Monthly Streamflow Research Articles

Catchment response to climatic variability: implications for root zone storage and streamflow predictions

Abstract. This paper investigates the influence of multi-decadal climatic variability on the temporal evolution of root zone storage capacities (Sr,max) and its implications for streamflow predictions in the Meuse basin. Through a comprehensive analysis of 286 catchments across Europe and the US that are hydro-climatically comparable to the Meuse basin, we construct inter-decadal distributions of past deviations in evaporative ratios (IE) from expected values based on catchment aridity (IA). These distributions of ΔIE were then used to estimate inter-decadal changes in Sr,max and to quantify the associated consequences for streamflow predictions in the Meuse basin. Our findings reveal that, while catchments do not strictly adhere to their specific parametric Budyko curves over time, the deviations in IE are generally very minor, with an average ΔIE=0.01 and an interquartile range (IQR) of −0.01 to 0.03. Consequently, these minor deviations lead to limited inter-decadal changes in Sr,max, mostly ranging between −10 and +21 mm (−5 % to +10 %). When these changes (ΔSr,max) are accounted for in hydrological models, the impact on streamflow predictions in the Meuse basin is found to be marginal, with the most significant shifts in monthly evaporation and streamflow not exceeding 4 % and 12 %, respectively. Our study underscores the utility of parametric Budyko-style equations for first-order estimates of future Sr,max in hydrological models, even in the face of climate change and variability. This research contributes to a more nuanced understanding of hydrological responses to changing climatic conditions and offers valuable insights for future climate impact studies in hydrology.

Application of coupled MIKE SHE/MIKE HYDRO on streamflow dynamics under climate change in Lake Tana sub basin, Abay Basin, Ethiopia

AbstractWater availability and quality are fluctuating due to climate change, which has disastrous effects on life. Modeling climate change impact on streamflow in the Lake Tana sub‐basin (LTSB) in selected watersheds was the main goal of the research. This research is unique in that it applies the coupled MIKE SHE/MIKE HYDRO model technique, which has not been applied before to investigate changes in streamflow due to climate change in LTSB. Streamflow in the LTSB was forecasted using the MIKE SHE/MIKE HYDRO model from seven ensembles of GCMs under two emission scenarios (SSP2–4.5 and SSP5–8.5) between 2041 and 2,100. According to the calibration and validation results in the period of 1985–2007, MIKE SHE performs well while simulating streamflow in LTSB. During calibration and validation, the Nash‐Sutcliffe efficiency (NSE) values in all watersheds were greater than 0.8, except for the Ribb watershed calibration, where it was 0.75. The mean monthly changes in temperatures indicate an incremental tendency in the next decades. According to the simulation result, the monthly mean rainfall for SSP2–4.5 and SSP5–8.5 will increase from 0.05 to 61.9% and 2.02 to 46.26% in the 2050s and 8.7 to 44.38% to 6.44 to 66.24% in the 2080s, respectively. Gilgle Abay, Gumara, and Ribb watersheds will lose 56.2, 56.65, 57.01% of their mean monthly streamflow in the 2050s, according to SSP2–4.5. For the Gilgle Abay, Gumara, and Ribb watersheds, streamflow will increase by 6.7% to 21.94%, 6.08% to 23.26%, and 6.51% to 22.88%, accordingly, in the 2050s under SSP5–8.5. Increased rainfall in the watershed could result in floods in the already‐flooded Gumara and Ribb flood plains. The government and community should implement coping mechanisms to lessen the impact of climate change like early warring and planting more trees. Overall, the study showed that it is expected that conditions related to streamflow may change in the future due to climate alteration. By offering a better understanding of current and future climates, the integrated modeling system developed can assist environmentalists, hydrologists, and policymakers in water management and policy intervention.

Coupling SWAT and Transformer Models for Enhanced Monthly Streamflow Prediction

The establishment of an accurate and reliable predictive model is essential for water resources planning and management. Standalone models, such as physics-based hydrological models or data-driven hydrological models, have their specific applications, strengths, and limitations. In this study, a hybrid model (namely SWAT-Transformer) was developed by coupling the physics-based Soil and Water Assessment Tool (SWAT) with the data-driven Transformer to enhance monthly streamflow prediction accuracy. SWAT is first constructed and calibrated, and then its outputs are used as part of the inputs to Transformer. By correcting the prediction errors of SWAT using Transformer, the two models are effectively coupled. Monthly runoff data at Yan’an and Ganguyi stations on Yan River, a first-order tributary of the Yellow River Basin, were used to evaluate the proposed model’s performance. The results indicated that SWAT performed well in predicting high flows but poorly in low flows. In contrast, Transformer was able to capture low-flow period information more accurately and outperformed SWAT overall. SWAT-Transformer could correct the errors of SWAT predictions and overcome the limitations of a single model. By integrating SWAT’s detailed physical process portrayal with Transformer’s powerful time-series analysis, the coupled model significantly improved streamflow prediction accuracy. The proposed models offer more accurate and reliable predictions for optimal water resource management, which is crucial for sustainable economic and societal development.

