.91 and an average Nash-Sutcliffe Efficiency (NSE) value of 0.93 when applied for an intermittent river system of Pamba in Southern Kerala, India. RF follows PROFINN in the imputation of extreme flows as well as long and short gap scenarios. KNN closely follows PROFINN for the imputation of continuous and discontinuous gap scenarios. This study augments the significance of tailored machine learning techniques in enhancing the integrity of hydrological datasets, offering valuable insights for effective decision-making in water resource management and related fields. Furthermore, the success of PROFINN in streamflow data imputation suggests its potential extension to other hydrological research domains, like historical data reconstruction assisting climate change studies and future water resources planning, emphasizing the broader relevance and applicability of this study's findings. [ABSTRACT FROM AUTHOR]

• FDC percentiles provide model performances similar to the time-domain results. • Combining signatures may improve model performance and reduce parametric uncertainty. • Considering an error model causes little impact on modeling results. • Signature-domain approach is more relevant to model performance than data origin. Uncertainty analysis and PUB are among the challenges related to hydrological modeling. This work evaluates the applicability of hydrological signatures and Approximate Bayesian Computation methods to estimate the parameters of hydrological models from a regional perspective. Using GR4H model to simulate streamflow in a Brazilian catchment, we established a paradigm solution, under a "classical" Bayesian approach in the time-domain, using an MCMC sampler and a formal likelihood function. Then, we tested different scenarios in the signature-domain using local data, to assess changes in model performance due to the consideration of punctual information. Finally, we used regional data to assess the viability of hydrological signatures to allow prediction in ungauged basins. In this last experiment, we found a worsening in the KGE value of about 33 %, compared to the time-domain results, and 23 %, compared to the local signatures scenario. This shows that the consideration of punctual information has a bigger impact than the data origin. [ABSTRACT FROM AUTHOR]

This study was designed to characterize and quantify the influence of surface soil moisture assimilation on estimated runoff (surface flow and baseflow) and hydraulically-routed streamflow across three large river basins in South Asia that are at risk of impending water stress. Soil Moisture Active Passive (SMAP) surface soil moisture retrievals were assimilated into the Noah-MP land surface model. The gridded runoff was hydraulically routed to obtain volumetric streamflow values for a river network using a runoff routing module (Hydrological Modeling and Analysis Platform, HyMAP). The open loop (OL, model-only) and data assimilated (DA, includes soil moisture retrieval assimilation) Noah-MP runoff estimates highlighted the improvements in estimated total runoff across irrigated areas. Soil moisture assimilation impacted baseflow more relative to surface runoff. The HyMAP-based OL and DA streamflow generally underestimated the streamflow at upstream stations and overestimated the streamflow at downstream stations within the Indus basin due to missing physics related to reservoir operations. For stations located in the Ganges–Brahmaputra basins, the OL and DA estimation performance varied. The OL and DA results showed that the assimilation of soil moisture retrievals improves gridded runoff and volumetric streamflow across irrigated areas. Considerable relative change in streamflow (> 70% increase in magnitude relative to OL) is noted after assimilation across the highly irrigated lower Indus basin and high precipitation regions in Bangladesh. Improving the modeling system's representativeness of ground conditions via inclusion of water management information could improve large-scale streamflow modeling across South Asia. • SMAP soil moisture assimilation generally leads to an increase in total runoff and volumetric streamflow across irrigated areas. • Relative to surface runoff, baseflow is generally more impacted by SMAP soil moisture assimilation. • Relative change in assimilated streamflow is considerable across irrigated areas but limited along river tributaries with high streamflow. [ABSTRACT FROM AUTHOR]

