No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Effect of sensitivity analysis on parameter optimization: Case study based on streamflow simulations using the SWAT model in China
Abstract
Parameter optimization is an essential step in hydrological simulations, especially for solving practical problems. However, parameter optimization is usually intractable for complex models with a large number of parameters. In this study, a parameter optimization system based on Sensitive Parameter Combinations (SPCs) was developed, which comprised four parameter sensitivity analysis (SA) methods and a sensitive parameter optimization method. In particular, parameter SA was used to screen out the relatively sensitive parameters with significant impacts on the model output, and instead of using All Parameter Combinations (APCs), the SPCs were optimized with a global optimization method. This system was applied to the Soil and Water Assessment Tool (SWAT) model for daily streamflow simulation and monthly evaluation in four watersheds of China. The results showed that no more than 10 sensitive parameters were identified from 27 adjustable parameters for each watershed. In particular, four parameters (CN2, SOL_K, ALPHA_BNK, and SLSUBBSN) were relatively sensitive in all watersheds. Compared with optimizing APCs, despite the number of parameters was reduced by almost 2/3 in the optimization of SPCs, the accuracy was still very close (the maximum Nash–Sutcliffe coefficient (NSE) difference was 0.024 and the minimum difference was 0.002) and the optimization speed was doubled. In the comparison of monthly streamflow optimization, the SPCs were in good agreement with the APCs and had an obvious improvement for the default simulation. The NSE values of the SPCs optimization were greater than 0.88 during the calibration period in all watersheds and greater than 0.83 during the validation period in three watersheds. These findings indicate that optimizing the sensitivity parameters can greatly reduce the computational costs of SWAT streamflow simulations while ensuring their accuracy.
... Based on several studies [45,49,50], we considered 24 commonly used parameters, along with their previously established reference ranges, for the parameter sensitivity analysis. For Sobol's quantitative SA, we specified 100 random sample points for 24 tunable parameter ranges, resulting in 5000 parameter samples [51,52]. We obtained the first-order sensitivity results of each parameter after 5000 SWAT+ model simulation runs. ...
... In this study, we used a quantitative sensitivity analysis method, Sobol's method, which could evaluate the impacts of each parameter (first-order effects) and its interactions (interaction effects) with other parameters on the model output [46,52]. Sobol's method calculates the contribution of the variance due to the perturbation of each parameter relative to the total variance of the model output [51]. We obtained the first-order sensitivities of 24 common parameters after 5000 SWAT+ model simulation runs based on the SWAT+ toolbox v1.0. ...
... However, the first-order sensitivity values were more uniform when using other quantitative sensitivity analysis methods, such as the McKay main effect analysis method [66]. Many studies [51,67] recommend that sensitive parameters should be chosen when the absolute value of the first-order sensitivity according to Sobol's method is greater than 0. Another reason for the low sensitivity value may have been that the interaction between parameters was not considered in this study. Gan et al. [66] concluded that parameters with lower main effect values may have significant interaction effects with other parameters, which should also be treated as important parameters. ...
Multisource meteorological re-analyses provide the most reliable forcing data for driving hydrological models to simulate streamflow. We aimed to assess different hydrological responses through hydrological modeling in the upper Lancang-Mekong River Basin (LMRB) using two gridded meteorological datasets, Climate Forecast System Re-analysis (CFSR) and the China Meteorological Assimilation Driving Datasets for the Soil and Water Assessment Tool (SWAT) model (CMADS). We selected the Pearson’s correlation coefficient (R), percent bias (PBIAS), and root mean square error (RMSE) indices to compare the six meteorological variables of the two datasets. The spatial distributions of the statistical indicators in CFSR and CMADS, namely, the R, PBIAS, and RMSE values, were different. Furthermore, the soil and water assessment tool plus (SWAT+) model was used to perform hydrological modeling based on CFSR and CMADS meteorological re-analyses in the upper LMRB. The different meteorological datasets resulted in significant differences in hydrological responses, reflected by variations in the sensitive parameters and their optimal values. The differences in the calibrated optimal values for the sensitive parameters led to differences in the simulated water balance components between the CFSR- and CMADS-based SWAT+ models. These findings could help improve the understanding of the strengths and weaknesses of different meteorological re-analysis datasets and their roles in hydrological modeling.
... The development of optimization system, such as based on sensitive parameter combinations (SPC), is other viable strategy to deal with multiparameter optimization problems. Although there are some limitations, due to the lack of sufficient observational data and the model itself, the use of SPC ensures significant improvements in the streamflow simulation of the SWAT model (Li et al. 2021a). Many satellite precipitation products (SPP) for the purpose of providing spatiotemporally continuous precipitation estimates. ...
