Streamflow Time Research Articles

A Tale of Two Storms: Inter‐Storm Variability of Stable Water Isotopes in a Solute Transport Model

ABSTRACTStable isotopic methods in hydroclimate monitoring are powerful for improving water resources management, but applications are limited, especially in semi‐arid regions where such management is needed most. Here, we show that we can address shortcomings related to the lack of a seasonal signal using stable water isotopic signatures measured in precipitation over the East San Francisco Bay area, California, during two contrasting events sampled at more than 20 locations in the winter of 2023. The observed range in δ18O in the rain samples is similar for both storms. However, the distributions do not overlap—the mean air temperature and δ18O during Winter Storm Olive (February 2023) were 2°C and − 12‰, respectively, while a warm atmospheric river event (March 2023) had a mean temperature of 9°C and δ18O of −6‰, close to the long‐term average δ18O measured in local precipitation. The Winter Storm showed expected trends in δ18O related to geography (i.e., lower with greater distance inland and elevation), while the atmospheric river δ18O pattern was more spatially uniform. We use hydrometric data from a gaged watershed in the study area and isotopic signatures of rain sampled during the two storm events and apply a solute transport model (StorAge selection) with a travel‐time approach to examine predicted watershed responses and potential water tracing applications. In this virtual experiment, we find that event size exerts a strong control on the relative amounts of runoff versus pre‐event water in the stream, while uncertainty in stream hydrograph separation is related to the degree of contrast between precipitation/runoff and pre‐event water. Key to flood prediction, adaptation, and mitigation, especially in coastal urban areas, is knowledge of the contributing water sources and timing of stream flow. The strong contrast in stable isotopes between these two events, close in time and over the same area, illustrates the potential to use stable isotope signatures to track the transport and mixing of events through natural and engineered watersheds that are threatened by climate whiplash events.

Has unsustainable groundwater use induced low flow regimes in the Urucuia Aquifer System? An urgent call for integrated water management

Groundwater depletion in the Urucuia Aquifer System (UAS) has concerned the Brazilian water agencies since it is the principal water source in the arid length of the São Francisco river. In contrast, the irrigation fields increasing and uncertainty regarding climate variability have impaired the management of this important hydrograph region of Brazil. Therefore, the proper management of the UAS relies on the correct definition of the factors that are impacting most of the ground and surface waters, i.e., whether they are from climate or anthropogenic activities. Relying on Artificial Intelligence (AI) modeling and hydrological signatures, this study proposed a methodology to disentangle the role of climatic variability and groundwater pumping on streamflow dynamics by rebuilding natural time series in four watersheds within the UAS’s boundaries. By comparing the natural to the observed streamflow time series, the long-term 90th percentile (Q_90) discharge, from which 80% can be granted for multiple uses in the State of Bahia, decreased by up to 50% in the UAS. Yet, the watersheds’ productivity, given by the 90% specific yield, ranged from -52.3% (3.8 L/s.km2 to 1.8 L/s.km2) to -74.0% (23.4 L/s.km2 to 6.1 L/s.km2). Low flow duration was ∼ 1,650 days in natural conditions and increased to 10,354 days in the current land use and cover scenario. Changes in the maximum low flow duration length ranged from 191.7% to 1,315.7%, and in the low flow, deficits were up to 7,150%. These results highlight that groundwater pumping is the principal factor of the UAS’s dryness since climate variability could not track streamflow decreases. However, climate variability is secondary because of the intensification of the atmosphere demand. Although it is a Brazilian application, the proposed methodology can be applied in other aquifer systems to guide decision-makers' management strategies worldwide.