• Use of multiple calibration parameter sets and RCMs helps in capturing upper and lower bounds of future predictions. • Hydrological model uncertainty is most important when climate model agreement is good. • Use of a single hydrologic model calibration, even with multiple RCMs, could lead to under-design. Assessing the uncertainty associated with projections of climate change impacts on hydrological processes can be challenging due to multiple sources of uncertainties within and between climate and hydrological models. Here we compare the effects of parameter uncertainty in a hydrological model to inter-model spread from climate projections on hydrological projections of urban streamflow in response to climate change. Four hourly climate model outputs from the RCP8.5 scenario were used as inputs to a distributed hydrologic model (SWMM) calibrated using a Bayesian approach to summarize uncertainty intervals for both model parameters and streamflow predictions. Continuous simulation of 100 years of streamflow generated 90 % prediction intervals for selected exceedance probabilities and flood frequencies prediction intervals from single climate models were compared to the inter climate model spread resulting from a single calibration of the SWMM model. There will be an increase in future flows with exceedance probabilities of 0.5 %-50 % and 2-year floods for all climate projections and all 21st century periods, for the modeled Ohio (USA) watershed. Floods with return periods of ≥ 5 years increase relative to the historical from mid-century (2046–2070) for most climate projections and parameter sets. Across the four climate models, the 90th percentile increase in flows and floods ranges from 17-108 % and 11–63 % respectively. Using multiple calibration parameter sets and climate projections helped capture the most likely hydrologic outcomes, as well as upper and lower bounds of future predictions. For this watershed, hydrological model parameter uncertainty was large relative to inter climate model spread, for near term moderate to high flows and for many flood frequencies. The uncertainty quantification and comparison approach developed here may be helpful in decision-making and design of engineering infrastructure in urban watersheds. [ABSTRACT FROM AUTHOR]

[Display omitted] • A method to deal with the spatial scale of the basin is proposed. • The runoff generation process is integrated into the modeling process. • The model realizes the combination of hydrological concepts and deep learning methods. • The proposed method shows its excellent performance in streamflow prediction. Accurate prediction of runoff is critical concerning reservoir management and disaster preparedness. Data-driven methods are progressively applied to runoff prediction tasks and have led to impressive results. However, existing data-driven methods are hardly considered to the runoff generation process and the spatial characteristics of basins in the models due to the lack of a priori knowledge guidance. Here a structured approach is provided to develop the perceptual model for runoff generation and model the behavior in groups at different locations and scales; considering the hierarchical structure of basin systems, a short-term runoff forecasting model with spatial perception and scale interaction, i.e., the hierarchical attention network, is developed based on the encoder-decoder structure and attention mechanism. Compared to the single- and multi-step prediction performance of the six baseline models, the NSE improved by an average of 2.41, 9.68, and 12.14%, respectively. This implies that incorporating basin-related knowledge in modeling and considering runoff generation processes and spatial connectivity can improve prediction accuracy, and the necessity of considering conceptual mechanisms in data-driven models. [ABSTRACT FROM AUTHOR]

• We develop a check dam module that is coupled into SWAT-PML to disentangle check dam impacts from vegetation changes. • With the continuous construction of check dams, there is an increasing trend in the impact of dams on streamflow. • Check dams under maximum operation have largest decrease in streamflow, followed by vegetation. • Normal operation of check dams has marginal impacts on decrease in streamflow. Check dam is an important soil and water conservation engineering measure to prevent and control the soil erosion in the Loess Plateau. However, how check dams influence catchment streamflow process remains largely unclear. This study develops a check dam operation module within SWAT model, which is then coupled with Penmen-Monteith-Leuning (PML)-V2 evapotranspiration model. The coupled model is finally used to disentangle streamflow impacts of check dam from vegetation changes by considering four scenarios: vegetation greening and normal operation of check dam, vegetation greening and maximum operation of check dam, vegetation greening only, and no-impact. Our results indicated a progressive increase in water reduction, with the gradual construction of check dams over the years. The vegetation greening plays a major role in reducing catchment streamflow by 9.9 % to 19.8 % in three studied catchments. The check dam under normal operation shows marginal impacted on streamflow (the three studied catchments experienced streamflow reduction ranging from 2.9 % to 4.9 %), though the maximum check dam operation had the largest impact (the proportion of streamflow reduction ranges from 28.2 % to 54.2 %). This study can help understanding the impacts of different check dam constructions and operations on streamflow, which is helpful for planning sustainable water resources. [ABSTRACT FROM AUTHOR]