Quantifying the Role of Calibration Strategies on Surface‐Subsurface Hydrologic Model Performance

ABSTRACTDistributed, coupled surface‐groundwater hydrologic models are high‐dimensional, given the necessity to reflect the spatially diverse nature of complex hydrologic processes. Furthermore, inverse/inference problems involving these high‐dimensional models are naturally ill‐posed, given the limited information content of state observations that are typically assimilated. Many inversion/inference algorithms do not cope well with high dimensionality, leaving the practitioner to make subjective choices related to uncertain model inputs. The objective of this study is to evaluate the impact of these subjective calibration choices within a formal sensitivity analysis, uncertainty analysis, and parameters estimation (SA‐UA‐PE) framework on model testing for a surface‐subsurface hydrologic model. In doing so, we address the concepts of ‘over‐parameterisation’ and ‘under‐parameterisation’. We completed a series of numerical experiments, testing several otherwise subjective aspects of the calibration process: (1) the number (5, 10, 15, 20) and type (soil, aquifer, land surface, channel) of calibration parameters selected); (2) the type of state observations assimilated (streamflow, groundwater head); and (3) the length of testing period (1 to 14 years), using monthly streamflow and groundwater head as testing data. The experiments were completed for models of the Winnebago River watershed (Minnesota, Iowa), (significant tile drainage) and the Nanticoke River watershed (Delaware, Maryland (significant groundwater‐channel interactions). The selected hydrologic model is SWAT+, using the gwflow module for physically based groundwater storage and flow modelling, and simulations are run for the 2000–2015 period. Through this process, we found that increasing the number of parameters from 5 to 15 improves the representation of streamflow, principally through an improvement of groundwater storage representation and baseflow generation, but minimal improvement when increasing to 20 parameters. Therefore, the SA‐UA‐PE process can be optimised based on an ideal number of parameters that yield adequate results while maintaining a lower computational burden. The method presented here can be used for any watershed, using integrated surface‐subsurface hydrologic models.

Future Projection of Water Resources of Ruzizi River Basin: What Are the Challenges for Management Strategy?

This study investigates the impact of climate change on hydrological dynamics in the Ruzizi River Basin (RRB) by leveraging a combination of observational historical data and downscaled climate model outputs. The primary objective is to evaluate changes in precipitation, temperature, and water balance components under different climate scenarios. We employed a multi-modal ensemble (MME) approach to enhance the accuracy of climate projections, integrating historical climate data spanning from 1950 to 2014 with downscaled projections for the SSP2-4.5 and SSP5-8.5 scenarios, covering future periods from 2040 to 2100. Our methodology involved calibrating and validating the SWAT model against observed hydrological data to ensure reliable simulations of future climate scenarios. The model’s performance was assessed using metrics such as R2, NSE, KGE, and PBIAS, which closely aligned with recommended standards. Results reveal a significant decline in mean annual precipitation, with reductions of up to 37.86% by mid-century under the SSP5-8.5 scenario. This decline is projected to lead to substantial reductions in surface runoff, evapotranspiration, and water yield, alongside a marked decrease in mean monthly stream flow, critically impacting agricultural, domestic, and ecological water needs. The study underscores the necessity of adaptive water resource management strategies to address these anticipated changes. Key recommendations include implementing a dynamic reservoir operation system, enhancing forecasting tools, and incorporating green infrastructure to maintain water quality, support ecosystem resilience, and ensure sustainable water use in the RRB. This research emphasizes the need for localized strategies to address climate-driven hydrological changes and protect future water resources.

