Evaluation Of Hydrological Models Research Articles

A short history of philosophies of hydrological model evaluation and hypothesis testing

A short history of philosophies of hydrological model evaluation and hypothesis testing

Evaluation of hydrological models for streamflow prediction: a case study of the Mille River, Lower Awash Basin, Ethiopia

ABSTRACT Accurate prediction of stream flow is essential for effective water resource management, particularly in regions with complex hydrological dynamics. The Mille River, which flows through this basin and has a watershed area of 4,862.3 km², plays a crucial role in the region's water resources. However, selecting the most suitable hydrological model for this task remains a challenge due to varying model performances. This study addresses the problem of identifying the most effective hydrological model for predicting stream flow in the Mille River by evaluating three models: Soil and Water Assessment Tool (SWAT), Hydrologiska Byråns Vattenbalansavdelning (HBV-IHMS), and Artificial Neural Network (ANN). The objectives are to compare these models’ abilities to simulate daily stream flow and to determine their accuracy. The significance of this evaluation lies to guide water management decisions by identifying which model provides the most reliable predictions. The methodology involves using performance metrics, Nash-Sutcliffe Efficiency (NSE) values, to assess the accuracy of each model. The results show that SWAT achieved the highest NSE values of 0.81 and 0.82, indicating the best performance in simulating stream flow. ANN followed with NSE values of 0.79 and 0.81, while HBV-IHMS had lower NSE values of 0.73 and 0.74. Although all models demonstrated potential, SWAT and ANN outperformed HBV-IHMS in capturing the complex hydrological processes of the Mille River. The findings from this study indicate that the choice of model significantly impacts the accuracy and reliability of streamflow predictions. The SWAT and ANN models show strong potential for use in operational streamflow forecasting in the Mille River and similar basins. In conclusion, further research is recommended to refine model calibration and validation processes. Incorporating additional data sources, such as remote sensing and climate projections, could further enhance model accuracy and provide a more comprehensive understanding of the region's hydrological dynamics.

Evaluation of hydrological models at gauged and ungauged basins using machine learning-based limits-of-acceptability and hydrological signatures

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

A brief history of information and disinformation in hydrological data and the impact on the evaluation of hydrological models

ABSTRACT This paper considers what we know about the potential for disinformation in hydrological data when used for the evaluation of hydrological models. This will generally arise from epistemic uncertainties associated with hydrological observations, particularly from nonstationary or extrapolated rating curves for discharges, and poor rainfall and snowmelt information when interpolated over basin areas. Approaches based on information theory are not well suited to consideration of such epistemic uncertainties in model evaluation and an alternative approach based on setting limits of acceptability independent of any model runs is suggested. This allows for both the rejection of all models tried, and for acceptability of models across different model structures and parameter sets. The paper concludes with some suggestions for future research on defining disinformative data for both point and spatial observables, studying model failures, and defining new observations with a view to having the greatest impact on reducing model uncertainties.

Evaluation of hydrological models on small mountainous catchments: impact of the meteorological forcings

Abstract. Hydrological modelling of small mountainous catchments is particularly challenging because of the high spatio-temporal resolution required for the meteorological forcings. In situ measurements of precipitation are typically scarce in these remote areas, particularly at high elevations. Precipitation reanalyses propose different alternative forcings for the simulation of streamflow using hydrological models. In this paper, we evaluate the performances of two hydrological models representing some of the key processes for small mountainous catchments (< 300 km2), using different meteorological products with a fine spatial and temporal resolution. The evaluation is performed on 55 small catchments of the northern French Alps. While the simulated streamflows are adequately reproduced for most of the configurations, these evaluations emphasize the added value of radar measurements, in particular for the reproduction of flood events. However, these better performances are only obtained because the hydrological models correct the underestimations of accumulated amounts (e.g. annual) from the radar data in high-elevation areas.