• Framework to reveal effect of hydrometeorological regime on drought propagation. • Meteorological–hydrological drought propagation is multidimensional/non-linear. • Accounting for regime shifts increases propagation thresholds by 2.81 %–16.90 %. Exploring drought propagation relationships and thresholds is crucial for drought control and mitigation. However, previous studies have overlooked the impacts of regime shifts in trends of hydrometeorological variables over time on drought propagation and its thresholds. This study proposes an integrated framework focusing on drought propagation from meteorological to hydrological drought. The regime shift periods, both increasing and decreasing trends, were identified using the cumulative difference curve-rank test on precipitation and streamflow at an annual time scale. The Standardized Precipitation Index (SPI) and Standardized Streamflow Index (SSI) were used as meteorological and hydrological drought indices, respectively. Additionally, the propagation relationships and thresholds from meteorological drought to hydrological drought were identified and compared across various regime shift periods. The integrated framework's applicability was verified by applying it to the Three-Rivers Headwaters region of China, which encompasses the Yellow River, Yangtze River, and Lancang River. This region is strategically important for China and has relatively little human activity. The framework revealed the variability in drought propagation and thresholds during different regime shift periods. The response sensitivity of SSI to SPI and the probabilities and thresholds for propagation from meteorological to hydrological drought differed between periods of increasing and decreasing trends in hydrometeorological variables. When regime shift characteristics were considered, the thresholds for meteorological drought propagating to hydrological drought were 2.81 %–16.90 % higher. These results indicated that, when examining drought propagation relationships and thresholds, accounting for regime shifts in hydrometeorological variables can improve early drought warning and enhance drought control and mitigation efforts. [ABSTRACT FROM AUTHOR]

• A novel spatiotemporal DL approach is developed to integrate Euclidean and non-Euclidean spatial information. • The responses of the spatiotemporal DL model to different precipitation inputs and graph representations are investigated. • The proposed model outperforms comparative models in streamflow simulations temporally, spatially and spatiotemporally. • The simulations of tributary streamflow gauges are particularly sensitive to model refinements and spatial input selection. Spatiotemporal deep learning (DL) has emerged as a promising paradigm for hydrological simulation compared with lumped models using basin-averaged inputs. However, existing research primarily focuses on either Euclidean data, characterized by regular structures such as grid-like meteorological forcing, or non-Euclidean data, which features irregular topological connectivity such as downstream information transfer. Here, a novel spatiotemporal DL-based approach is developed to integrate both types of spatial information for streamflow modeling. Our proposed model utilizes convolutional neural network (CNN)- and graph neural network (GNN)-based modules to extract the spatial features of Euclidean and non-Euclidean data, respectively. Spatial-attention mechanisms are incorporated into these modules to automatically focus on important features. Long Short-Term Memory (LSTM) networks are then employed to capture temporal dependencies. The adaptability and response of the spatiotemporal DL-based model to different Euclidean and non-Euclidean inputs are investigated. The proposed model is applied to simulate multi-step and multi-gauge streamflow in the mid-lower Hanjiang River basin (ML-HRB), one of the largest sub-basins of the Yangtze River. Results show that when integrating Euclidean and non-Euclidean spatial data, the proposed model outperforms comparative models temporally, spatially and spatiotemporally. The average model improvements across all steps and gauges are of 17.75 % for the Nash-Sutcliffe Efficiency coefficient (NSE), 9.02 % for the correlation coefficient (CC), and 14.79 % for the Kling–Gupta efficiency metric (KGE), compared with the commonly-used lumped LSTM model. The simulations of the tributary stream gauges are particularly sensitive to model refinements and spatial input selection. This study sheds light on the potential of integrating different kinds of spatial information for spatiotemporal DL-based simulation. [ABSTRACT FROM AUTHOR]

