Streamflow Time Research Articles

Coupling ANFIS with ant colony optimization (ACO) algorithm for 1-, 2-, and 3-days ahead forecasting of daily streamflow, a case study in Poland

Finding an efficient and reliable streamflow forecasting model has always been an important challenge for managers and planners of freshwater resources. The current study, based on an adaptive neuro-fuzzy inference system (ANFIS) model, was designed to predict the Warta river (Poland) streamflow for 1 day, 2 days, and 3 days ahead for a data set from the period of 1993–2013. The ANFIS was additionally combined with the ant colony optimization (ACO) algorithm and employed as a meta-heuristic ANFIS-ACO model, which is a novelty in streamflow prediction studies. The investigations showed that on a daily scale, precipitation had a very weak and insignificant effect on the river’s flow variation, so it was not considered as a predictor input. The predictor inputs were selected by the autocorrelation function from among the daily streamflow time lags for all stations. The predictions were evaluated with the actual streamflow data, using such criteria as root mean square error (RMSE), normalized RMSE (NRMSE), and R2. According to the NRMSE values, which ranged between 0.016–0.006, 0.030–0.013, and 0.038–0.020 for the 1-day, 2-day, and 3-day lead times, respectively, all predictions were classified as excellent in terms of accuracy (prediction quality). The best RMSE value was 1.551 m3/s and the highest R2 value was equal to 0.998, forecast for 1-day lead time. The combination of ANFIS with the ACO algorithm enabled to significantly improve streamflow prediction. The use of this coupling can averagely increase the prediction accuracies of ANFIS by 12.1%, 12.91%, and 13.66%, for 1-day, 2-day, and 3-day lead times, respectively. The current satisfactory results suggest that the employed hybrid approach could be successfully applied for daily streamflow prediction in other catchment areas.

Regional variability and changing water distributions drive large-scale water resource availability in Alberta, Canada

Geographic inequalities in water distribution can lead to relatively small regions driving or buffering changes in water availability in major watersheds. This research provides a baseline quantification of water distribution in Alberta, Canada, and evaluates changes in water yield, streamflow timing, and climate from 1976 to 2015. Annual water yields for 77 contributing watershed areas were calculated and assessed for trends, and 45 unregulated watersheds were evaluated for changes in key streamflow timing metrics. Mountain headwaters supply 22–38% of annual flow of major rivers, while plains watersheds provide relatively low annual yield. Annual yield and precipitation decreases dominated northern watersheds, with increases in southern watersheds, and mixed trends in central watersheds. Later spring freshet timing and earlier onset of the low flow season were detected for most watersheds, leading to more water over a shorter duration when paired with increasing yield, and water losses in late summer-autumn when paired with decreasing yield. Yield and flow timing metrics are strongly related to precipitation with temperature as a secondary driver, except freshet timing, which is driven by spring temperature. This study provides comprehensive information about geographic drivers of changing water distribution and quantifies the disparity between regions of water surplus and deficit.

Using Machine Learning to Identify Hydrologic Signatures With an Encoder–Decoder Framework

AbstractHydrologic signatures are quantitative metrics that describe a streamflow time series. Examples include annual maximum flow, baseflow index and recession shape descriptors. In this paper, we use machine learning (ML) to learn encodings that are optimal ML equivalents of hydrologic signatures, and that are derived directly from the data. We compare the learned signatures to classical signatures, interpret their meaning, and use them to build rainfall‐runoff models in otherwise ungauged watersheds. Our model has an encoder–decoder structure. The encoder is a convolutional neural net mapping historical flow and climate data to a low‐dimensional vector encoding, analogous to hydrological signatures. The decoder structure includes stores and fluxes similar to a classical hydrologic model. For each timestep, the decoder uses current climate data, watershed attributes and the encoding to predict coefficients that distribute precipitation between stores and store outflow coefficients. The model is trained end‐to‐end on the U.S. CAMELS watershed data set to minimize streamflow error. We show that learned signatures can extract new information from streamflow series, because using learned signatures as input to the process‐informed model improves prediction accuracy over benchmark configurations that use classical signatures or no signatures. We interpret learned signatures by correlation with classical signatures, and by using sensitivity analysis to assess their impact on modeled store dynamics. Learned signatures are spatially correlated and relate to streamflow dynamics including seasonality, high and low extremes, baseflow and recessions. We conclude that process‐informed ML models and other applications using hydrologic signatures may benefit from replacing expert‐selected signatures with learned signatures.