Hydrological Evaluation of CRA40 and ERA5 Reanalysis Precipitation Products over Ganjiang River Basin in Humid Southeastern China

This study evaluates two reanalysis precipitation products (CRA40 and ERA5) over the Ganjiang River Basin with precipitation data from 37 ground rainfall gauges and surface-observed stream flow data from January 1998 to December 2008. Direct comparison with rain gauge observations shows that both CRA40 and ERA5 can capture the spatial and temporal characteristics of precipitation at the basin scale of the Ganjiang River and reflect most of the precipitation events, but there are pronounced differences in the quality of precipitation between them. ERA5 performs better on the daily scale, capturing precipitation changes more accurately over short periods of time, while CRA40 performs better on the monthly scale, providing more stable and long-term precipitation trends. The results of stream flow simulations using two reanalysis precipitation products driving the VIC hydrological model show that (1) CRA40 outperforms ERA5 with a better Nash–Sutcliffe Efficiency (NSE, 0.65 and 0.6) and higher CC (0.96 and 0.91) in daily and monthly scale stream flow simulations, and ERA5 has a good CC (0.86 and 0.93, respectively), but its NSE is poor (0.29 and 0.30, respectively); (2) both CRA40 and ERA5 generally overestimate basin stream flows, especially during the flood season (April–September), with ERA5’s overestimation being more pronounced. This study is expected to provide a basis for the selection of reliable reanalysis products for Ganjiang River Basin precipitation and hydrological simulation.

Assessing hydrological changes in the Ganga River with the range of variability approach

ABSTRACT The study quantitatively assesses the hydrological alterations induced by the Luv Kush Barrage on the Ganga River, India, using the Range of Variability Approach (RVA) to analyze changes in flow conditions before and after the barrage construction. The Ganga River, a crucial water source for millions, has experienced significant hydrological changes due to human interventions, notably the construction of the Luv Kush Barrage. Employing indicators of hydrological alteration (IHAs), we compared pre-impact (1972–1994) and post-impact (2001–2014) flow data to ascertain the barrage’s effects on the river’s flow regime. Our findings reveal significant alterations in the magnitude of monthly streamflow, especially during the monsoon season, with notable decreases in July, August, and September. Alterations were observed in the annual extreme flow conditions, frequency, and duration of high and low pulses, indicating a comprehensive impact on the river’s natural flow patterns. These alterations affect the river’s ecological health, affecting aquatic habitats, fishery resources, and community livelihoods dependent on the river. The study underscores the necessity for integrated water resource management practices that consider the ecological and socio-economic implications of hydrological alterations.

RNN-Based Monthly Inflow Prediction for Dez Dam in Iran Considering the Effect of Wavelet Pre-Processing and Uncertainty Analysis

In recent years, deep learning (DL) methods, such as recurrent neural networks (RNN). have been used for streamflow prediction. In this study, the monthly inflow into the Dez Dam reservoir from 1955 to 2018 in southwestern Iran was simulated using various types of RNNs, including long short-term memory (LSTM), bidirectional long short-term memory (Bi-LSTM), gated recurrent unit (GRU), and stacked long short-term memory (Stacked LSTM). It was observed that considering flow discharge, temperature, and precipitation as inputs to the models yields the best results. Additionally, wavelet transform was employed to enhance the accuracy of the RNNs. Among the RNNs, the GRU model exhibited the best performance in simulating monthly streamflow without using wavelet transform, with RMSE, MAE, NSE, and R2 values of 0.061 m3/s, 0.038 m3/s, 0.556, and 0.642, respectively. Moreover, in the case of using wavelet transform, the Bi-LSTM model with db5 mother wavelet and decomposition level 5 was able to simulate the monthly streamflow with high accuracy, yielding RMSE, MAE, NSE, and R2 values of 0.014 m3/s, 0.008 m3/s, 0.9983, and 0.9981, respectively. Uncertainty analysis was conducted for the two mentioned superior models. To quantify the uncertainty, the concept of the 95 percent prediction uncertainty (95PPU) and the p-factor and r-factor criteria were utilized. For the GRU, the p-factor and r-factor values were 82% and 1.28, respectively. For the Bi-LSTM model, the p-factor and r-factor values were 94% and 1.06, respectively. The obtained p-factor and r-factor values for both models are within the acceptable and reliable range.