• Hyporheic flow under rough ice cover conditions was addressed for the first time through numerical models. • Compared to open channel flow, the rough underside of ice cover increases hyporheic flux through shallower zones. • Parsimonious models are provided for predicting ice cover roughness-affected hyporheic flow. Many rivers are subject to ice-covered flow conditions in cold climate regions or during the winter months. The presence of ice cover can change stream flow hydrodynamics and thus affect the hyporheic exchange involved in many biogeochemical processes in aquatic environments. Among the various features of ice cover, its underside roughness is a main factor modifying the surface–subsurface flow hydrodynamics. To date, there has been little research on how ice cover affects the hyporheic exchange, even less on the effect of the ice cover roughness. This leaves open questions on the mechanical changes initiated at the ice-water interface and transferred to the water–sediment interface. In this study, a coupled surface–subsurface computational approach was used to model conjunctively the channel flow over dune bed and the underlying porous flow under different ice cover conditions. The results show that ice cover can enhance the hyporheic flux and reduce the hyporheic zone depth when compared to open channel flow at equivalent discharge. As the river flow depth increases to 4 times the height of the dune, the effect of the ice cover on the hyporheic flow becomes negligible. Within this effective range, the hyporheic flux via the channel water depth follows a power-law function. Ice cover roughness greatly augments the exchange rate while compresses the exchange space, therefore reduces flow residence time in the hyporheic zone. Based on the modeling results, prediction models were proposed for evaluating the impacts of ice cover roughness on the hyporheic exchange flux and depth. Our study indicates that hyporheic flow exchange is significantly impacted by the roughness of the ice cover which in turn the hyporheic exchange is likely to affect the river ice processes as well as the water quality and ecological functions throughout the river corridors. [ABSTRACT FROM AUTHOR]

• First time Transformer achieved the same performance as LSTM on CAMELS dataset. • Unmodified vanilla Transformer could not reach LSTM's performance. • A non-recurrent connection was added to the Transformer, supporting parallelism. • Transformers may have scale advantages for larger datasets. • LSTMs and Transformers are likely nearing the prediction limits of the dataset. For a number of years since their introduction to hydrology, recurrent neural networks like long short-term memory (LSTM) networks have proven remarkably difficult to surpass in terms of daily hydrograph metrics on community-shared benchmarks. Outside of hydrology, Transformers have now become the model of choice for sequential prediction tasks, making it a curious architecture to investigate for application to hydrology. Here, we first show that a vanilla (basic) Transformer architecture is not competitive against LSTM on the widely benchmarked CAMELS streamflow dataset, and lagged especially prominently for the high-flow metrics, perhaps due to the lack of memory mechanisms. However, a recurrence-free variant of the Transformer model obtained mixed comparisons with LSTM, producing very slightly higher Kling-Gupta efficiency coefficients (KGE), along with other metrics. The lack of advantages for the vanilla Transformer network is linked to the nature of hydrologic processes. Additionally, similar to LSTM, the Transformer can also merge multiple meteorological forcing datasets to improve model performance. Therefore, the modified Transformer represents a rare competitive architecture to LSTM in rigorous benchmarks. Valuable lessons were learned: (1) the basic Transformer architecture is not suitable for hydrologic modeling; (2) the recurrence-free modification is beneficial, so future work should continue to test such modifications; and (3) the performance of state-of-the-art models may be close to the prediction limits of the dataset. As a non-recurrent model, the Transformer may bear scale advantages for learning from bigger datasets and storing knowledge. This work lays the groundwork for future explorations into pretraining models, serving as a foundational benchmark that underscores the potential benefits in hydrology. [ABSTRACT FROM AUTHOR]

• An integrated water system model with a reduced-complexity groundwater module. • The model performs well in simulating river and groundwater temporal dynamics. • A novel flux tracking approach to reveal groundwater and river baseflow origins. • Assessment of aquifer bodies' contributions to baseflow and groundwater abstraction. • Strategies informed by flux tracking results improve baseflow more efficiently. In systems where surface water and groundwater interact, management of the water resource often involves conflicting objectives between water supply and baseflow maintenance. Balancing such objectives requires understanding of the role of groundwater in integrated water systems to inform the design of an efficient strategy to minimise abstraction impacts. This study first develops a reduced-complexity, processed-based groundwater model within the water systems integration modelling framework (WSIMOD). This model is applied to the Lea catchment, UK, as a case study and evaluated against monitored groundwater level and river flow data. A flux tracking approach is developed to reveal the origins of both river baseflow at a critical assessment point and abstracted groundwater across the systems. The insights obtained are used to design two strategies for groundwater abstraction reduction. Results show that the model has good performance in simulating the groundwater and river flow dynamics. Three aquifer bodies that contribute the most to the river baseflow in the dry season at the assessment point are identified; contributions being 17 %, 15 %, and 5 %. The spatial distribution of abstracted groundwater originating from these aquifer bodies is illustrated. Compared to the default equal-ratio reduction, the strategy prioritising abstraction reduction in these three aquifer bodies increases a similar amount of baseflow (13 %) by reducing much less abstraction (23 %). The other strategy, which further decreases abstraction in the adjacent aquifer bodies, increases more baseflow (16 %) with a similar abstraction reduction (30 %). Both strategies can more efficiently improve the baseflow. The flux tracking approach can be further implemented to trace water from other origins, including runoff, stormwater, and wastewater, to enable coordinated management for better systems-level performance. [ABSTRACT FROM AUTHOR]