GTAR: a new ensemble evolutionary autoregressive approach to model dissolved organic carbon

AbstractThis article explores the forecasting capabilities of three classic linear and nonlinear autoregressive modeling techniques and proposes a new ensemble evolutionary time series approach to model and forecast daily dynamics in stream dissolved organic carbon (DOC). The model used data from the Oulankajoki River basin, a boreal catchment in Northern Finland. The models that were evolved used both accuracy and parsimony measures including autoregressive (AR), vector autoregressive (VAR), and self-exciting threshold autoregressive (SETAR). The new method, called genetic-based SETAR (GTAR), evolved through the integration of state-of-the-art genetic programming with SETAR. To develop the models, high-resolution DOC concentration and daily streamflow (as the external input for VAR) were measured at the same gauging station throughout the ice free season. The results showed that all the models characterize the DOC dynamics with an acceptable 1-day-ahead forecasting accuracy. Use of the streamflow time series as an exogenous variable did not increase the predictive accuracy of AR models. Moreover, the hybrid GTAR provided the best accuracy for the holdout testing data and proved to be a suitable approach for predicting DOC in boreal conditions.

Toward an improved estimation of flood frequency statistics from simulated flows

AbstractThe estimation of extreme flood frequency for ungauged or poorly gauged catchments is a longstanding problem of great practical importance. Simulated streamflow derived from distributed hydrological models can be used to address this issue, but their representation of extreme flood peaks is often prone to large biases. This study evaluates the potential of a nonasymptotic statistical approach able to consider all the independent flood peaks instead of extremes only, the Simplified Metastatistical Extreme Value (SMEV), for the estimation of extreme flood frequency from time series of simulated streamflow. We examined 28 years of simulated daily streamflow across the contiguous United States and compared SMEV to traditional statistical models based on annual maxima. Our results suggest that when its assumptions are met SMEV can moderate the impact of hydrological model biases in the quantification of extreme flood frequency. SMEV exhibits a lower relative difference between quantiles derived from observations and simulations for all return periods and forcing dataset. Quantiles estimated from simulated streamflow time series (28‐year records) using SMEV are usually in better agreement with the estimates based on 70‐year‐long observations. Geographical variations in the results of SMEV are noticed, with a better performance of SMEV in the east and west coasts (California, New England, and Mid‐Atlantic) and in the southwestern regions (Texas‐Gulf). These results indicate that the potential of SMEV for flood frequency analyses in ungauged and poorly gauged basins deserves further investigations.

Spatial variability in Alpine reservoir regulation: deriving reservoir operations from streamflow using generalized additive models

Abstract. Reservoir regulation affects various streamflow characteristics, from low to high flows, with important implications for downstream water users. However, information on past reservoir operations is rarely publicly available, and it is hardly known how reservoir operation signals, i.e. information on when water is stored in and released from reservoirs, vary over a certain region. Here, we propose a statistical model to reconstruct reservoir operation signals in catchments without information on reservoir operation. The model uses streamflow time series observed downstream of a reservoir that encompass a period before and a period after a known year of reservoir construction. In a first step, a generalized additive model (GAM) regresses the streamflow time series from the unregulated pre-reservoir period on four covariates including temperature, precipitation, day of the year, and glacier mass balance changes. In a second step, this GAM, which represents natural conditions, is applied to predict natural streamflow, i.e. streamflow that would be expected in the absence of the reservoir, for the regulated period. The difference between the observed regulated streamflow signal and the predicted natural baseline should correspond to the reservoir operation signal. We apply this approach to reconstruct the seasonality of reservoir regulation, i.e. information on when water is stored in and released from a reservoir, from a dataset of 74 catchments in the central Alps with a known reservoir construction date (i.e. date when the reservoir went into operation). We group these reconstructed regulation seasonalities using functional clustering to identify groups of catchments with similar reservoir operation strategies. We show how reservoir management varies by catchment elevation and that seasonal redistribution from summer to winter is strongest in high-elevation catchments. These elevational differences suggests a clear relationship between reservoir operation and climate and catchment characteristics, which has practical implications. First, these elevational differences in reservoir regulation can and should be considered in hydrological model calibration. Furthermore, the reconstructed reservoir operation signals can be used to study the joint impact of climate change and reservoir operation on different streamflow signatures, including extreme events.