Tradeoffs Between Temporal and Spatial Pattern Calibration and Their Impacts on Robustness and Transferability of Hydrologic Model Parameters to Ungauged Basins

AbstractOptimization of spatially consistent parameter fields is believed to increase the robustness of parameter estimation and its transferability to ungauged basins. The current paper extends previous multi‐objective and transferability studies by exploring the value of both multi‐basin and spatial pattern calibration of distributed hydrologic models as compared to single‐basin and single‐objective model calibrations, with respect to tradeoffs, performance and transferability. The mesoscale Hydrological Model (mHM) is used across six large central European basins. Model simulations are evaluated against streamflow observations at the basin outlets and remotely sensed evapotranspiration patterns. Several model validation experiments are performed through combinations of single‐ (temporal evaluation through discharge) and multi‐objective (temporal and spatial evaluation through discharge and spatial evapotranspiration patterns) calibrations with holdout experiments saving alternating basins for model evaluation. The study shows that there are very minimal tradeoffs between spatial and temporal performance objectives and that a joint calibration of multiple basins using multiple objective functions provides the most robust estimations of parameter fields that perform better when transferred to ungauged basins. The study indicates that particularly the multi‐basin calibration approach is key for robust parametrizations, and that the addition of an objective function tailored for matching spatial patterns of ET fields alters the spatial parameter fields while significantly improving the spatial pattern performance without any tradeoffs with discharge performance. In light of model equifinality, the minimal tradeoff between spatial and temporal performance shows that adding spatial pattern evaluation to the traditional temporal evaluation of hydrological models can assist in identifying optimal parameter sets.

When best is the enemy of good – critical evaluation of performance criteria in hydrological models

Abstract. Performance criteria play a key role in the calibration and evaluation of hydrological models and have been extensively developed and studied, but some of the most used criteria still have unknown pitfalls. This study set out to examine counterbalancing errors, which are inherent to the Kling–Gupta efficiency (KGE) and its variants. A total of nine performance criteria – including the KGE and its variants, as well as the Nash–Sutcliffe efficiency (NSE) and the modified index of agreement (d1) – were analysed using synthetic time series and a real case study. Results showed that, when assessing a simulation, the score of the KGE and some of its variants can be increased by concurrent overestimation and underestimation of discharge. These counterbalancing errors may favour bias and variability parameters, therefore preserving an overall high score of the performance criteria. As bias and variability parameters generally account for two-thirds of the weight in the equation of performance criteria such as the KGE, this can lead to an overall higher criterion score without being associated with an increase in model relevance. We recommend using (i) performance criteria that are not or less prone to counterbalancing errors (d1, modified KGE, non-parametric KGE, diagnostic efficiency) and/or (ii) scaling factors in the equation to reduce the influence of relative parameters.

Guidance on large scale hydrologic model calibration with isotope tracers

Standard hydrologic model evaluation and calibration approaches focus on the accurate simulation of streamflow, disregarding internal process simulations. Stable isotope tracers can provide additional information on water sources, and process flux and storage, which can be used to inform model calibration. This study assesses the added value of isotope data in comparison to current best-practice flow-only calibration methods and evaluates the merits and limitations of isotope simulation performance metrics for the purposes of hydrological model calibration. Following several years of regular isotope sampling and measurement, an isotope-enabled process-based hydrologic model was tested on a large watershed in western Canada (Athabasca River), which allowed model calibration using global sensitivity analyses, Monte Carlo simulations, and multi-objective optimizations. Isotope tracer data were found to improve both process and streamflow component identifiability and produced some minor improvement in individual parameter value identifiability. Calibrating to optimize both flow and isotope simulation performance produced better flow simulation ensembles, with improved observation capture and validation performance, relative to calibrating to optimize flow simulations alone. Using an isotope simulation performance metric which includes timing error as a secondary optimization objective led to more robust streamflow modeling, even in mesoscale watersheds with limited isotope observation datasets.

Hydrological Model Evaluation of Ground, GPM IMERG, and CHIRPS precipitation data for Shabelle Basin in Ethiopia