• Performance of calibration-free Dynamic Budyko (DB) model is compared with regionalized HBV (HBV reg) model. • DB performed as good as HBV reg for daily streamflow prediction in 3534 study basins. • Climate exerts first-order control over streamflow dynamics. • Catchment similarity needs to be focused on for developing meaningful hydrological models. Hydrological processes leading to flow in river channels are extremely complex, which is why we cannot use any fundamental law to fully explain river flow dynamics at basin scale. The general approach for hydrological modelling considers that every catchment's response to precipitation input is unique, which means different catchments respond differently to the same precipitation input. In other words, river basins are sufficiently dissimilar to warrant having hydrological models with multiple free-parameters for predicting discharge. It is assumed that free-parameters are related to static catchment characteristics. However, past efforts to relate model parameters with observable catchment characteristics have not been very successful. We thus need a framework to quantify catchment dissimilarities exhibited by catchment characteristics because there is no need of free-parameters if we decide to ignore dissimilarities. Thus, the relative roles of climate and catchment characteristics in the streamflow response of the basin needs to be quantified. In this study, we have considered daily data from 3534 gauging stations spread across the globe to compare HBV, a widely used hydrological model, with Dynamic Budyko (DB) model, a recently proposed calibration-free model. The HBV model assumes that both climate and catchment characteristics are important in modeling streamflow response, whereas a calibration-free DB model essentially assumes that streamflow response is largely governed by climate. The performance of HBV run with regionalized parameters and DB in terms of median coefficient of determination (R 2 ) and weighted R 2 (WR 2 ) are (0.43, 0.45) and (0.24, 0.27), respectively. The corresponding values of R 2 and WR 2 from quantile-quantile comparison are (0.94, 0.94) and (0.60, 0.59). The results here seem to support the idea of developing calibration-free models by appreciating similarities between basins, particularly if the objectives are prediction in ungauged basins and assessment of climate change impact on streamflow. [ABSTRACT FROM AUTHOR]

• A novel Fourier operator-based DL model is built to model river flow hydrodynamics. • New model architecture is used to model unsteady, complex, realistic river flow. • FNOCL efficiently addresses uncertainties in inflow and cross-sectional profiles. • FNOCL drastically reduces the computational time by nearly two orders of magnitude. Effective river management relies heavily on the accurate simulation of river flow dynamics to develop scenario-based strategies and inform decision making. Deep learning (DL) models, as alternatives to conventional simulation techniques, offer a compelling blend of precision and efficiency. This study introduces FNOCL (Fourier Neural Operator with ConvLSTM units), a novel DL neural network for modelling unsteady natural river flows. The network is adept at handling the complexities of spatial and temporal variations. FNOCL builds upon the Fourier neural operator (FNO) by incorporating convolutional long short-term memory (ConvLSTM) layers, combining robust performance with high proficiency in terms of capturing spatiotemporal patterns. Notably, FNOCL accommodates complex river cross-section profiles, thus enabling the accurate modelling of flow dynamics within intricate river morphologies. Our study applies FNOCL to two distinct river networks with varying morphology complexity levels. With reference to the HEC-RAS model, FNOCL showcases its ability to attain high accuracy in model prediction tasks and superior generalization with regard to addressing the uncertainty associated with inflow variations and cross-sectional changes. In comparative analyses with benchmark models (i.e., FNO and ConvLSTM), FNOCL performs most efficiently during the training process and obtains the most accurate water level and river flow predictions in the spatiotemporal domain. Furthermore, FNOCL exhibits a substantial speedup over HEC-RAS, notably reaching up to 50 times in a complex river network. These results highlight FNOCL as an efficient surrogate for conventional simulators, facilitating precise analyses across various scenarios. It provides efficient computations for system responses within complex river flow networks, thereby enhancing applications in river management, such as flood risk assessment, flood control, and the planning and design of hydraulic control structures. [ABSTRACT FROM AUTHOR]

