Subjective decisions in hydrologic model calibration can have drastic impacts on our understanding of basin processes and simulated fluxes. Here, we present a multicase calibration approach to determine three pillars of an appropriate hydrological model configuration, i.e. calibration data length, spin-up period, and spatial resolution, using a spatially distributed meso-scale hydrological model (mHM) together with a dynamically dimensioned search (DDS) algorithm and Nash-Sutcliffe efficiency (NSE) for the Moselle basin. The results show that a 10-year calibration data length, 2-year spin-up period, and 4-km model resolution are appropriate for the Moselle basin to reduce the computational burden while simulating streamflow with a decent performance. Although the calibration data length and spatial resolution are related to the extent and quality of the data, and the spin-up period is basin dependent, analysing the combined effects further allowed us to understand the interactions of these three usually overlooked pillars in the mHM configuration.
ABSTRACT The popular approach to select a suitable distribution to characterize extreme rainfall events relies on the assessment of its descriptive performance. This study examines an alternative approach to this task that evaluates, in addition to the descriptive performance of the models, their performance in estimating out-of-sample events (predictive performance). With a numerical experiment and a study case in São Paulo state, Brazil, we evaluated the adequacy of seven probability distributions widely used in hydrological analysis to characterize extreme events in the region and compared the selection process of both popular and altenative frameworks. The results indicate that (1) the popular approach is not capable of selecting distributions with good predictive performance and (2) combining different predictive and descriptive tests can improve the reliability of extreme event prediction. The proposed framework allowed the assessment of model suitability from a regional perspective, identifying the Generalized Extreme Value (GEV) distribution as the most adequate to characterize extreme rainfall events in the region.
The standard method for characterizing the variability of overland flow focuses primarily on a single scale, usually the smallest scale available (i.e. the highest resolution). However, the extremes of overland flow are generally variable over a wide range of scales. Thus, the smallest scale has no specific hydrological significance, and a scale-independent characterization is more physically relevant. This study investigated the spatial variability of overland flows and evaluated the hydrological performance of the Nature-Based Solutions scenarios in terms of a possible improvement of the morphological functioning of the catchment at several scales.
We are very grateful to the authors Aksoy and Cavus for their constructive contributions to the discussion of our earlier paper "Drought assessment in a south Mediterranean transboundary catchment." Our paper focused on the variability of rainfall and drought patterns in the Medjerda catchment in Tunisia and analysed the underlying causes resulting in these changes. Scope exists to improve the integration of the space-time variability in the drought assessment methodology. Different drought indices and statistical methods are employed in the study. The discussion of the article concentrates almost exclusively on the use of the Standardized Precipitation Index (SPI) and some mistyped equations. In this reply, we clarify the problematic points of misunderstanding and discuss the classification of droughts based on the SPI values.
This study investigated the lake–aquifer hydraulic interactions in Lake Urmia (LU) as the second largest hypersaline lake in the world. Due to the scarcity of hydrogeological data required for modelling, a method based on Darcy’s Law and lake water budget was used to quantify the lake–aquifer interaction. Long-term ground- and satellite-based hydrological datasets over the time frame 2001–2019 were used. Results indicate that the groundwater flux between LU and the aquifers controls 18.74 ± 1.67% of the lake’s water storage. While 10 out of 14 adjacent aquifers recharge LU at a rate of less than 180 m3/m.month, one phreatic aquifer recharges the LU up to 1400 m3/m.month. Two aquifers are recharged from the seawater of LU at a rate of 110–640 m3/m.month. The results of this study lead to a holistic understanding of the hydraulic interaction of LU–aquifers, which is essential for the sustainable management of both settings.