A multi-variable calibration framework at the grid scale for integrating streamflow with evapotranspiration data to improve the simulation of distributed hydrological model

Study regionThe Ganjiang River Basin, China Study focusParameter calibration is crucial for the accurate and reliable operation of hydrological models. Traditional methods face challenges in calibrating spatially heterogeneous parameters of distributed hydrological models, and existing multi-variable calibration strategies often fall short in comprehensively improving model performance, particularly in streamflow simulations. To address these challenges, this study proposes a multi-objective calibration framework that integrates observed streamflow data and satellite-based evapotranspiration (ET) data. The spatiotemporal information of the merged ET is utilized to calibrate six hydrological parameters of the Variable Infiltration Capacity (VIC) model at the grid scale, enhancing hydrological simulations for the Ganjiang River basin. New hydrological insightsCompared to the benchmark scheme based solely on streamflow, the proposed calibration framework improves simulations of area-average ET at the sub-basin scale and soil moisture content in the Ganjiang River basin, without compromising the accuracy of daily streamflow simulations. Additionally, notable enhancements are observed in monthly streamflow simulations. This study provides a promising and comprehensive calibration framework using satellite-based data to constrain parameters and enhance the performance of distributed hydrological models.

Exploring climate-change impacts on streamflow and hydropower potential: insights from CMIP6 multi-GCM analysis

ABSTRACT The generation of hydropower is profoundly influenced by shifts in streamflow patterns induced by climate change. This research examines changes in streamflow and the potential surge in hydropower generation over a span of 35 years (2015–2050) at the Bhakra Dam site within the Upper Sutlej River basin. Employing a deep learning methodology, particularly the long short-term memory (LSTM) model, in conjunction with Coupled Model Intercomparison Project (CMIP) 6 multi-global climate model (GCM), facilitates a thorough analysis of these dynamics. Six out of 14 bias-corrected statistically downscaled datasets (0.25° × 0.25° grid resolution) from CMIP6 multi-GCM were selected based on entropy and combined compromise solution techniques. This innovative approach is utilized to assess streamflows and project potential increases in hydropower at the Bhakra Dam site under shared socioeconomic pathways (SSP) scenarios, specifically SSP245 and SSP585. The results indicate a maximum increase of approximately 15% and 17% in mean monthly streamflow under SSP245 and SSP585, respectively. Moreover, dependability flows calculated at Q50, Q75, and Q90 show respective rises of 13%, 16%, and 17% under SSP245 and 21%, 17%, and 18% under SSP585. The projected hydropower potential exhibits an increase of up to 15.9% and 17.3% under SSP245 and SSP585, respectively.

Streamflow Intermittence in Europe: Estimating High‐Resolution Monthly Time Series by Downscaling of Simulated Runoff and Random Forest Modeling

AbstractKnowing where and when rivers cease to flow provides an important basis for evaluating riverine biodiversity, biogeochemistry and ecosystem services. We present a novel modeling approach to estimate monthly time series of streamflow intermittence at high spatial resolution at the continental scale. Streamflow intermittence is quantified at more than 1.5 million river reaches in Europe as the number of no‐flow days grouped into five classes (0, 1–5, 6–15, 16–29, 30–31 no‐flow days) for each month from 1981 to 2019. Daily time series of observed streamflow at 3706 gauging stations were used to train and validate a two‐step random forest modeling approach. Important predictors were derived from time series of monthly streamflow at 73 million 15 arc‐sec (∼500 m) grid cells that were computed by downscaling the 0.5 arc‐deg (∼55 km) output of the global hydrological model WaterGAP, which accounts for human water use. Of the observed perennial and non‐perennial station‐months, 97.8% and 86.4%, respectively, were correctly predicted. Interannual variations of the number of non‐perennial months at non‐perennial reaches were satisfactorily simulated, with a median Pearson correlation of 0.5. While the spatial prevalence of non‐perennial reaches is underestimated, the number of non‐perennial months is overestimated in dry regions of Europe where artificial storage abounds. Our model estimates that 3.8% of all European reach‐months and 17.2% of all reaches were non‐perennial during 1981–2019, predominantly with 30–31 no‐flow days. Although estimation uncertainty is high, our study provides, for the first time, information on the continent‐wide dynamics of non‐perennial rivers and streams.