Toward Utilizing Similarity in Hydrologic Data Assimilation

Similarity to reality is a necessary property of models in earth sciences. Similarity information can thus possess a large potential in advancing geophysical modeling and data assimilation. We present a formalism for utilizing similarity within the existing theoretical data assimilation framework. Two examples illustrate the usefulness of utilizing similarity in data assimilation. The first, theoretical example shows changes in the accuracy of the amplitude estimate in the presence of a phase error in a sine function, where correcting the phase error prior to the assimilation reduces the degree of ill-posedness of the assimilation problem. This signifies the importance of accounting for the phase error in order to reduce the error in the amplitude estimate of the sine function. The second, real-world example illustrates that timing errors in simulated flow degrade the data assimilation performance, and that the flow gradient-informed shifting of rainfall time series improved the assimilation results with less adjusting model states. This demonstrates the benefit of utilizing streamflow gradients in shifting rainfall time series in a way to improve streamflow timing—vital information for flood early warning and preparedness planning. Finally, we discuss the implications, potential issues, and future challenges associated with utilizing similarity in hydrologic data assimilation.

Investigating the Performance of the Informer Model for Streamflow Forecasting

Recent studies have shown the potential of transformer-based neural networks in increasing prediction capacity. However, classical transformers present several problems such as computational time complexity and high memory requirements, which make Long Sequence Time-Series Forecasting (LSTF) challenging. The contribution to the prediction of time series of flood events using deep learning techniques is examined, with a particular focus on evaluating the performance of the Informer model (a particular implementation of transformer architecture), which attempts to address the previous issues. The predictive capabilities of the Informer model are explored and compared to statistical methods, stochastic models and traditional deep neural networks. The accuracy, efficiency as well as the limits of the approaches are demonstrated via numerical benchmarks relating to real river streamflow applications. Using daily flow data from the River Test in England as the main case study, we conduct a rigorous evaluation of the Informer efficacy in capturing the complex temporal dependencies inherent in streamflow time series. The analysis is extended to encompass diverse time series datasets from various locations (>100) in the United Kingdom, providing insights into the generalizability of the Informer. The results highlight the superiority of the Informer model over established forecasting methods, especially regarding the LSTF problem. For a forecast horizon of 168 days, the Informer model achieves an NSE of 0.8 and maintains a MAPE below 10%, while the second-best model (LSTM) only achieves −0.63 and 25%, respectively. Furthermore, it is observed that the dependence structure of time series, as expressed by the climacogram, affects the performance of the Informer network.

Drought‐Rich Periods Are More Likely Than Flood‐Rich Periods in Brazil

AbstractStreamflow exhibits persistent decadal variability; however, it is unclear if the magnitude and spatial extent of these variabilities are symmetric for droughts and floods. Here, we examine drought‐rich and flood‐rich periods in 319 streamflow gauges in Brazil from 1940 to 2020. Drought‐ and flood‐rich periods are detected by computing annual streamflow minima and maxima time series and using scan statistics to verify if events exceeding a threshold follow a Bernoulli process. We contrast streamflow time clustering with rainfall, evaporation, water abstraction, the Atlantic Multidecadal Oscillation (AMO), and the Pacific Decadal Oscillation (PDO). We detected a higher spatial frequency of drought‐ than flood‐rich periods. For 5‐year return period thresholds, drought‐rich periods are observed in 81% of the basins, 16.7 times the false positive rate (4.8%) and 4.7 times flood‐rich periods (17%). This asymmetry is linked with sharp increases in water abstractions since the 1990s and a higher prevalence of rainfall‐poor periods (41% of gauges) compared to rainfall‐rich (22% of gauges), which we interpret as being further amplified into drought‐rich periods due to an interannual persistence of water storage deficits. Brazil experienced a dry period until the 1960s, extensive flooding in the 1980s, and record low flows from the 2000s onward. Drought and flood‐rich periods are well aligned with rainfall clustering, water abstractions, the AMO and PDO. Droughts‐rich periods are more frequent in shorter time scales (several years to one decade) and flood‐rich periods in longer time scales (a few decades). Our findings highlight the nonlinearity and asymmetry of drought and flood change at decadal scales.