Monitoring and predicting river floods have always been of concern for hydrologists. Anomalies in flow patterns alert us of upcoming events. The effect of such anomalies on the multifractality and nonlinear dynamics of the river flow is investigated in this research. The river flow of the River Trent in England from 2018 to 2019 was dissected using the multifractal, power spectrum, and phase space reconstruction techniques. Results show a reduction in multifractality strength before the anomalies in the time series. Moreover, receding trajectories from the attractor in the phase space and augmentation of the power spectrum confirmed a decline in flow multifractality due to anomalies. A compliance comparison of multifractality reduction before anomalies in river flow leads to predicting river floods approximately 12 days in advance. Furthermore, if the multifractal strength does not return to its equilibrium, this is a warning that floods are expected in the near future (less than 10 days).
Hydrological model calibration is a quintessential step in model development, and the time scale of calibration depends on the application. However, the implications of choice of time scale of calibration have not been explored extensively. Here, we evaluate the effect of the time scale of calibration on model sensitivity, best parameter ranges, and predictive uncertainty for three river basins using the Soil and Water Assessment Tool (SWAT) model. Multiple models were set up for three different catchments from southern India. Our results showed that the sensitivity of the parameters, best parameter ranges, and model performance are conditioned on the time scale of calibration. The models calibrated at coarser time scales marginally outperformed the models calibrated at fine time scale in terms of Nash-Sutcliffe efficiency and percentage bias. Transfer of parameters across scales (both from coarse to fine and from fine to coarse) have a general tendency to worsen the model performance in all three catchments, with few exceptions.
Groundwater is vital in Chennai Metropolitan Area, and thus its sustainable management is essential. This study aims to estimate groundwater recharge in the Chennai River basin (CRB) using an empirical method, a rainfall infiltration factor (RIF) method, a geographical information system (GIS)-distributed model, and the water table fluctuation (WTF) method. The average recharge estimates were 196 mm/year (empirical), 127 mm/year (WTF), and 122 mm/year (RIF). These results agree with the estimate of the GIS-distribution model, where 59% of the area has high recharge (>100 mm/year). The effective recharge from rainfall was around 10% with the RIF and WTF methods and 16% using the empirical method. All the estimates were statistically significant with confidence levels of over 98%, using a 95% confidence interval. The WTF method (with a 99.9% confidence level) was the most reliable of the four approaches. However, we suggest using multiple methods to cross-check the results and that measuring rainfall and water level is necessary for accurate recharge estimates in the future.
ABSTRACT This study assesses bias error of rainfall from climate models and related error propagation effects to simulated streamflow in the Gidabo sub-basin, Ethiopia. Rainfall is obtained from a combination of four global and regional climate models (GCM-RCMs), and streamflow is simulated by means of the Hydrologiska Byråns Vattenbalansavdelning (HBV-96) rainfall-runoff model. Five bias correction methods were tested to reduce the rainfall bias. To assess the effects of rainfall bias error propagation, percent bias (PBIAS), difference in coefficient of variation (CV), and 10th and 90th percentile indicators were applied. Findings indicate that the bias of the uncorrected rainfall caused large errors in simulated streamflow. All five bias correction methods improved the HBV-96 model performance in terms of capturing the observed streamflow. Overall, the findings of this study indicate that the magnitude of the error propagation varies subject to the selected performance indicator, bias correction method and climate model.
Regional flood frequency analyses are highly affected by the number of gauged catchments and lengths of observation periods at the individual gauges. Therefore, information is unevenly distributed in the region of interest. In particular, the occurrence of single extreme floods that are observed or not observed at some gauges, depending on the observation period, may have a large impact on the regionalization. We evaluated the impact of the sample length on the regionalization of a type-specific statistical mixture-model. In a case study it is shown how the regionalization error can be reduced by more than 50% if the sample sizes increase. We compare three approaches to handling this stochastic uncertainty in regionalization. The alignment of the statistical distribution parameters to consider the impact of extreme floods proved to be most beneficial when aiming to obtain homogeneous regionalized flood quantiles for hydrologically similar regions in a region with heterogeneous observation periods.