Enhanced SWAT calibration through intelligent range-based parameter optimization

Hydrological models are vital tools in environmental management. Weaknesses in model robustness for hydrological parameters transfer uncertainties to the model outputs. For streamflow, the optimized parameters are the primary source of uncertainty. A reliable calibration approach that reduces prediction uncertainty in model simulations is crucial for enhancing model robustness and reliability. The optimization of parameter ranges is a key aspect of parameter calibration, yet there is a lack of literature addressing the optimization of parameter ranges in hydrological models. In this paper, we introduce a parameter calibration strategy that applies a clustering technique, specifically the Self-Organizing Map (SM), to intelligently navigate the parameter space during the calibration of the Soil and Water Assessment Tool (SWAT) model for monthly streamflow simulation in the Baishan Basin, Jilin Province, China. We selected the representative algorithm, the Sequential Uncertainty Fitting version 2 (SUFI-2), from the commonly used SWAT Calibration and Uncertainty Programs for comparison. We developed three schemes: SUFI-2, SUFI-2-Narrowing Down (SUFI-2-ND), and SM. Multiple diagnostic error metrics were used to compare simulation accuracy and prediction uncertainty. Among all schemes, SM outperformed the others in describing watershed streamflow, particularly excelling in the simulation of spring snowmelt runoff (baseflow period). Additionally, the prediction uncertainty was effectively controlled, demonstrating the SM's adaptability and reliability in the interval optimization process. This provides managers with more credible prediction results, highlighting its potential as a valuable calibration tool in hydrological modeling.

SIMULATION OF SPATIO-TEMPORAL DYNAMICS OF STREAM FLOW TO CLIMATE VARIABILITY IN NJORO RIVER CATCHMENT, KENYA

Climate variability continues to alter hydrological regimes and the response of many catchments globally thus threatening water security. Njoro River Catchment has not been an exception. There is a steady recognition of climate variability and adverse influences like deterioration of ecosystems, surface, and groundwater sources. Therefore, the study focused on modelling the impact of climate variability on the spatial and temporal distribution of stream flow in the Njoro River Catchment, Kenya. Modelling of stream flow response to climate variability using SWAT was carried out based on USGS/NASA downloaded Digital Elevation Model, FAO soil data, Landsat (MSS 1-5) LULC of 1978, and meteorological data for the period (1978 - 2017). Simulation of spatial and temporal impacts of climate variability was then carried out. Based on the analysis, SUFI-2 algorithm results revealed that the most sensitive parameter in Njoro River Catchment is SCS curve number for antecedent moisture condition (II) (CN2), ranked according to the highest sensitivity tested at p < 0.05. The least sensitive parameter is maximum canopy storage (CANMX). Overall uncertainty analysis results showed a good performance value of P-factor (0.72) and R-factor (0.38), which indicated that SWAT reproduced the dynamics of the catchment hydrologic response relatively well. The values of R2, NSE, and PBIAS for calibration and validation of monthly stream flow were 0.88 and 0.77, 0.86 and 0.74, and 5.51 % and -15.42 % respectively. From the spatio-temporal impacts of climate variability on hydrologic variables, actual evapotranspiration increased by about 5.6 %, 3 %, and 6.62 % in the 2nd, 3rd, and 4th decades. Surface runoff decreased in the 2nd decade by 34.56 % and increased by 4.55 % and 34.02 % in the 3rd and 4th decades. Annual water yield decreased in the 2nd decade by 29.69 % and increased by 4.25 % and 60.18 % in the 3rd and 4th decades respectively. On average, stream flow reduced by 30.91 % in the 2nd decade and increased by 5.47 % and 63.63 % in the 3rd and 4th decades respectively. These findings provide pertinent insights, which may perhaps enlighten decision-making in designing adaptable mitigation measures, catchment rehabilitation, and strategic initiatives for the integrated management of water resources.

Comparison of parameter optimization methods for a runoff forecast model based on a support vector machine

Long-term flow prediction is crucial in water resource management, with support vector regression (SVR) commonly used for runoff prediction. This paper aims to clarify SVR parameter optimization through a case study, focusing on the penalty coefficient (C), the kernel function parameter (γ), and the insensitive loss coefficient (ε), using different heuristic methods including particle swarm optimization (PSO), a complex method (CM), and simulated annealing (SA). A runoff prediction model based on SVR was established to forecast the monthly inflow runoff of Zipingpu Reservoir. The study evaluates the applicability of each optimization method for the SVM-based monthly streamflow forecast model. Results indicate CM has the best operational efficiency with an average running time of less than 2 s, significantly lower than PSO and SA. SA shows the best stability with an average standard deviation of fitness values of 0.00032, significantly lower than PSO's 0.0015 and CM's 0.021. The SA-SVM model reduces RMSE and MARE by 8.35% and 15% respectively, compared to the standard SVM model, while NS and R improve by 11.67% and 6.57%, achieving the highest prediction accuracy. Thus, the SA algorithm is the most applicable, followed by the PSO algorithm.