Hydrographic shifts in coastal waters reflect climate-driven changes in hydrological regimes across Northwestern Patagonia

Climate-driven changes in freshwater inputs have been shown to affect the structure and function of coastal ecosystems. We evaluated changes in the influence of river runoff on coastal systems of Northwestern Patagonia (NWP) over recent decades (1993–2021) by combined analysis of long-term streamflow time series, hydrological simulation, satellite-derived and reanalysis data on sea surface conditions (temperature, turbidity, and salinity). Significant decreases in minimum streamflow across a zone spanning six major river basins were evident at weekly, monthly, and seasonal scales. These changes have been most pronounced in mixed-regime northern basins (e.g., Puelo River) but appear to be progressing southward to rivers characterised by a nival regime. In the adjacent two-layer inner sea, reduced freshwater input corresponds with a shallower halocline and increased surface temperatures across northern Patagonia. Our results underscore the rapidly evolving influence of rivers on adjacent estuarine and coastal waters in NWP. We highlight the need for cross-ecosystem observation, forecasting, mitigation and adaptation strategies in a changing climate, together with corresponding adaptive basin management of systems that supply runoff to the coastal marine waters.

Changes in snow‐dominated streamflow quantity and timing following an extensive wildfire in British Columbia

AbstractThe length and frequency of extreme fire weather has increased across the globe in recent decades, with potential deleterious consequences to streamflow quantity, timing and quality. Changes in the hydrologic regime following wildfire can have substantial downstream consequences, affecting communities and ecosystems through flooding, erosion, loss of habitat and degraded water quality. While there are many studies that address post‐wildfire hydrology across the globe, there are few studies in the snow‐dominated regions. The 2017 Elephant Hill wildfire in south‐central BC burned across or adjacent to four watersheds with long‐term streamflow gauges providing a rare opportunity to evaluate hydrologic change. Several approaches were used to identify patterns of change following the wildfire, all of which suggest increased post‐fire flows. The before‐after‐control‐impact design showed significant increases in annual, spring and summer water yield from the small (49 km2) Arrowstone Creek watershed (30%, 21% and 86%, respectively). Significant increases in spring water yield were observed in the larger (5318 km2) Bonaparte River watershed (48%). Annual and summer water yield increased in the Bonaparte River (31% and 58%, respectively) but these changes were not statistically significant. In both the Bonaparte River and Arrowstone Creek, the onset of spring freshet (26 days earlier in both) was significantly advanced, however, the timing of maximum snowmelt discharge was significantly advanced (27 days earlier) only in Arrowstone Creek. Smaller changes were also observed in the reference watersheds; however, these were not statistically significant. The difference in results between the small and large watershed, as well as the effects of weather and watershed attributes, highlight the need for continued research into the relationships between wildfire and hydrologic regime across diverse landscapes.

Multi-mission virtual monitoring station for streamflow monitoring and hydrodynamic model calibration

A comprehensive understanding of the streamflow dynamic along the river system is one of the significant components for preserving biodiversity and ensuring sustainable ecological processes. Understanding this requirement, streamflow estimation using remote sensing (RS) has evolved as an alternate approach in the hydrologic literature during the last decade due to its extensive spatiotemporal coverage. However, the existing RS-based approaches have limited applications for deriving the continuous time series in tropical river reaches due to narrow water widths during the low flow and frequent cloud covers during the high flow periods. Therefore, this study proposed frameworks to establish a virtual monitoring station (VMS) where multi-mission satellites are being used to derive continuous streamflow time series. From the multi-mission satellites, the optical RS images are used to develop a Copula-based fusion (CFUS) model by integrating the Frank copula with the 30m × 1-day resolution synthetic Landsat images, derived from the fusion of 250m × 1-day resolution MODIS images and 30m × 16-days resolution Landsat images; whereas the water levels are retrieved from the altimeters to derive discharge using the rating curve. Additionally, the potential utility of the established VMS for hydrodynamic model (MIKE11-NAM-HD) calibration under data-scarce conditions is also demonstrated. Finally, a coupled RS-hydrodynamic (RS-HD) framework is also proposed to establish VMS both at semi-gauged and ungauged sections of the rivers. The efficacy of the aforementioned framework was tested along the lower Brahmani river reach. The results reveal that the advocated framework could perform satisfactorily with the Nash-Sutcliffe efficiency (NSE) and Kling-Gupta efficiency (KGE) ≥0.8. This approach has the potential to be upscaled to other river reaches as one of the next-generation hydrometry for daily streamflow monitoring using high-frequent observation of multi-mission satellites.