Consistent and accurate precipitation measurement is a fundamental input component of the hydrological model. However, most developing countries, including Ethiopia, lack consistent and precise precipitation measurements. Nevertheless, satellitebased precipitation data may play a crucial role in bridging the data gap and providing precipitation inputs for rainfall-runoff models in regions with limited ground data. The study compares the ground precipitation data from the Ethiopian meteorological agency and SWALIM to the satellite-based GPM_IMERG and CHIRPS precipitation products. HECHMS software was used to model rainfall-runoff simulation, and the study area’s spatial characteristics and model development were processed with ArcGIS and HEC- GeoHMS. Since the data on hand is enough and fit for continuous event assessment, the deficit and constant loss method, a continuous event-based loss method, was tested, and found good model performance results. The onhand 14 years of precipitation and discharge data were divided into seven years for calibration and seven years for validation. Using Nash–Sutcliffe efficiency (ENS), coefficient of determination (R2), Root Mean Square Error (RMSE), and percentage bias objective functions, the performance of the satellite precipitation was evaluated. During calibration periods (Jan 2013 to Dec 2009), the model performance showed ENS values of 0.65, R2 values of 0.68, RMSE values of 0.6, a percent bias of 0.58% for the metrological stations, ENS values of 0.61, R2 values of 0.64, RMSE values of 0.6, a percent bias of 5.38% for CHIRPS, and ENS values of 0.63, R2 values of 0.66, RMSE values of 0.7, and a percent bias of 18.9% for GMP_IMERG satellite precipitation products. During validation periods (Jan 2010 to Dec 2016), the model performance showed ENS values of 0.75, R2 values of 0.78, RMSE values of 0.6, a percent bias of -16.9% for the metrological stations, ENS values of 0.71, R2 values of 0.74, RMSE values of 0.6, a percent bias of -22.12% for CHIRPS2, and ENS values of 0.76, R2 values of 0.76, RMSE values of 0.5, and a percent bias of -0.03% for GPM_IMERG satellite precipitation products. The study showed that the HEC-HMS model performed well and gave very good results for the hydrological model for the CHIRPS and GPM_IMERG rainfall products. The study also indicates that the model outperformed well during the validation period. Overall, the study found that the simulated GPM_IMERG product gave better results than the simulated CHIRPS product.

Evaluating the accuracy of gridded water resources reanalysis and evapotranspiration products for assessing water security in poorly gauged basins

Abstract. Achieving water security in poorly gauged basins is critically hindered by a lack of in situ river discharge data to assess past, current, and future evolution of water resources. To overcome this challenge, there has been a shift toward the use of freely available satellite and reanalysis data products. However, due to inherent bias and uncertainty, these secondary sources require careful evaluation to ascertain their performance before being applied in poorly gauged basins. The objectives of this study were to evaluate river discharge and evapotranspiration estimates from eight gridded water resources reanalysis (WRR), six satellite-based evapotranspiration (ET) products, and ET estimates derived from complimentary relationship (CR–ET) across eight river basins located in Central–West Africa. Results highlight strengths and weaknesses of the different WRR in simulating discharge dynamics and ET across the basins. Likewise, satellite-based products also show some strength and weaknesses in simulating monthly ET. Our results further revealed that the performance of the different models in simulating river discharge and evapotranspiration is strongly influenced by model structure, input data, and spatial resolution. Considering all hydrological model evaluation criteria, FLDAS-Noah, Lisflood, AWRAL, and Terra were among the best performing WRR products while for ET estimates, FLDAS-Noah, Terra, GLEAM3.5a and 3.5b, and PMLV2 outperformed the rest of the products. Given the plethora of WRR and ET products available, it is imperative to evaluate their performance in representative gauged basins to identify products that can be applied in each region. However, the choice of a particular product will depend on the application and user requirements. Taking this together, results from this study suggest that gridded WRR and ET products are a useful source of data for assessing water security in poorly gauged basins.

Corrected ERA5 Precipitation by Machine Learning Significantly Improved Flow Simulations for the Third Pole Basins

Abstract Precipitation is one of the most important atmospheric inputs to hydrological models. However, existing precipitation datasets for the Third Pole (TP) basins show large discrepancies in precipitation magnitudes and spatiotemporal patterns, which poses a great challenge to hydrological simulations in the TP basins. In this study, a gridded (10 km × 10 km) daily precipitation dataset is constructed through a random-forest-based machine learning algorithm (RF algorithm) correction of the ERA5 precipitation estimates based on 940 gauges in 11 upper basins of TP for 1951–2020. The dataset is evaluated by gauge observations at point scale and is inversely evaluated by the Variable Infiltration Capacity (VIC) hydrological model linked with a glacier melt algorithm (VIC-Glacier). The corrected ERA5 (ERA5_cor) agrees well with gauge observations after eliminating the severe overestimation in the original ERA5 precipitation. The corrections greatly reduce the original ERA5 precipitation estimates by 10%–50% in 11 basins of the TP and present more details on precipitation spatial variability. The inverse hydrological model evaluation demonstrates the accuracy and rationality, and we provide an updated estimate of runoff components contribution to total runoff in seven upper basins in the TP based on the VIC-Glacier model simulations with the ERA5_cor precipitation. This study provides good precipitation estimates with high spatiotemporal resolution for 11 upper basins in the TP, which are expected to facilitate the hydrological modeling and prediction studies in this high mountainous region. Significance Statement The Third Pole (TP) is the source of water to the people living in the areas downstream. Precipitation is the key driver of the terrestrial hydrological cycle and the most important atmospheric input to land surface hydrological models. However, none of the current precipitation data are equally good for all the TP basins because of high variabilities in their magnitudes and spatiotemporal patterns, posing a great challenge to the hydrological simulation. Therefore, in this study, a gridded daily precipitation dataset (10 km × 10 km) is reconstructed through a random-forest-based machine learning algorithm correction of ERA5 precipitation estimates based on 940 gauges in 11 TP basins for 1951–2020. The data eliminate the severe overestimation of original ERA5 precipitation estimates and present more reasonable spatial variability, and also exhibit a high potential for hydrological application in the TP basins. This study provides long-term precipitation data for climate and hydrological studies and a reference for deriving precipitation in high mountainous regions with complex terrain and limited observations.