... Paul et al. (2021) showed that remotely sensed data integrated into hydrological models may be used as decision support tool on water productivity and crop yield estimates, especially in the semi-arid regions, where there are limited observations, can greatly benefit from incorporating high-resolution MODIS LAI data. In general, strategies like these can be used and applied in watersheds with impoverished data, with the purpose of detecting protection (Andrade et al. 2017b;Aghakhani Afshar et al. 2018;Luan et al. 2018b;Li et al. 2021a;Paul et al. 2021;Zhang et al. 2022b). Despite the resolution or temporal limitations that satellite products can generate for the SWAT output data, the cited studies prove the good performance of the model in environments with limited data, which include dry environments. ...
The use of models affords an understanding of hydrological processes under different environmental conditions, and aids action planning in the face of changes in land use, soil, and climate. Arid and semi-arid regions are the most vulnerable to climate change. To this effect, the SWAT model has been used in different locations around the world and in various applications. This work analysed recent studies published between 2009 and 2022, to understand the following: (1) how the SWAT model has been applied in arid and semi-arid regions of the world, (2) the criteria used in applying the model, and (3) what alternatives can be used to overcome the limitation of input data and improve SWAT performance in dry environments? A systematic search was conducted based on scientific articles. A total of 234 articles were identified where SWAT was used drylands. The largest and the smallest applications were implemented in Asia and Oceania respectively, in such segments as agriculture, water governance, environmental conservation, hydrological processes, and extreme events. The number of years of data used for calibration and validation was generally low (<20 years), with a median of 9 for calibration and 6 years for validation. The NSE (92%) was the most applied in analysing modelling efficiency, followed by the R² (78%) and PBIAS (51%). The lack of observational data in the drylands is still a major challenge for studies que involving hydrologic modelling, and to improve the hydrological responses, researchers use strategies such as the use of data measured in the field, remote sensing, and even application of other models with or inside the SWAT.
... Hence, from Table 7, it can be seen that the parameters that are the most sensitive are CN2 and ESCO. [42][43][44][45]. Figure 4a,b illustrates the monthly calibrated SF. Table 8 explains the performance of the calibrated model. ...
... The validation follows the same steps as the calibration. The monthly step model was [42][43][44][45]. Figure 5a,b illustrate the monthly validated streamflow. Table 8 explains the performance of the validated model. ...
Understanding the likely impacts of climate change (CC) and Land Use Land Cover (LULC) on water resources (WR) is critical for a water basin's mitigation. The present study intends to quantify the impact of (CC) and (LULC) on the streamflow (SF) of the Parvara Mula Basin (PMB) using SWAT. The SWAT model was calibrated and validated using the SWAT Calibration Uncertainty Program (SWAT-CUP) for the two time periods (), respectively. To evaluate the model's performance, statistical matrices such as R 2 , NSE, PBIAS, and RSR were computed for both the calibrated and validated periods. For both these periods, the calibrated and validated results of the model were found to be very good. In this study, three bias-corrected CMIP6 GCMs (ACCESS-CM2, BCC-CSM2-MR, and CanESM5) under three scenarios (ssp245, ssp370, and ssp585) have been adopted by assuming no change in the existing LULC (2018). The results obtained from the SWAT simulation at the end of the century show that there will be an increase in streamflow (SF) by 44.75% to 53.72%, 45.80% to 77.31%, and 48.51% to 83.12% according to ACCESS-CM2, BCC-CSM2-MR, and CanESM5, respectively. A mean ensemble model was created to determine the net change in streamflow under different scenarios for different future time projections. The results obtained from the mean ensembled model also reveal an increase in the SF for the near future (2020-2040), mid future (2041-2070), and far future (2071-2100) to be 64.19%, 47.33%, and 70.59%, respectively. Finally, based on the obtained results, it was concluded that the CanESM5 model produces better results than the ACCESS-CM2 and BCC-CSM2-MR models. As a result, the streamflow evaluated with this model can be used for the PMB's future water management strategies. Thus, this study's findings may be helpful in developing water management strategies and preventing the pessimistic effect of CC in the PMB.
... Furthermore, the optimal parameters will be found when the convergence criterion (i.e., the function value is not improved after a certain shuffling loop number, n loop ) is met after repeating the evolution and shuffling steps (Francés et al., 2007;Kan et al., 2018). In this paper, we set n loop is 20, which has been proved able to capture the changes between the optimization process of SPs and APs by Li et al. (2021). In addition, the maximum number of iterations n maxr is set to 20,000, and the optimization will also stop once the model runs exceed this value. ...
... Moreover, the optimization accuracy of SPs was close to APs, and it was also sufficient for practical application in hydrological simulations. This is consistent with the conclusion obtained by Li et al. (2021). Thus, the SPs obtained by GDISCs system contribute to shortening the parameter optimization process well, with satisfactory model performance. ...