The present study aims to improve the efficacy of water budget (WB) estimations from various hydrological data products, by (1) evaluating the uncertainties of hydrological data products, (2) merging four precipitation and six evapotranspiration products using their error variances, and (3) employing the constrained Kalman filter (CKF) method to distribute residual errors among water budget components based on their relative uncertainties. The results show that applying bias correction before the merging process improved estimations of precipitation products with decreasing root mean square error (RMSE), except Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN). Variable Infiltration Capacity (VIC) and bias-corrected Climate Prediction Center Morphing Technique (CMORPH) products outperformed other evapotranspiration and bias-corrected precipitation products, respectively, in terms of mean merging weights. The terrestrial water storage change is the primary reason for non-closure errors, mainly caused by the coarse resolution of Gravity Recovery and Climate Experiment (GRACE). The CKF results were insensitive to variations in uncertainties of runoff. Precipitation derived from the CKF was the best precipitation output, with the highest correlation coefficient (CC) and smallest root mean square deviation (RMSD).
Adequate and consistent monsoonal rainfall is essential for the economic and societal well-being of India. Although many past studies analysed trends in Indian summer monsoon rainfall (ISMR) to capture any changes, the nature and consistency (stability) of those trends remained unexplored. Therefore, this study investigates the monotonic, non-monotonic, and trendless nature as well as the stability of ISMR trends from 1897 to 2016 using both graphical and statistical approaches for different meteorological sub-divisions of India. Also, the trend stability is examined considering high, medium, and low clusters of ISMR. Results show a predominance of monotonic trends in the long-term series, and maximum eastern sub-divisions exhibit a decreasing trend and vice versa. However, further analysis revealed instabilities in the ISMR trend and its respective clusters. The possible mechanisms behind ISMR instabilities and the spatio-temporal changes are discussed, along with certain methodological challenges, which might aid in formulating climate action plans.
This paper explores how crowdsourced social media data complements urban flood modelling to improve model performance and achieve a better classification of impacts. In addition to georeferencing flood impacts, Twitter allows monitoring the events in terms of hazards and impacts, and YouTube facilitates a retrospective analysis from audiovisual data. The analysis of 2800 tweets collected during four storm events and of almost 900 videos of the recent history of the basin, together with the implementation of a high-resolution model, contributed to the expansion of the capacity to represent the temporal and spatial scales of the problem. The complementation of crowdsourced social media data and urban modelling enhances the understanding of the flood dynamics, thus offering a framework of greater certainty for the generation of flood risk management products.
Moments of rainfall spatial variability, which quantify how flood response time scales are affected when spatially variable rainfall is considered, compared to when rainfall is spatially uniform, have been suggested as a useful tool for forecasters to guide their choice between lumped or distributed rainfall information for runoff modelling. However, the approaches used to evaluate the validity of moments suffer from limitations. Hence, we adopt a novel approach for their evaluation by comparing moments to the relationship between observed hydrograph characteristics generated by spatially variable and by uniform rainfall events in the same catchment. We further investigate the usefulness of moments by testing whether the performance of a lumped hydrological model for events classified by moments as spatially variable is lower than for uniform events. Results confirmed that moments can identify spatially variable events and characterize differences in hydrograph features compared to uniform events, providing a useful tool for forecasters.
In this study the capability of two advanced hybrid artificial intelligence-based methods is investigated in modelling meteorological droughts based on the standardized precipitation index (SPI) for various time windows. The outcomes are compared with adaptive neuro-fuzzy inference system (ANFIS), support vector regression (SVR), autoregressive integrated moving average (ARIMA) and seasonal autoregressive integrated moving average (SARIMA). The best-fitted distribution functions are found to vary with respect to stations and time windows. For Canakkale, Istanbul and Tekirdag stations, the correlation coefficients (CC) of hybridized method of ANFIS with gray wolf optimization (ANFIS-GWO) and ARIMA are in the range of 0.88-0.94, 0.88-0.96 and 0.86-0.94, respectively. The performance index also showed that the ANFIS-GWO provides superior accuracy in modelling droughts, with the minimum value (PI = 0.78) for SPI12 of Canakkale station. Forward-chaining cross-validation and the P value of the Chi-squared test also confirm that ANFIS-GWO is the superior model, in which there is not a significant difference between the trend of the predicted categories and that of the real categories for all stations and SPIs.