Spatial pattern‐based performance evaluation and uncertainty analysis of a distributed hydrological model

AbstractAdvancement in data acquisition techniques (e.g., satellites) has provided enormous opportunities to improve the predictions of distributed hydrological models by incorporating new types of information. Moreover, such advancements have paved the way for the spatial scale evaluation rather than the conventional temporal evaluation of models for more accurate water management by considering the spatial variability. The present study focuses on the spatial pattern‐oriented model evaluation by intercomparing the satellite remote sensing data‐driven patterns of actual evapotranspiration (AET) and soil moisture with simulated patterns from a hydrological model for the 19 276 km2 Subarnarekha basin in Eastern India. Pattern comparison is carried out by incorporating three spatial performance metrics as diagnostic tools, that is, joint empirical orthogonal functions (EOF), fractional skill score (FSS), and spatial efficiency (SPAEF). Moreover, the present study made a novel contribution to spatial pattern evaluation by incorporating the uncertainty associated with the patterns through the “Quantile Regression’ technique”. The EOF analysis indicated a significant similarity between simulated and observed patterns. High scores for both the FSS and SPAEF metrics also revealed similarities between simulated and observed spatial patterns. Overall, the results show an acceptable spatial pattern performance of the hydrological model mimicking the observed spatial patterns across time. The uncertainty analysis represents reasonable predictive ranges, with a 95 per cent prediction uncertainty (PPU) band enclosing 80%–95% of AET and 79%–92% of the soil moisture observations. The present study demonstrated a framework for improving the basin's overall water balance analysis by incorporating a spatial pattern‐based evaluation of hydrological models.

On the Calibration of Spatially Distributed Hydrologic Models for Poorly Gauged Basins: Exploiting Information from Streamflow Signatures and Remote Sensing-Based Evapotranspiration Data

Spatially distributed hydrologic models are useful for understanding the water balance dynamics of catchments under changing conditions, thereby providing important information for water resource management and decision making. However, in poorly gauged basins, the absence of reliable and overlapping in situ hydro-meteorological data makes the calibration and evaluation of such models quite challenging. Here, we explored the potential of using streamflow signatures extracted from historical (not current) streamflow data, along with current remote sensing-based evapotranspiration data, to constrain the parameters of a spatially distributed Soil and Water Assessment Tool (SWAT) model of the Mara River Basin (Kenya/Tanzania) that is forced by satellite-based rainfall. The result is a reduced bias of the simulated estimates of streamflow and evapotranspiration. In addition, the simulated water balance dynamics better reflect underlying governing factors such as soil type, land cover and climate at both annual and seasonal time scales, indicating the structural and behavioral consistency of the calibrated model. This study demonstrates that the judicious use of available information can help to facilitate meaningful calibration and evaluation of hydrologic models to support decision making in poorly gauged river basins around the world.