Exploring cumulative probability functions for streamflow drought magnitude: Global scale analysis and parametric vs. non-parametric comparisons

Streamflow droughts pose significant challenges to water resource management and environmental sustainability. The accurate characterization of streamflow drought magnitude is essential for effective mitigation, assessment and adaptation strategies. In this study, we explore the implications of choosing different cumulative probability distribution functions (cdfs) on the determination of streamflow drought magnitude, as assessed by the z-score values of the standardized streamflow index (SSI) and return periods. Our investigation encompasses a comprehensive global-scale analysis, with the aim of identifying the most suitable cdfs for characterizing streamflow droughts. To accomplish this objective, we incorporate global simulated streamflow values from the model WaterGAP 2.2d and observed streamflow values from 280 Global Runoff Data Centre (GRDC) stations for the period 1981 – 2010. We evaluate the performance of 118 parametric cdfs for each of the 12 calendar monthly streamflow values at each model grid cell and GRDC station to determine the best fitting parametric cdf. Moreover, we go beyond the conventional scope of such studies by further assessing the best-fitting cdfs alongside non-parametric cdfs such as the Empirical cumulative distribution function (ecdf) and the cumulative distribution function derived from the Kernel density estimation function (kcdf). Among our findings, we establish a subset of 23 parametric cdfs that exhibit the best fit for the global streamflow values concerning all 12 calendar monthly streamflow values at individual GRDC stations or grid cells within the specified period. Furthermore, we observe that the accuracy of streamflow simulation and the length of the reference period directly influence the selection of the most suitable parametric cdf. Our comparative analysis reaffirms the existing understanding that opting for the best-fitting parametric cdf, rather than a non-parametric alternative, mitigates the risk of overestimating drought magnitudes. However, we emphasize the crucial importance of considering the intended purpose, approach-specific sensitivities, and uncertainties when choosing the most suitable parametric cdf for any calendar monthly streamflow values.

Modularity Density Based-Edge Betweenness Method for Catchment Classification in Sabah

One of the most significant studies in hydrology is the classification of catchments. There are various reasons why catchment classification studies have been conducted, but most importantly is to anticipate the prediction of ungauged basin (PUB), among others. There are several ways of classification, each with its own set of assumptions and basis, that have been used for catchment classification. Recently, complex network concepts, specifically community structure, have emerged as key classification methods, and are now attaining traction in catchment classification. As a result, in this study, community structure approach, specifically the Modularity Density based- Edge Betweenness (MDEB) method is implemented to split the network into communities. By using the proposed method, a network of 30 monthly streamflow stations across Sabah is considered for catchment classification. The impact of correlation threshold, which vary from 0 to 1, denoting the strength between a pair of catchments is also investigated. As a result, four threshold values are chosen (T = 0.5, 0.55, 0.6, and 0.65), and the communities established with threshold value, T=0.6 are interpreted. The relationship between catchment characteristics and flow characteristics are investigated as well as distance-correlation relationships for communities identified to understand the Sabah catchment’s behaviors.

Spatial and Temporal Assessment of Baseflow Based on Monthly Water Balance Modeling and Baseflow Separation

Baseflow is the part of streamflow that is mainly replenished by groundwater. The protection of the biological environment and the growth of its water resources greatly depend on the spatial and temporal evolution of baseflow. Therefore, the Baizhiao (BZA) and Shaduan (SD) catchments of the Jiaojiang River Basin (JRB) in the Zhejiang province of China were selected as study areas. The ABCD model and Eckhardt method were used to calculate baseflow and baseflow index (BFI). The temporal and spatial evolution patterns of baseflow were analyzed through statistical analysis and the Mann–Kendall test. The results showed that the ABCD model performs well in simulating overall hydrological processes on the monthly streamflow at BAZ and SD stations with NSE (Nash–Sutcliffe Efficiency) values of 0.82 and 0.83 and Pbias (Percentage Bias) values of 9.2% and 8.61%, respectively. The spatial–temporal distribution of the BFI indicates the higher baseflow contribution in upstream areas compared to downstream areas at both stations. The baseflow and BFI had significant upward trends at the BZA and SD stations in the dry season, while their trends were not uniform during the wet period. These findings are essential guidance for water resource management in the JRB regions.