Factors influencing fish migration in one of the world's largest inland fisheries

Fish from Cambodia's Tonle Sap Lake are economically, culturally, and nutritionally significant for people in the Lower Mekong Basin, providing income, livelihoods, and protein. Fish in this system generally migrate toward upstream Mekong River in dry season and return in early wet season. However, drivers of fish migration from Tonle Sap Lake to the Mekong River are not well-understood. In this paper, we utilized Mixed Effects Random Forest to predict the catch weight of six fish species migrating from the Tonle Sap Lake to the Mekong River using precipitation, lunar cycle, and hydrologic conditions like river stage, streamflow, flow magnitude, and timing as predictors. As a surrogate for fish migration, we used daily fish catch weight from 2002 through 2008 at the bagnet, or Dai, fisheries along Tonle Sap River, a migration corridor connecting Tonle Sap Lake to the Mekong River. We found that migration of large fish was mainly cued by streamflow and flow magnitude, while smaller fish migrate depending on the combination of streamflow and flow timing. Streamflow less than average cumulative flow was the most important driver for migration of Pangasianodon hypophthalmus, and Cirrhinus microlepis. Migration of Cyclocheilichthys enoplos and Osteochilus melanopleurus was highly dependent on the number of low- and minimum-flow days. Cumulative flows, period of high flow and water level were the main predictors of the small mud-carp Henicorhynchus entmema's migration, while individuals of Labiobarbus leptocheilus migrated out of the Tonle Sap Lake depending on the number of days after 7-, 30-, and 90-day minimum flows. These results suggest that flow characteristics can be used to aid conservation and adaptive management of Cambodia's Dai fisheries.

Streamflow Prediction Using Complex Networks.

The reliable prediction of streamflow is crucial for various water resources, environmental, and ecosystem applications. The current study employs a complex networks-based approach for the prediction of streamflow. The approach consists of three major steps: (1) the formation of a network using streamflow time series; (2) the calculation of the clustering coefficient (CC) as a network measure; and (3) the use of a clustering coefficient-based nearest neighbor search procedure for streamflow prediction. For network construction, each timestep is considered as a node and the existence of link between any node pair is identified based on the difference (distance) between the streamflow values of the nodes. Different distance threshold values are used to identify the critical distance threshold to form the network. The complex networks-based approach is implemented for the prediction of daily streamflow at 142 stations in the contiguous United States. The prediction accuracy is quantified using three statistical measures: correlation coefficient (R), normalized root mean square error (NRMSE), and Nash-Sutcliffe efficiency (NSE). The influence of the number of neighbors on the prediction accuracy is also investigated. The results, obtained with the critical distance threshold, reveal that the clustering coefficients for the 142 stations range from 0.799 to 0.999. Overall, the prediction approach yields reasonably good results for all 142 stations, with R values ranging from 0.05 to 0.99, NRMSE values ranging from 0.1 to 12.3, and the NSE values ranging from -0.89 to 0.99. An attempt is also made to examine the relationship between prediction accuracy and the catchment characteristics/streamflow statistical properties (drainage area, mean flow, coefficient of variation of flow). The results suggest that the prediction accuracy does not have much of a relationship with the drainage area and the mean streamflow values, but with the coefficient of variation of flow. The outcomes from this study are certainly promising regarding the application of complex networks-based concepts for the prediction of streamflow (and other hydrologic) time series.

An ecohydrological approach to assess water provisioning and supporting ecosystem services in the Budhabalanga River Basin, India.