Frequency-based performance measure for hydrologic model evaluation

Scale-dependent and -independent error and performance measures are used to evaluate hydrologic forecasting and simulation models. However, these single-valued measures may not always provide a comprehensive assessment of model performance developed for a specific hydrologic process and application. A new frequency-based performance measure (FBPM) that can incorporate application-specific information for model evaluation is proposed to address the limitations of existing measures. FBPM is derived using a data classification scheme to partition the observed data into several classes and evaluate frequencies of each class's chronologically paired observed and forecasted values.A variant of FBPM, composite performance measure (CPM), is also developed to include additional indices that evaluate error, variance, and other statistical characteristics of the observed and forecasted series. Several univariate and multivariate data classification schemes are initially evaluated, and the best one is selected for use in the measures.FBPM and CPM are used to assess the performance of daily streamflow forecasting models developed for the Tapi River, India, using a data-driven model tree (MT) approach. The measures are also evaluated using two synthetic datasets representing different model forecast scenarios. A comprehensive evaluation of streamflow forecasts using FBPM and CPM with the best data classification scheme (i.e., geometricinterval schemein this work) indicates a better and more robust assessment of the forecasting model performance than that from existing error and performance measures. The traditional measures such as coefficient of determination, index of agreement, Nash Sutcliffe, and Kling Gupta efficiency have overestimated model performances compared to FBPM and CPM on an average by at least 100%. The traditional measures have also failed to identify inferior models mainly due to their inflated numerical values compared to those from FBPM and CPM. The measures developed in this study allow the user-specific definition of classesandassignment of weightstothese classes based on the intended purpose of the hydrologic modeling effort. The FBPM and CPM are scale-independent, informative, interpretable, andoutlier-resistant. These measures can be used for visual and/or statistical evaluation of the performance ofsingleor multiple hydrologic simulation models.

Integrating remote sensing data in optimization of a national water resources model to improve the spatial pattern performance of evapotranspiration

Hydrological models have traditionally been calibrated and evaluated against point-scale observations, such as streamflow, emphasizing the temporal, but not necessarily the spatial component of a model. The main goal of this study is to improve the spatial pattern performance of simulated actual evapotranspiration (AET), a key variable in the land–atmosphere interface, for the entire land-phase of Denmark. This is achieved by integrating fully distributed remote-sensing (RS) data and robust objective functions in the regionalization and optimization of the national water resources model of Denmark (DK-model). For this, monthly spatial patterns of MODIS16 ET at 1 km resolution are used as a reference RS ET estimate in the model calibration. We applied a gradient-based nonlinear Gauss-Marquardt-Levenberg (GML) algorithm for parameter estimation via inverse modeling. The proposed optimization framework includes three calibration strategies: Cal01 represents the traditional calibration approach against discharge and groundwater head observations only, Cal02 expands Cal01 by means of a spatial pattern-oriented objective function, namely the spatial efficiency (SPAEF) metric, targeting ET pattern performance, Cal03 adds to Cal02 an additional parameter regionalization scheme, where the original fixed land use class dependent distribution and seasonal development of leaf area index (LAI), crop coefficient (Kc) and root depth (RD) are replaced with empirical schemes driven by distributed vegetation (NDVI) and soil texture (clay fraction) data. Our results reveal that a significant improvement in the simulated spatial patterns of AET is obtained when combining regionalization schemes and a spatial pattern oriented objective function in Cal03, which is highlighted by a Copula analysis also. At the same time, a very limited trade-off in the model performance of AET patterns, streamflow and groundwater-heads is observed. Finally, the regionalization scheme (Cal03) affected the simulated groundwater recharge according to changes of AET. The findings of this study contribute to further improve spatial representation and evaluation of hydrological models and highlight the need to carefully design regionalization schemes as well as objective functions in fully distributed models.

Comment on Liu (2020): A rational performance criterion for hydrological model

The performance criteria have been often used in the hydrological model evaluation or calibration. Since the performance criteria have their own advantages and disadvantages, it is necessary to understand the impact of each performance criterion on the model evaluation or calibration to be used for the purposes of model applications. The Liu Mean Efficiency (LME) was proposed for the design flow estimates from hydrological modeling data by reformulating the three metric components (correlation, variability, and bias measures) in the Nash and Sutcliffe efficiency (NSE) or the Kling and Gupta Efficiency (KGE) to improve the flow variability in the runoff simulations optimized with the NSE or KGE. This article has discussed that the LME criterion has weaknesses of an underdetermined system that can potentially induce the vagueness in comparative assessments for alternative model performance as well as an overestimation system that can substantially reduce the reliability in design flow estimates through hydrological modeling. A complement approach to the limitations of the LME is necessary for the LME to be made fit for a rational performance criterion.