• Streamflow prediction in 2,610 basins worldwide using Temporal Fusion Transformers (TFTs). • TFTs outperform LSTMs and Transformers for streamflow prediction on single basins. • Combining recurrence with attention has significant value for time series modeling of streamflow. Over the past few decades, the hydrology community has witnessed notable advancements in streamflow prediction, particularly with the introduction of cutting-edge machine-learning algorithms. Recurrent neural networks, especially Long Short-Term Memory (LSTM) networks, have become popular due to their capacity to create precise forecasts and realistically mimic the system dynamics. Attention-based models, such as Transformers, can learn from the entire data sequence concurrently, a feature that LSTM does not have. This work tests the hypothesis that combining recurrence with attention can improve streamflow prediction. We set up the Temporal Fusion Transformer (TFT) architecture, a model that combines both of these aspects and has never been applied in hydrology before. We compare the performance of LSTM, Transformers, and TFT over 2,610 globally distributed catchments from the recently available Caravan dataset. Our results demonstrate that TFT indeed exceeds the performance benchmark set by the LSTM and Transformers for streamflow prediction, when models are trained individually for each basin. Additionally, being an explainable AI method, TFT helps in gaining insights into the streamflow generation processes. [ABSTRACT FROM AUTHOR]

• Variable riverbed topography greatly modulates the relationship of hyporheic and regional groundwater flow process. • The exist of hyporheic zone produced by variable topography affect the exchange flux between surface water and wider regional groundwater. • The variation of topography wavelength has more significant effect on exchange flow paths and travel times compared with amplitude. Hyporheic flow, which occurs in the shallow zone where river flow interacts with groundwater (GW) flow, involves spatiotemporally complex water mixing and storage processes of high ecological importance. The role of naturally complex riverbed morphology on the interaction of hyporheic flow in meandering stream and groundwater patterns is still poorly understood. In this study, groundwater conceptual and numerical models were established using field datasets from investigations completed along the Yellow River in Henan province (China), which is characterized by heavy sediment load and varied riverbed topography transversally to the river channel. The change of transversal riverbed topography was conceptually simplified and described as single sinusoidal with varying amplitude and wavelength. Numerical results showed that increasing topography amplitude and wavelength produced a thicker and wider hyporheic zones (HZs), where surface water mixes with groundwater, under both gaining and losing conditions. Increasing topography wavelength led to significant decrease of both the total water flux (Q TF) across the riverbed sediment interface and the hyporheic flux (Q HF, the flux of these water that flows into the subsurface and returns to the stream after spending some time in the bed) under both gaining and losing conditions. The total exchange flux between groundwater and surface water (Q GW-sw) increased with the increase of topography wavelength under gaining conditions, but decreased under losing conditions. These trends for Q TF , Q HF and Q GW-sw could be reasonably well fitted with power law relationships. Furthermore, the increased wavelength also increased groundwater exchange travel times between GW and surface water (SW) under losing conditions, but decreased under gaining conditions. The change of topography wavelength also compress/extend the GW flow paths under gaining and losing conditions, respectively. These findings demonstrate the major control of river morphology on spatial patterns and magnitude of hyporheic exchanges, and further highlight the importance of considering riverbed topography not only when evaluating hyporheic exchange but also the interactions between GW and SW. [ABSTRACT FROM AUTHOR]

• Inadequate hydrologic data restricts sound water resource management decisions. • Distributed models can improve estimates of natural flow regimes. • Results quantified the impact of environmental flow decisions on hydropower. • Equitable management of resources depends on confidence in management consequences. Estimating streamflow regimes on small tropical islands is challenging due to limited observational data, poor accuracy of remote sensing, and the limitations of coarse-scale climate models, restricting our understanding of hydrological processes and the management of water resources. The inadequacy of data has also affected the development of instream flow standards (i.e., environmental flows, e-flows) needed to protect ecological and cultural values central to island communities. Estimates of water availability for agriculture, hydropower, and drinking water supply are also critical for informed investments in water resource planning. We utilized high-resolution (250 m) gridded daily rainfall data to parameterize the Soil and Water Assessment Tool (SWAT) to characterize streamflow regimes at ungauged locations in a watershed on Kaua'i Island. The resultant flow statistics were then used to describe the daily, seasonal, and annual availability of surface water to evaluate the implications of potential e-flow scenarios on water available for run-of-river hydropower. Compared to rainfall data derived from a single station within the watershed, the gridded daily rainfall product increased the accuracy of model output from not satisfactory to good. Model techniques using high resolution rainfall data may substitute for long-term hydrological data collection across complex tropical environments and provide data for making sustainable management decisions at ungauged locations. [ABSTRACT FROM AUTHOR]