Enhancing spatial streamflow prediction through machine learning algorithms and advanced strategies

Forecasting and extending streamflow is a critical aspect of hydrology, especially where the time series are locally unavailable for a variety of reasons. The necessity of preprocessing, model fine-tuning, feature selection, or sampling to enhance prediction outcomes for streamflow forecasting using ML techniques is evaluated in this study. In this regard, the monthly streamflow at Pol-Chehr station is analyzed using various monthly rainfall and streamflow time series data from different stations. The results of streamflow prediction in the k-folds cross-validator approach are generally better than those of the time series approach, except when raw data with no preprocessing or feature selection is used. Applying the simple SVR model to raw data leads to the weakest result, but using the GA-SVR model on raw data significantly increases the Nash coefficient by about 215% and 72%, decreases the NRMSE by about 48% and 36% in the k-fold and time series approaches, even with no feature selection. On the other hand, standardization produces highly accurate model predictions in both the k-fold and time series approaches, with a minimum Nash coefficient of 0.83 and 0.73 during the test period in the simple SVR model, respectively. Finally, using optimization algorithms like GA to fine-tune ML models and feature selection does not always yield improved prediction accuracy, but it depends on whether raw or preprocessed data is chosen. In conclusion, combining k-fold cross-validator and preprocessing typically yields highly accurate predictive results, with an R value exceeding 93.7% (Nash = 0.83, SI = 0.55, NRMSE = 0.09), without requiring any additional fine-tuning or optimization. Using feature selection is only significant when utilizing the TS approach as well.

Optimal combinations of global evapotranspiration and terrestrial water storage products for catchment water balance

ABSTRACT During the last two decades, a great number of global products about evapotranspiration ( E ) and terrestrial water storage ( S ) have been released. This has led to numerous combinations for describing catchment water balance, and hence determining an appropriate combination has become an important issue. The main objective of this study is to evaluate various global products of E and S at the catchment scale and determine the most appropriate combination, taking two Korean catchments as examples. For E , we evaluated global evapotranspiration (Global ET), Global Land Evaporation Amsterdam Model (GLEAM), and Penman–Monteith–Leuning (PML). For S , six individual datasets of Gravity Recovery and Climate Experiment (GRACE) were evaluated, with Global Land Data Assimilation System (GLDAS) products used for comparison. We found that the performance is sensitive to the choice of E product while various S products displayed minimal differences in results. Based on four evaluation criteria, the combination of PML and GRACE-SH-GFZ is suggested as the most suitable pair. For the period between 2003 and 2013, these data exhibit noteworthy trends of increasing E , offset by decreasing S . Warming climate is suspected to be behind these trends. Our research presents an approach that allows for the estimation of monthly streamflow exclusively with global data products, which is an advancement in hydrological analysis and particularly useful for regions that lack in-situ data networks. This approach provides a new perspective in the application of global datasets for the assessment of water balance and could significantly improve predictions in ungauged basins.

Enhancing Monthly Streamflow Prediction Using Meteorological Factors and Machine Learning Models in the Upper Colorado River Basin

Streamflow prediction is crucial for planning future developments and safety measures along river basins, especially in the face of changing climate patterns. In this study, we utilized monthly streamflow data from the United States Bureau of Reclamation and meteorological data (snow water equivalent, temperature, and precipitation) from the various weather monitoring stations of the Snow Telemetry Network within the Upper Colorado River Basin to forecast monthly streamflow at Lees Ferry, a specific location along the Colorado River in the basin. Four machine learning models—Random Forest Regression, Long short-term memory, Gated Recurrent Unit, and Seasonal AutoRegresive Integrated Moving Average—were trained using 30 years of monthly data (1991–2020), split into 80% for training (1991–2014) and 20% for testing (2015–2020). Initially, only historical streamflow data were used for predictions, followed by including meteorological factors to assess their impact on streamflow. Subsequently, sequence analysis was conducted to explore various input-output sequence window combinations. We then evaluated the influence of each factor on streamflow by testing all possible combinations to identify the optimal feature combination for prediction. Our results indicate that the Random Forest Regression model consistently outperformed others, especially after integrating all meteorological factors with historical streamflow data. The best performance was achieved with a 24-month look-back period to predict 12 months of streamflow, yielding a Root Mean Square Error of 2.25 and R-squared (R2) of 0.80. Finally, to assess model generalizability, we tested the best model at other locations—Greenwood Springs (Colorado River), Maybell (Yampa River), and Archuleta (San Juan) in the basin.