Parameter optimization is mandatory in many numerical modeling studies of hydrology sciences. However, optimizing all parameters (APs) is highly inefficient, including the equifinality phenomenon. This paper proposes an integrated sensitivity analysis system consisting of, multiple Global sensitivity analysis methods, Design of experiment algorithms, Indicators, and Sample size design Combinations (GDISCs), for screening the sensitive parameters (SPs) robustly and reducing parameter optimization burden efficaciously. In this study, four design of experiment (DoE) algorithms are applied to collect samples with varying sizes, and the corresponding three indicator values. We evaluate the effectiveness and efficiency of one qualitative and two quantitative global sensitivity analysis (GSA) methods by identifying the appropriate DoE algorithm and the sufficient sample size. To apply the GDISCs, the Block-wise use of the TOPMODEL (BTOP) model is set up for the hydrological simulation in the upper Min River Basin, China. The results show that the GDISCs system robustly screens out three SPs of the BTOP model, and the DoE algorithms significantly impact the effectiveness and efficiency of the GSA methods. The SA performance with applying the modified Kling–Gupta efficiency (KGE') as the indicator is more robust than the root mean squared error (RMSE) and the Nash-Sutcliffe efficiency (NSE). Compared to the traditional APs optimization, the optimal values are identified quickly through almost 1/2 or even fewer model runs and duration for the SPs optimization, with a very close accuracy (the maximum NSE, KGE' and relative bias (RBias) differences are 0.137, 0.037 and 2.70%, and the minimum are 0.002, 0.001 and 0.20%). Optimizing SPs also has a noticeable improvement for the default parameter values, and the maximum differences of NSE, KGE' and RBias reach 0.764, 0.321, and -20.80%. In summary, the GDISCs system is especially applicable for obtaining the robust SA results to considerably improve model calibration efficiency and reduce model computational burden while ensuring satisfactory model simulation performance.
... These parameters should be calibrated to the best local values to obtain the optimal match between the simulated and measured values. Many different studies have been done on the calibration of the SWAT model (Athira 2021;Li et al. 2021;Shah et al. 2021;Ma et al. 2022). There are different methods for analyzing uncertainty in watershed distribution models such as GLUE, ParaSol, MCMC and SUFI2 methods. ...
... The final range of model parameters is shown in Table 5. By examining different methods of sensitivity analysis, Li et al. (2021) found parameters for calibration with daily and monthly streamflow data that were similar to those of the present study (Table 5). Similarly, the results of the SWAT model calibration in that study provided acceptable NSE values, which are consistent with those in this study (Table 6). ...
Considering the importance of rainfed agriculture in adaptation to nature and long-term sustainability in the human food supply and livelihood of farmers, the main purpose of this study is to investigate the potential of rainfed agriculture in the Zarrinehroud basin as this basin is one of the most important sub-basins of Lake Urmia. For this study, the remote sensing data of surface soil moisture and evapotranspiration were combined with the SWAT model using the Data Assimilation method, Ensemble Kalman Filter (EnKF). Calibration of runoff flow rate in the SWAT model showed the correlation coefficient ranging between 0.69 and 0.84 in the calibration period (2000–2009) and between 0.64 and 0.86 for the validation period (2010–2014). The assimilation of the remote sensing data with the calibrated SWAT model showed that the model simulations for both the variables of surface soil moisture and actual evapotranspiration improved by at least 25% in both 2010 and 2014. It has been determined that 10.5 and 25.4% of the region's lands have a Very Appropriate and Appropriate potential for rainfed wheat agriculture, respectively. Areas with Moderate and Inappropriate potential occupy 64.1% of the lands in the region. HIGHLIGHTS
The efficiency of the SWAT model in predicting the yield of rainfed wheat was evaluated in improvement with remote sensing data.;
Assimilation of remote sensing data significantly improved the simulation results of the calibrated SWAT model.;
The results of this study could be an efficient tool in order to cope with water scarcity in the region for agricultural and water resources decision makers.;
... Hydrological models serve as the primary solution for understanding and predicting hydrological processes, considering the economic constraints in measuring components and their non-linear and complex relationships [1,2]. However, hydrologists often face challenges in data collection for evaluating and modeling performance [3][4][5][6]. Developing countries, including Iran, confront data shortages, sparse station distribution, data gaps, a lack of updates, and high costs associated with data collection, particularly in arid regions. To address these limitations, alternative sources of climate input for hydrological models are required in areas without comprehensive data [7][8][9]. ...