• Nationwide streamflow projections for Australia informed by CMIP5 and CMIP6. • Vast majority of CMIP5 and CMIP6 GCMs project less winter rainfall across Australia. • Lower winter rainfall translates to a significant reduction in annual runoff. • Modelled median projection of runoff shows a 50% reduction in far southwest Australia. • Modelled median projection of runoff shows a 20% reduction in far southeast Australia. Projections of streamflow under future climate are essential for developing adaptation strategies in the water and related sectors. This paper presents nationwide streamflow projections for Australia informed by climate change signals in CMIP6 GCMs and compares the projections with those from CMIP5 GCMs. The modelled future runoff projections driven by CMIP5 and CMIP6 GCMs are relatively similar for far southern Australia. The median projection is a 50% reduction in mean annual runoff in far southwest Australia and 20% reduction in far southeast Australia by 2046–2075 relative to 1976–2015 under the high SSP5-8.5/RCP8.5 global warming scenario. The vast majority of CMIP5 and CMIP6 GCMs project less winter rainfall across Australia. As most of the runoff in far southern Australia occurs in winter and spring, the lower winter rainfall translates to a significant reduction in annual runoff. It is therefore prudent to plan and manage for a reduction in future water resources, particularly in the densely populated and important agricultural regions in southeast Australia. There is less agreement between GCMs in the summer rainfall projection, with the CMIP6 GCMs generally projecting a wetter future (or smaller decrease in rainfall) than the CMIP5 GCMs. The modelled median projection for mean annual runoff is a 10% reduction in the south-east coast, 5% reduction in the north-east coast and little change in northern Australia. More GCMs project an increase rather than a decrease in the interannual variability of rainfall, and this would further amplify multi-year hydrological droughts. [ABSTRACT FROM AUTHOR]

[Display omitted] • MVMD explores coupling relationships between streamflow and associated variables. • The Transformer establishes connections between pairs of nodes in each mode. • MVMD-Transformer outperforms benchmark models in streamflow forecasting for three major inland rivers in China. Accurate streamflow forecasting is critical in the domain of water resources management. However, the inherently non-stationary and stochastic nature of streamflow poses a significant challenge to achieving accuracy in streamflow forecasting. In this study, we introduce an MVMD-ensembled Transformer model (MVMD-Transformer), which incorporates the MVMD for concurrent time–frequency analysis of streamflow and related potential influencing variables. The model aligns common modes in the decomposition results, ensuring that the different variables corresponding to each mode have the same center frequency. This alignment overcomes frequency mismatches and helps uncover the intrinsic patterns and essential features between streamflow and associated variables. During the forecasting phase, the Transformer component of the MVMD-Transformer model establishes connections among streamflow and other influencing variables across pairs of nodes in each mode. We tested the performance of the MVMD-Transformer model in forecasting streamflow across 1-, 3-, 5-, and 7-day horizons within the Shiyang River, Heihe River, and Shule River basins situated in the Hexi Corridor of Northwest China. The MVMD-Transformer model harnesses MVMD to concurrently decompose both predictor variables (precipitation, air temperature, air pressure, soil moisture) and the response variable (streamflow). Subsequently, the modes drived from the MVMD were fed into the Transformer, serving as the predictive analytics engine, to forecast streamflow. Furthermore, we conducted a comprehensive performance evaluation by comparing the MVMD-Transformer model against four alternatives: the VMD-ensembled Transformer model (VMD-Transformer), CEEMDAN-ensembled Transformer model (CEEMDAN-Transformer), stand-alone Transformer model, and LSTM model. The results indicate that MVMD-Transformer outperformed all other models, achieving Nash-Sutcliffe coefficient (NSE) values exceeding 0.85 in the majority of the forecasting scenarios. This superior performance highlights the proficiency of the MVMD approach in more accurately unraveling the intricate interdependencies between streamflow and its various potential influencing variables, thus significantly improving the precision of streamflow forecasting. [ABSTRACT FROM AUTHOR]