Hydrological changes caused by the construction of dams and reservoirs: The CECP analysis.

We investigated the influence of the construction of cascade dams and reservoirs on the predictability and complexity of the streamflow of the São Francisco River, Brazil, by using complexity entropy causality plane (CECP) in its standard and weighted form. We analyzed daily streamflow time series recorded in three fluviometric stations: São Francisco (upstream of cascade dams), Juazeiro (downstream of Sobradinho dam), and Pão de Açúcar station (downstream of Sobradinho and Xingó dams). By comparing the values of CECP information quantifiers (permutation entropy and statistical complexity) for the periods before and after the construction of Sobradinho (1979) and Xingó (1994) dams, we found that the reservoirs' operations changed the temporal variability of streamflow series toward the less predictable regime as indicated by higher entropy (lower complexity) values. Weighted CECP provides some finer details in the predictability of streamflow due to the inclusion of amplitude information in the probability distribution of ordinal patterns. The time evolution of streamflow predictability was analyzed by applying CECP in 2 year sliding windows that revealed the influence of the Paulo Alfonso complex (located between Sobradinho and Xingó dams), construction of which started in the 1950s and was identified through the increased streamflow entropy in the downstream Pão de Açúcar station. The other streamflow alteration unrelated to the construction of the two largest dams was identified in the upstream unimpacted São Francisco station, as an increase in the entropy around 1960s, indicating that some natural factors could also play a role in the decreased predictability of streamflow dynamics.

Insights From Dayflow: A Historical Streamflow Reanalysis Dataset for the Conterminous United States

AbstractReconstructed historical streamflow time series can supplement limited streamflow gauge observations. However, there are common challenges of typical modeling approaches: process‐based hydrologic models can be data/computation‐intensive, and statistics‐based models can be region/stream‐specific. Here we present a nationally scalable modeling framework integrating the simulated runoff from the Variable Infiltration Capacity (VIC) model with the Routing Application for Parallel computatIon of Discharge (RAPID) routing model leveraging high‐performance computing. We demonstrate an efficient method of assimilating streamflow at US Geological Survey (USGS) streamflow monitoring sites using a simple hierarchical approach in the VIC‐RAPID framework. The result is a reconstructed 36‐year (1980–2015) daily and monthly streamflow dataset (Dayflow) at ∼2.7 million NHDPlusV2 stream reaches in the conterminous US (CONUS). We perform a comprehensive evaluation at 7,526 USGS sites and characterize their error statistics. The results demonstrate that 49% of the USGS sites demonstrate Kling–Gupta Efficiency (KGE) > 0.5 and 58% of the sites show percentage bias within ±20% for the daily naturalized streamflow. Streamflow data assimilation across CONUS shows an overall improvement over naturalized streamflow, notably in the western semiarid‐to‐arid regions. Comparison to other national and global streamflow reanalysis datasets such as the National Water Model and Global Reach‐scale A priori Discharge Estimates for SWOT demonstrates improved KGE, reduced bias, and directions for Dayflow improvements. Investigations of error statistics with key hydrologic, hydroclimatic, and geomorphologic basin characteristics reveal region‐specific patterns which may help improve future framework applications. Overall, Dayflow may enable a better understanding of hydrologic conditions in a changing environment, especially in locations currently not represented by streamflow monitoring networks.

A 21st‐Century perspective on snow drought in the Upper Colorado River Basin

AbstractIn the Upper Colorado River Basin (UCRB), there is a deep reliance on seasonal snowpack for maintenance of water resources. The term “snow drought” has recently emerged to describe periods of anomalously low snowpack. Unique seasonal patterns in precipitation and temperature that drive snow drought can have distinct hydrologic signatures, and these relationships have not been carefully studied in the UCRB. Here we examine snow drought with a new classification scheme using peak snow water equivalent (SWE) and the ratio of basin‐wide modeled peak SWE to accumulated (onset to peak) precipitation (SWE/P) that clusters snow drought years into three distinct groups—“warm,” “dry,” and “warm & dry”—that minimize within‐group variance. Over the period 1916–2018, we identify 14 warm years ( = 160 mm; = 0.24), 24 dry years ( = 117 mm; = 0.35), and 21 warm & dry years ( = 94 mm; = 0.23). An elevation‐based analysis reveals two distinct patterns: warm snow droughts see severe SWE reductions primarily at lower (<2600 m) elevations (65% at lower elevations, 37% overall), whereas “dry” scenarios exhibit a consistent reduction across all elevations (39% overall). Using naturalized streamflow data, we also differentiate snow droughts by their earlier streamflow timing and decreased peakedness (warm: 7 days, 2%; dry: 7 days, 2%; warm & dry: 13 days, 5%). This research provides new insights into snow drought patterns relevant for regional water management.