Rivers are vital and complex natural systems that provide a wide range of ecosystem services. This study presents a methodology for assessing the riverine provisioning and supporting ecosystem services, whose applicability has been demonstrated over the Budhabalanga River Basin of India. The Soil and Water Assessment Tool (SWAT) is used to generate streamflow time series at various ungauged sites, and then the streamflow is characterized for the evaluation of provisioning services. Further, the diversity and abundance of macroinvertebrates, along with the Lotic-invertebrate Index for Flow Evaluation (LIFE), is used to study the riverine supporting ecosystem services. The streams show intermittent behavior and strong seasonality for low flows, which limits the water availability, particularly during pre-monsoon season. The Baseflow Index (BFI) is greater than 0.6, indicating that groundwater contributes more than 60% of the total streamflow. Interestingly, despite the high BFI, the streams did not conform to the prevailing opinion that a greater baseflow contribution results in a later commencement of the low-flow period in the hydrological year. Furthermore, the study depicts significant variations in the diversity and abundance of the macroinvertebrates across the various sampling sites. However, the LIFE score across the sites remained consistent within a narrow range, i.e., 8 to 9, suggesting a steady supply of supporting ecosystem services. The results of the study can help the policymakers towards an informed decision making and the simplistic methodology proposed in this study can be replicated in other river basins for identifying vulnerable watersheds and prioritizing management actions.

Seasonal meteorological forcing controls runoff generation at multiple scales in a Mediterranean forested mountain catchment

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.

Streamflow seasonality in a snow-dwindling world.

Climate warming induces shifts from snow to rain in cold regions1, altering snowpack dynamics with consequent impacts on streamflow that raise challenges to many aspects of ecosystem services2-4. A straightforward conceptual model states that as the fraction of precipitation falling as snow (snowfall fraction) declines, less solid water is stored over the winter and both snowmelt and streamflow shift earlier in season. Yet the responses of streamflow patterns to shifts in snowfall fraction remain uncertain5-9. Here we show that as snowfall fraction declines, the timing of the centre of streamflow mass may be advanced or delayed. Our results, based on analysis of 1950-2020 streamflow measurements across 3,049 snow-affected catchments over the Northern Hemisphere, show that mean snowfall fraction modulates the seasonal response to reductions in snowfall fraction. Specifically, temporal changes in streamflow timing with declining snowfall fraction reveal a gradient from earlier streamflow in snow-rich catchments to delayed streamflow in less snowy catchments. Furthermore, interannual variability of streamflow timing and seasonal variation increase as snowfall fraction decreases across both space and time. Our findings revise the 'less snow equals earlier streamflow' heuristic and instead point towards a complex evolution of seasonal streamflow regimes in a snow-dwindling world.

Interbasin river mapping between Californian and Turkish rivers for climate change assessment

ABSTRACT A climate change assessment for streamflow availability of the selected rivers in Turkey is presented. Using an Index Basin Mapping (IBM) approach, climate change information is transferred across hydrologically similar rivers. This approach maps rivers of interest without downscaled climate information to index rivers where climate projections are available. Then, monthly perturbation ratios of index rivers, conveying projected climatic changes, are applied to the mapped rivers. Climate change effects on monthly streamflow availability and timing in eight selected rivers are evaluated under 20 scenarios produced from 10 General Circulation Models (GCMs) under medium and high emission cases. Results show that winter streamflow availability will increase due to more precipitation falling as rainfall rather than snowfall. Spring and summer streamflow availability will decrease due to reduced snowmelt runoff. Monthly streamflow variability will increase in all evaluated rivers. Out of eight selected rivers, the Çoruh, Yeşilırmak and Zamantı Rivers will be the most affected by climatic changes, with 14.6, 4.1, and 5.6% reductions in overall water availability under the high emission ensemble scenario. Overall water availability is projected increase in the Ceyhan and Göksu Rivers with climate change. Increased monthly streamflow variability can complicate water management in the region.