• EFS-EMA was proposed to optimize the input variables and hyperparameters parallelly. • ESVR and LSTM were incorporated with EFS-EMA to forecast the monthly streamflow. • Ten study areas from the CAMELS dataset were used to verify the model performance. • EFS-EMA helped ESVR and LSTM get overall performance improvement. • The proposed EFS-ESVREMA model was relatively superior to other tested models. Accurate forecast of monthly streamflow is helpful to improve the social capability in risk management. Diverse input features are crucial to the accuracy of machine learning-based monthly streamflow forecast models. The embedded feature selection (EFS) method based on binary-coded metaheuristic algorithm is adept at parallel optimization of feature selection and model hyperparameter. It has been rarely studied coupled with deep learning models in monthly streamflow forecasting. This study further investigated the potential of EFS method using the well-known Long short-term memory network (LSTM) and a state-of-the-art model named Extreme support vector regression (ESVR), with the Evolutionary mating algorithm (EMA) as the search algorithm. The benchmark models were established using EMA for hyperparameter optimization without feature selection. Ten basins with distinct hydroclimate characteristics minimally impacted by human activities from the Catchment Attributes and Meteorology for Large-sample Studies (CAMELS) data set were used to comprehensively examine the proposed models. Nonparametric approaches including the Friedman test and Wilcoxon signed ranks test were utilized to rank the models and detect the significance of the performance variation brought by EFS. The kernel ESVR model coupled with EFS-EMA (EFS-KESVREMA) ranked the first in the 1-month-ahead experiment with NSE = 0.734–0.944. EFS-ESVREMA and EFS-LSTMEMA ranked second and third, with NSE = 0.746–0.94 and NSE = 0.701–0.94 respectively. The results also demonstrated that EFS-EMA helps the benchmark models improve prediction accuracy at a confidence level of 95 % but it has no obvious effect on the standard SVR. As the forecast horizon increased to 3- and 5-month-ahead, EFS-ESVREMA maintained a reliable performance. Whereas EFS-LSTMEMA experienced a significant accuracy decrease in the 5-month-ahead experiment. The proposed models and relevant findings in this study have certain reference value for researchers in monthly streamflow forecasting. [ABSTRACT FROM AUTHOR]

• Improved the accuracy of GFS meteorological forecasts after bias-correction. • Enhanced bi-directional LSTM through Wavelet analysis. • Coupled six hierarchical data driven models with SWAT-Pothole. • Assessed PSWAT-WBiLSTM hybrid model as the best for 1–10 days' flood forecasting. The development of a streamflow forecasting tool becomes a challenging task due to the sophisticated nonlinear catchment response, varying crop management practices, and limited in situ data availability. For real-time streamflow forecasting with up to 10-days lead-time, this study investigates the potential of SWAT-pothole (PSWAT) module forced with bias-corrected Global Forecasting System (GFS) meteorological variables; wherein the error-updating is carried out for the streamflow forecasts simulated by PSWAT. The error is updated by hierarchical data-driven sub-models, such as AutoRegressive (AR), AutoRegressive Moving Average with eXogenous inputs (ARMAX), Wavelet-based Neural Network (WNN), Wavelet-based Non-linear AutoRegressive with eXogenous inputs (WNARX), Long-Short Term Memory (LSTM), and a novel Wavelet-based Bidirectional LSTM (WBiLSTM). The efficacy of the standalone PSWAT module is tested against the hybrid models in the Brahmani-Baitarani (≈ 49,000 km2) compound River Basin in eastern India. The results revealed that the novel PSWAT-WBiLSTM hybrid model is the best for reliable streamflow forecasts up to 8-days' and 9-days' lead-times in the Brahmani and Baitarani River basins, respectively. Conclusively, this model can be a potential medium-range streamflow forecasting tool for paddy-dominated catchments. [ABSTRACT FROM AUTHOR]