Stochastic Volatility Modeling of Daily Streamflow Time Series

AbstractUnder the changing climate, the natural characteristics of hydrological processes are assumed to have been intensified, and the volatility of these processes to have increased significantly. However, the applicability of traditional time series analysis methods and the commonly used Gaussian GARCH (Generalized AutoRegressive Conditional Heteroskedasticity) to streamflow modeling has been limited. Although SV‐type (stochastic volatility) models have appeared as competitive alternatives to GARCH‐type models owing to their flexibility in modeling volatilities, they have not been applied in hydrological studies. Therefore, this study assesses the applicability of SV models to streamflow modeling using daily streamflow series from 21 hydrological stations in the Yellow River basin. Considering the influence of different types of GARCH and residual distributions, 7 hybrid models with variance fluctuation (fractional autoregressive integrated moving average‐GARCH) based on 9 distributions are compared to determine the optimal model for each station. Then 4 SV‐type models are introduced and compared with the results of the optimal GARCH‐type model to verify their applicability. The results show that: (a) the Gaussian distribution is not applicable in both GARCH‐type and SV‐type models for modeling daily streamflow; (b) although the GARCH‐type models are shown to describe the volatility of streamflow processes and improve the modeling performance, large residuals have been observed in the results of the analysis during the peak flow period; and (c) SV‐type models can better describe the streamflow series with time‐varying variance and accurately capture the occurrence of peak streamflow. The findings of this study offer practical and promising time series analysis methods for daily streamflow modeling.

Recurrent neural networks for dynamical systems: Applications to ordinary differential equations, collective motion, and hydrological modeling.

Classical methods of solving spatiotemporal dynamical systems include statistical approaches such as autoregressive integrated moving average, which assume linear and stationary relationships between systems' previous outputs. Development and implementation of linear methods are relatively simple, but they often do not capture non-linear relationships in the data. Thus, artificial neural networks (ANNs) are receiving attention from researchers in analyzing and forecasting dynamical systems. Recurrent neural networks (RNNs), derived from feed-forward ANNs, use internal memory to process variable-length sequences of inputs. This allows RNNs to be applicable for finding solutions for a vast variety of problems in spatiotemporal dynamical systems. Thus, in this paper, we utilize RNNs to treat some specific issues associated with dynamical systems. Specifically, we analyze the performance of RNNs applied to three tasks: reconstruction of correct Lorenz solutions for a system with a formulation error, reconstruction of corrupted collective motion trajectories, and forecasting of streamflow time series possessing spikes, representing three fields, namely, ordinary differential equations, collective motion, and hydrological modeling, respectively. We train and test RNNs uniquely in each task to demonstrate the broad applicability of RNNs in the reconstruction and forecasting the dynamics of dynamical systems.

AI‐based runoff simulation based on remote sensing observations: A case study of two river basins in the United States and Canada

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

Enhancing robustness of monthly streamflow forecasting model using embedded-feature selection algorithm based on improved gray wolf optimizer

Accurate streamflow prediction plays an essential role in guaranteeing the sustainable utilization and management of water resources. In recent years, Artificial Intelligence (AI) models have been widely used for flow prediction. The performance of these models depends on the appropriate calibration of the input features and model parameters. Theoretically, the embedded feature selection method directly takes the final prediction model as a prediction indicator, which has the unique advantage of parallel optimization of feature and prediction model parameters compared with other methods. Despite being widely used in many other fields, its streamflow forecasting abilities are thus far unknown. In this paper, an embedded prediction model (EFS-SVMIGWO) with improved gray wolf optimizer (IGWO) and support vector machine (SVM) is proposed based on the principle of embedded feature selection method and validated with monthly runoff prediction at Kizil reservoir station in Xinjiang, China. The validation results demonstrate that the EFS-SVMIGWO model has consistently better accuracy and stable values than the benchmark methods (Including autoregressive integrated moving average, random forest, neural network and SVM models based on filtered selection methods). Moreover, IGWO is compared to differential evolution (DE), particle swarm optimization (PSO), whale optimization algorithm (WOA), sparrow search algorithm (SSA), and gray wolf optimizer (GWO), and the results show that IGWO has better convergence speed and solution quality in feature and model parameter parallel optimization tasks. Overall research and analysis indicate that the EFS-SVMIGWO model can exhibit convincing performance in monthly streamflow forecasting. Thus, it is of great importance to carefully choose the input variables and parameters to develop more effective models for forecasting monthly streamflow time series.