Topographic LIDAR can be used to estimate elevation values for dry areas down to the river water level during the extraction of river cross-sections (XS). However, LIDAR cannot accurately predict the submerged topography, which causes uncertainty in river XS area estimation. This uncertainty affects the channel water level and flood inundation depth estimation in in situ sparse data. Therefore, an alternative approach is presented to estimate unknown submerged topography (UST) using topographic LIDAR. The one dimension/two dimension Hydrologic Engineering Center River Analysis System (1D/2D HEC-RAS) model is used to simulate the estimated river XS with the help of in situ river water level and flow data which is later validated using in situ data. The results show that the proposed approach accurately estimates water level (error >0.5 m), channel flow areas, and floodplain water depths. Notably, the extent of the estimated floodplain overflow by UST models was in 94% agreement with the real XS.
Defining river networks from a digital elevation model (DEM) is a critical step in many hydrological studies. Most methods rely on analyses of flow accumulation based on a single drainage area threshold value. Landscape heterogeneity hinders the definition of drainage pixels by exceeding a threshold. In this study, we propose a method that automatically accounts for variable drainage density in the definition of river networks from relatively coarse DEMs by using the topographic position index (TPI) as an auxiliary variable, therefore calling it the Topographic Position-based Stream definition (TPS). We tested the TPS method with the 90 m resolution MERIT-DEM (Multi-Error-Removed Improved-Terrain DEM) in Brazil, applying it to three river basins with contrasting landforms and climates. The method improved drainage representation in all analysed regions. It successfully represented stream networks with intra-basin variable drainage density by using a single DEM and three parameters, showing potential for hydrological applications at multiple spatial scales.
Monitoring terrestrial water storage anomalies (TWSA) is essential for better understanding the influences of climatic variability on hydrological cycles. Here we use Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-on satellite data and meteorological observations to analyse the inter-annual trends of TWSA across the Qinghai-Tibet Plateau (QTP) during 2003-2020 and investigate the relationships between climatic variability and annual changes in TWSA. Results indicate that TWSA across the QTP generally decreased, at a rate of -0.5 +/- 1.4 mm/year during 2003-2020, which mainly arises from the coupled effects of Potential evapotranspiration (PET) and precipitation. In terms of the effects of climatic variability, annual changes in TWSA show a negative correlation coefficient (r = -0.76) with PET, which is greater than that with precipitation (r = 0.67). The cross-wavelet transformation analysis revealed that the evolution of TWSA across the QTP is closely associated with the Atlantic Multidecadal Oscillation. Our conclusions can help decision makers to formulate appropriate policies for the assessment and management of water resources over the QTP.
ABSTRACT We congratulate Heal et al. for initiating an important discussion on how to broaden the scope of the water–energy–food nexus. We agree that more explicit inclusion of water quality into the nexus is an important step forward. At the same time, water quality is itself an indicator of e.g. ecosystem services and biodiversity, and improvement of water quality comes with a cost in terms of resource consumption that is typically not included in models studying the water–energy–food nexus. We already see hesitation in using the nexus for policy development, and further complexity may be an additional barrier to its practical implementation. So, while the consideration of water quality is indeed important for the nexus, it also suggests that perhaps it is necessary to consider more local contexts than striving for one global framing for analysis of the water–energy–food nexus.
This study presents an extension of an existing performance-based weighting system to combine statistical streamflow estimates for ungauged basins. Three statistical methods, namely the drainage area ratio (DAR) method, the standardization with mean (SM) method, and the inverse similarity weighting (ISW) method, have been applied to two sub-basins of the Euphrates basin in Turkey. To improve the effectiveness of the daily streamflow estimation in ungauged basins, ensemble approaches that combine the results of two or three statistical methods have been proposed based on a weighting system using various performance measures. The overall results demonstrate that the proposed ensemble approaches with the performance-based weighting system are beneficial to enhance the performance of the statistical methods in estimating daily streamflow at ungauged basins.