Estimating Increased Transient Water Storage with Increases in Beaver Dam Activity

Dam building by beaver (Castor spp.) slows water movement through montane valleys, increasing transient water storage and the diversity of residence times. In some cases, water storage created by beaver dam construction is correlated to changes in streamflow magnitude and timing. However, the total amount of additional surface and groundwater storage that beaver dams may create (and, thus, their maximum potential impact on streamflow) has not been contextualized in the water balance of larger river basins. We estimate the potential transient water storage increases that could be created at 5, 25, 50, and 100% of maximum modeled beaver dam capacity in the Bear River basin, USA, by adapting the height above nearest drainage (HAND) algorithm to spatially estimate surface water storage. Surface water storage estimates were combined with the MODFLOW groundwater model to estimate potential increases in groundwater storage throughout the basin. We tested four scenarios to estimate potential transient water storage increases resulting from the construction of 1179 to 34,897 beaver dams, and estimated surface water storage to range from 57.5 to 72.8 m3 per dam and groundwater storage to range from 182.2 to 313.3 m3 per dam. Overall, we estimate that beaver dam construction could increase transient water storage by up to 10.38 million m3 in the Bear River basin. We further contextualize beaver dam-related water storage increases with streamflow, reservoir, and snowpack volumes.

Evaluating the significance of wetland representation in isotope-enabled distributed hydrologic modeling in mesoscale Precambrian shield watersheds

Evaluation of regional-scale models against both streamflow and stable isotopes of water is a relatively new and evolving area of distributed hydrological modeling. While applications of isoWATFLOOD, an isotope-enabled distributed hydrological model are increasing, evaluation of the impacts of model structural features, such as wetland connectivity, on simulation of runoff components is still limited. Wetlands are important water bodies that impact streamflow generation and its spatial variation in Precambrian shield watersheds in parts of Canada and elsewhere. This study investigates the efficiency of isoWATFLOOD in simulating streamflow and instream δ18O and the effects of model percent connected wetlands (%CW) in a mesoscale (∼12,000 km2) Precambrian Shield watershed in Northeastern Ontario, Canada. Model calibration and validation are performed using daily streamflow for five hydrometric gauging stations and biweekly to monthly instream δ18O from 11 river sampling locations from across the watershed. Five model versions with different %CW are generated, calibrated, and validated separately. Results show all models can simulate the timing and pattern of streamflow and δ18O. %CW impacts simulated streamflow during both calibration and validation periods, calibrated parameter sets and simulated internal hydrological processes. Model performance varies spatially with no single value of %CW providing best performance across sub-catchments with 40 % CW providing best mean streamflow simulations in calibration. Increasing %CW (10 % to 40 %) results in higher contribution of lower soil zone water to streamflow that markedly improves baseflow simulation. This study builds on recent isotope-enabled hydrologic simulation capacity of isoWATFLOOD, contributing new examination of model representation of wetland connectivity, a critical component of Precambrian Shield watersheds and evidence of how stable isotopes can contribute to addressing the challenge of equifinality. Results suggest the need to develop model capability for representation of spatial variation in wetland connectivity to recognize this important aspect of heterogeneity in watersheds influencing hydrologic function.

Interpretable and explainable hybrid model for daily streamflow prediction based on multi-factor drivers.

Streamflow time series data typically exhibit nonlinear and nonstationary characteristics that complicate precise estimation. Recently, multifactorial machine learning (ML) models have been developed to enhance the performance of streamflow predictions. However, the lack of interpretability within these ML models raises concerns about their inner workings and reliability. This paper introduces an innovative hybrid architecture, the TCN-LSTM-Multihead-Attention model, which combines two layers of temporal convolutional networks (TCN) followed by one layer of long short-term memory (LSTM) units, integrated with a Multihead-Attention mechanism for predicting streamflow with streamflow causation-driven prediction samples (RCDP), employing local and global interpretability studies through Shapley values and partial dependency analysis. The find_peaks method was used to identify peak flow events in the test dataset, validating the model's generality and uncovering the physical causative patterns of streamflow. The results show that (1) compared to the LSTM model with the same hyperparameter settings, the proposed TCN-LSTM-Multihead-Attention hybrid model increased the R2 by 52.9%, 2.5%, 43.1%, and 10.7% respectively at four stations in the test set predictions using RCDP samples. Moreover, comparing the prediction results of the hybrid model under different samples in Hengshan station, the R2 for RCDP increased by 5.06% and 1.22% compared to streamflow autoregressive prediction samples (RAP) and meteorological-soil volumetric water content coupled autoregressive prediction samples (MCSAP) respectively. (2) Historical streamflow data from the preceding 3days predominantly influences predictions due to strong autocorrelation, with flow quantity (Q) typically emerging as the most significant feature alongside precipitation (P), surface soil moisture (SSM), and adjacent station flow data. (3) During periods of low and normal flow, historical data remains the most crucial factor; however, during flood periods, the roles of upstream inflow and precipitation become significantly more pronounced. This model facilitates the identification and quantification of various hydrodynamic impacts on flow predictions, including upstream flood propagation, precipitation, and soil moisture conditions. It also elucidates the model's nonlinear relationships and threshold responses, thereby enhancing the interpretability and reliability of streamflow predictions.