Gap-filling of daily precipitation and streamflow time series: a method comparison at random and sequential gaps

ABSTRACT This study compared the performance of seven gap-filling methods in daily streamflow and precipitation data and assessed the maximum number of gap days on which the methods perform well. Random (occurring randomly in the data series) and sequential (consecutive days with missing record) gaps were considered. Results show that the type of gaps affects the performance of the methods for gap-filling streamflow and precipitation data. Concerning random gaps, the best methods for streamflow were autoregressive integrate moving average and spline interpolation. For precipitation, the best methods were inverse distance weighting and linear regression (LR). Regarding sequential gaps, LR and multiple regression perform well in gap-filling up to 60 consecutive days in streamflow series. The other methods perform well up to 15 consecutive gap days. For precipitation series, the methods performed well up to seven consecutive missing days. For longer gaps (15, 30, 45 and 60 days), the methods performed poorly.

Past and future changes toward earlier timing of streamflow over Pakistan from bias-corrected regional climate projections (1962–2099)

Pakistan has experienced seasonal changes of streamflow, causing a lack of available water resources for agriculture. However, understanding of future seasonal changes of streamflow over Pakistan remains limited. This study assessed the past and future changes in streamflow timing along the four major rivers of Pakistan (Upper Indus, Kabul, Jhelum, and Chenab River basins), using observational data and bias-corrected hydrological projections. Firstly, the VIC-river routing model was simulated forced by simulated daily surface and base runoff data from six CORDEX-South Asia regional climate models (1962–2099). Secondly, the minimum and seasonality bias in simulated daily streamflow data were corrected based on observational records. To quantify seasonal changes of the hydrologic regime, half of annual cumulative streamflows (HCSs) and center-of-volume dates (CVDs) were computed from observed and bias-corrected simulated streamflow data. Over 1962–2019, observational records showed a significant decreasing trend in CVD (that is, an earlier onset of the wet season) by a range between −4.5 and −12.6 days across the three river basins, except for Chenab River basin. Bias-corrected hydrologic projections showed decreased CVD across the four study river basins by −4.2 to −6.3 days during the record period (1962–2019). The decreased CVDs ranges from −5 to −20 days in the near future (the 2050–2059 average) and −11 days to −37 days in the far future (the 2090–2099 average). This study reported diverse hydrologic responses to a similar magnitude of near-surface temperature in Pakistan, highlighting a need to develop basin-specific water resources mamangement and policies for climate change adaptation.

Attribution analysis of different driving forces on vegetation and streamflow variation in the Jialing River Basin, China

Abstract As an important tributary of the Yangtze River basin, the Jialing River Basin (JRB) has a variable topography and fragile ecological environment. Under the influence of climate warming and human activities, the runoff and vegetation cover of the area are undergoing significant changes. For evaluating the impact of different factors on vegetation and streamflow variation in the JRB, the univariate linear regression method was employed to analyze the variation trend of streamflow and vegetation, and a heuristic segmentation algorithm was applied for identifying the mutation year of streamflow and NDVI time-series data in the JRB. Then, the influence degree of different factors on streamflow variation in the JRB was calculated by the Budyko equation and elastic coefficient method. Finally, the contribution rates of different factors to vegetation variation in the JRB were calculated by the multiple linear regression method. The results indicated that the mutation year of streamflow and NDVI time series data were 1985 and 2006, respectively. The contribution rates of climate factors and human activities to streamflow change in the JRB were 42.7 and 57.3%, respectively. The contribution rates to vegetation change were 28.89 and 71.11%, respectively. In general, human activities are the main driving force leading to runoff and vegetation changes in the JRB. This study can provide a theoretical basis for the ecological environment protection and water resources management of the JRB.