Goodness-of-fit criteria for hydrological models: Model calibration and performance assessment

This study provides guidelines for the selection of proper goodness-of-fit criteria for the calibration and evaluation of hydrological models. Popular goodness-of-fit criteria and good practices for hydrological modeling are reviewed. The review discusses the advantages and disadvantages of several criteria and is followed by a case study that focuses on the review’s main findings. The main recommendation is for hydrologists to avoid using threshold values to assess model performance and preferably set a proper benchmark series. The case study was developed using the GR5J hydrological model and data from 179 watersheds in the Brazilian Cerrado biome. Several single- and multi-objective functions are used in optimization runs to assess the outcome for different goodness-of-fit criteria. The model performance is evaluated for each optimization run considering overall conditions, i.e., entire time series, and conditions under low- and peak-flow conditions. The study case reinforces that the popular Nash-Sutcliffe efficiency index should be avoided as an objective function. Alternatively, the Kling-Gupta efficiency index showed to be a more reliable criterion, resulting in lower bias for both calibration and validation, and balanced results for both low- and peak-flow conditions. Additionally, combining different criteria in multi-objective functions can result in robust trade-offs. General guidelines are summarized and additional emphasis is given to tropical watersheds where low flows deserve due attention.

TOSSH: A Toolbox for Streamflow Signatures in Hydrology

We present a Matlab toolbox to calculate hydrologic signatures, which are metrics that quantify streamflow dynamics. Signatures are widely used for catchment characterisation, hydrologic model evaluation, and assessment of instream habitat, but standardisation across applications and advice on signature selection is lacking. The toolbox provides accessible, standardised signature calculations, with clear information on methodological decisions and recommended parameter values. The toolbox implements three categories of signatures: basic signatures that describe the five components of a natural streamflow regime, signatures from benchmark papers, and an extended set of process-based signatures. The toolbox is designed for ease of use, including documentation, workflow scripts and example data to demonstrate implementation procedures, and visualisation options. We demonstrate the accuracy and robustness of the signature calculations by applying reproducible workflows to large streamflow datasets. The modular design of the toolbox allows for flexibility and easy future expansion. The toolbox is available from https://github.com/TOSSHtoolbox/TOSSH (https://doi.org/10.5281/zenodo.4313275).

Comprehensive evaluation of hydrological models for climate change impact assessment in the Upper Yangtze River Basin, China

Climate change has substantial impacts on regional hydrology in the major river basins. To figure out such latent hydrological impacts of changing climate, more reliable hydrological simulations are imperative. In this study, we evaluated the impacts of climate change on hydrological regime in the Upper Yangtze River Basin based on four downscaled and bias-corrected Global Climate Model outputs from Coupled Model Intercomparison Project Phase 5 under four Representative Concentration Pathways (RCP2.6, RCP4.5, RCP6.0, and RCP8.5) driving three hydrological models. Two model evaluation approaches were applied: simple and comprehensive. The comprehensive approach was used to evaluate models in the historical period, optimizing objective function at four gauges, and hydrological models were weighted for impact assessment based on their performance. In such a way, projected streamflow time series are obtained under different emission scenarios. Results show that the annual average discharge is projected to increase by 4.1–10.5% under the RCP scenarios at the end of twenty-first century relative to the reference period (1970–1999). Moreover, the high flow is projected to increase and the low flow to decrease indicating a higher probability of flood and drought occurrence in the basin. The severity of floods and droughts may increase. In comparison with the simple one-site model evaluation approach, the comprehensive method reveals that the anticipated extreme flow events would be less severe, and annual mean discharge slightly lower. The projected results imply that application of the comprehensive model evaluation approach could narrow the simulated spreads of projections significantly, and might provide more credible results.

Impacts of precipitation and topographic conditions on the model simulation in the north of China

AbstractThe evaluation of hydrological models for a specific catchment is normally based on the model performance according to the selected performance criteria. However, the catchment rainfall-runoff characteristics could be used for the selection of a suitable hydrological model in study area, which, also, for the problem solve of the model application in ungauged basins. In this study, six conceptual models were applied in three semi-humid or semi-arid catchments to investigate the correlation between catchment characteristics and model structure selection. In addition, the impacts of precipitation and topography in model simulation were analyzed. The results show that runoff generation are highly impacted by catchment topographic index and land cover change, and the influence of slope for river channel is greater than mean slope for the whole catchment due to the runoff generation for partial area. For the catchments under similar climate condition, the impact of topographic features for runoff generation process is greater than the difference of precipitation. It indicates that for a specific catchment, the selection of appropriate model should base on better understanding of the rainfall-runoff relationship. The method of incorporating additional runoff generation module in the traditional model can significantly improve the accuracy of flood simulation.