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.

Do non-linearity and non-Gaussianity truly matter in streamflow forecasting? A comparative study between PAR(p) and vine copula for Brazilian streamflow time series.

This study evaluates the joint impact of non-linearity and non-Gaussianity on predictive performance in 23 Brazilian monthly streamflow time series from 1931 to 2022. We consider point and interval forecasting, employing a PAR(p) model and comparing it with the periodic vine copula model. Results indicate that the Gaussian hypothesis assumed by PAR(p) is unsuitable; gamma and log-normal distributions prove more appropriate and crucial for constructing accurate confidence intervals. This is primarily due to the assumption of the Gaussian distribution, which can lead to the generation of confidence intervals with negative values. Analyzing the estimated copula models, we observed a prevalence of the bivariate Normal copula, indicating that linear dynamic dependence is frequent, and the Rotated Gumbel 180°, which exhibits lower tail dependence. Overall, the temporal dynamics are predominantly shaped by combining these two types of effects. In point forecasting, both models show similar behavior in the estimation set, with slight advantages for the copula model. The copula model performs better during the out-of-sample analysis, particularly for certain power plants. In interval forecasting, the copula model exhibits pronounced superiority, offering a better estimation of quantiles. Consistently demonstrating proficiency in constructing reliable and accurate intervals, the copula model reveals a notable advantage over the PAR(p) model in interval forecasting.

Developing Non-Stationary Frequency Relationships for Greater Pamba River basin, Kerala India incorporating dominant climatic precursors

Abstract. Global climate changes significantly contribute to increased frequency of hydrologic extremes. This significantly underestimates the hydrologic design parameters, bringing of hydro systems to increased failure risk. In order to address this concern, the current practice of development of hydrologic frequency tools need to be updated accounting for non-stationarity. This study first considered a diverse set of statistical tests to examine the trend, change points, non-stationarity and randomness of streamflow, rainfall and temperature time series of scales ranging from daily to annual. The annual maxima time series indicated non stationarity against the stationary behaviour of daily series of hydro-meteorological datasets of the basin. Subsequently, this study developed the Temperature Duration Frequency (TDF), Rainfall Intensity Duration Frequency (IDF) and Flood Frequency (FF) curves of Greater Pamba river basin in Kerala India, the part of which was most severely affected by the near century return period flood event of 2018. The analysis was performed for a multitude of combinations of variations in distribution parameters with time and climatic drivers as physical covariates in the extreme value formulations. The study proposed a novel wavelet coherence (WC) based driver selection of most dominant combination of climatic precursors in developing FF and IDF relations of three locations of Kalloopara, Malakkara and Thumpamon and TDF curve of Kuttanad region in the basin, considering data of 1985–2015 period. The proposed WC framework considers bi-multi-and partial effects of climatic oscillations (COs) like El Niño Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), Pacific Decadal Oscillation (PDO) and North Atlantic Oscillation (NAO) in identifying potential drivers. The different WC formulations captured in-phase relationships of streamflows and rainfall with COs at intra-annual, annual and inter annual scales up to 4 years. The methods showed that addition of climatic precursors improved the NS estimates of flood and rainfall quantiles by more accurately capturing the magnitudes of extreme streamflows and rainfalls of 2018, 2021 than the time covariate formulations. However, the role of COs on extreme temperature is not found to be influential in developing TDF relationships, which needs further investigation.