Unveiling the climatic origin of streamflow persistence through multifractal analysis of hydro-meteorological datasets of India

ABSTRACT Persistence and multifractality are two fundamental properties of hydrological time series. This study aims at understanding the origin of the Hurst exponent of streamflow time series. Thus, we have obtained streamflow, rainfall, and potential evapotranspiration (PET) time series data pertaining to 101 stations in 17 river basins in India. The persistence and multifractal properties of the time series were estimated using multifractal detrended fluctuation analysis (MFDFA). Results suggest that most of the rainfall time series (85%) exhibit short-term persistence while all the PET time series and 59% of the streamflow time series exhibit long-term persistence. The Hurst exponent exhibited by the river flow series simulated by the dynamic Budyko (DB) model that considers only rainfall and PET as inputs displayed a good agreement with the Hurst exponent of observed streamflow, supporting our hypothesis that climatic factors typically shape the value of the Hurst exponent.

Daily Streamflow Time Series Modeling by Using a Periodic Autoregressive Model (ARMA) Based on Fuzzy Clustering

The behavior of hydrological processes is periodic and stochastic due to the influence of climatic factors. Therefore, it is crucial to develop the models based on their periodicity and stochastic nature for prediction. Furthermore, forecasting the streamflow, as one of the main components of the hydrological cycle, is a primary subject. In this study, a statistical method, Fuzzy C-means clustering, was used to find the periodicity in the daily discharge time series, whereas autoregressive moving average, ARMA, was used in modeling every cluster. Dividing the daily stream flow time series into smaller groups based on their similar statistical behavior by using a statistical method for analyzing and a combination of Fuzzy C-means clustering and ARMA modeling is the innovation of this study. We draw on the daily discharge data of four different river stations in Hesse state in Germany. The collected data cover 18 years, from 2000 to 2017. Root mean square error (RMSE) was used to evaluate the accuracy. The results revealed that the performance of ARMA in four stations for predicting every cluster was reliable. In addition, it must be highlighted that by clustering the daily stream flow time series into smaller groups, forecasting different days of the year will be possible.

Investigation of cross-entropy-based streamflow forecasting through an efficient interpretable automated search process

Streamflow forecasting has always been important in water resources management, particularly the peak flow, which often determines the seriousness of the impending flood. However, the highly imbalanced flow distribution often hinders the machine learning algorithm's performance. In this paper, streamflow forecasting was approached through the formulation of two distinct machine learning problems: categorical streamflow forecast and regression streamflow forecast. Due to the distinctive characteristics of these two adopted forms, selecting the correct algorithm for the machine learning problem along with their hyperparameter tuning process is critical to the realization of the desired results. For the distinct streamflow formulated scenarios, three neural network algorithms and their hyperparameter tuning strategy were investigated. The comparative empirical studies had revealed that formulated categorical-based streamflow forecast is a better choice than a regression-based streamflow forecast, regardless of the algorithms used; for instance, the f1-score of 0.7 (categorical based) is obtained compared to the 0.53 (regression based) for the LSTM in scenario 1 (binary). Furthermore, forest-based algorithms were investigated and shown to be superior at forecasting high streamflow fluctuations in situations featuring low-dimensional streamflow input. Besides, encoding the streamflow time series as images (input) for forecasting purposes would require a thorough analysis as there is a discrepancy in the results, revealing that not all approaches are suitable for streamflow image transformation. The functional ANOVA analysis provided evidence to substantiate the Bayesian optimization results, implying that the hyperparameters were effectively optimized.

Perspective Impact on Water Environment and Hydrological Regime Owing to Climate Change: A Review

This study summarizes reviews on climate change’s impact on the water environment and hydrological regime. The results indicate a strong relationship between the climatological parameters and hydrological patterns. This relationship can be determined in two steps: (1) define the variations in climatological factors, particularly temperature and precipitation, and (2) measure the variations in runoff and inflows to streams and river systems using different statistical and global climate modeling approaches. It is evident that the increasing global temperatures have significant positive effects on runoff variations and evapotranspiration. Similarly, the increase in temperature has speeded up the melting of glaciers and ice on hilly terrains. This is causing frequent flash floods and a gradual rise in the sea level. These factors have altered the timing of stream flow into rivers. Furthermore, the accumulation of greenhouse gases, variations in precipitation and runoff, and sea-level rise have significantly affected freshwater quality. These effects are likely to continue if timely mitigation and adaptation measures are not adopted.