Hydrological modeling is essential for runoff simulations in line with climate studies, especially in remote areas with data scarcity. Advancements in climatic precipitation datasets have improved the accuracy of hydrological modeling. This research aims to evaluate the APHRODITE, PERSIANN-CDR, and ERA5-Land climatic precipitation datasets for the Hablehroud watershed in Iran. The datasets were compared with interpolated ground station precipitation data using the inverse distance weighted (IDW) method. The variable infiltration capacity (VIC) model was utilized to simulate runoff from 1992 to 1996. The results revealed that the APHRODITE and PERSIANN-CDR datasets demonstrated the highest and lowest accuracy, respectively. The sensitivity of the model was analyzed using each precipitation dataset, and model calibration was performed using the Kling-Gupta efficiency (KGE). The evaluation of daily runoff simulation based on observed precipitation indicated a KGE value of 0.78 and 0.76 during the calibration and validation periods, respectively. The KGE values at the daily time scale were 0.64 and 0.77 for PERSIANN-CDR data, 0.62 and 0.75 for APHRODITE precipitation data, 0.50 and 0.66 for ERA5-Land precipitation data during the calibration and validation periods, respectively. These results indicate that despite varying sensitivity, climatic precipitation datasets present satisfactory performance, particularly in poorly gauged basins with infrequent historical datasets.
... Parameters related to the soil type and cover, management practices and basin characteristics were also important: SOL_Z, CANMX, SOL_AWC, BIOMIX and SURLAG. Verma and Dash (2016), Daramola et al. (2019), Musie et al. (2020), Li et al. (2021), Xiang et al. (2022) and Wu et al. (2022b) also reported these parameters as among the most sensitive in terms of flow. ...
Hydrosedimentological models make it possible to better understand the dynamics of water and sediment production in watersheds when properly calibrated. The objective of this study was to analyze the effects of the curve number (CN) and Green and Ampt (GA) methods and of seasonal calibration of the Soil and Water Assessment Tool (SWAT) model for estimating flow and sediment production in an agricultural basin. In this research, we presented an original application with the hourly suspended sediment concentration (SSC) generated by artificial neural networks (ANNs) for use in SWAT model calibration. This method was applied in the Taboão basin (77.5 km²), with data from 2008 to 2018. The best Nash–Sutcliffe (NS) coefficient values were obtained using the combination of wet years for calibration and the GA method for both daily flow (NScalibration: 0.74; and NSvalidation: 0.68) and daily sediment production (NScalibration: 0.83; and NSvalidation: 0.77). The CN method did not result in satisfactory values during daily flow calibration (NScalibration 0.39). The results showed that it is possible to employ the SWAT model for hydrosedimentological prediction in the Taboão basin, with a favorable efficiency, using the GA method and calibration with data for wet periods.
... Although the lumped models have been widely applied in the NPS pollution simulation, it ignores the spatial-temporal distribution of hydrological elements, and the model is suitable for small watershed. Therefore, the distributed model came into being (Pradhan et al. 2020;Li et al. 2021). The typical distributed models for NPS pollution simulation consist of the AGNPS (Agricultural Non-point Source) model, HSPF (Hydrological Simulation Program-Fortran) model, SWAT (Soil Water Assessment Tool) model, and MIKESHE (MIKE System Hydrological European) model Zhuang et al. 2022). ...
... The temporal analysis of runoff trends under different scenarios (SSP245 and SSP585) offers insights into how future runoff may evolve over time, with a focus on the period of 2023-2100. The consideration of different SSP scenarios allows for a comprehensive assessment of potential future runoff changes based on varying greenhouse gas emissions, enhancing the study's reliability [72]. ...
Climate and land use changes are major factors affecting runoff in regional basins. Understanding this variation through considering the interactions among hydrological components is an important process for water resource management. This study aimed to assess the variation of future runoff in the Upper Chi Basin, Northeastern Thailand. The QSWAT hydrological model was integrated into three CMIP6 GCMs—ACCESS-CM2, MIROC6, and MPI-ESM1-2-LR—under SSP245 and SSP585 scenarios for the period 2023–2100. The Land Change Modeler (LCM) was also used for future land use simulation. The results revealed that the future average long-term precipitation and temperature tended to increase while forest land tended to decrease and be replaced by sugarcane plantations. The accuracy assessment of the baseline year runoff calculation using QSWAT for the period 1997–2022 showed an acceptable result, as can be seen from the R2, NSE, RSR, and PBIAS indices. This result could lead to the temporal and spatial simulation of future runoff. Likewise, the runoff of the two SSP scenarios tended to increase consecutively, especially in the SSP585 scenario. In addition, in cases of long-term spatial changes in the subbasins scale, over 90% of the area—from upstream to the outlet point—tended to be higher due to two major factors; namely, future increased precipitation and changes in cultivation, which would be influential to groundwater and interflow components, respectively. The methodology and result of this study can be useful to stakeholders in understanding changes in hydrological systems so that they can apply it to developing a strategy for water resource management and to handling factors affecting different dimensions properly and sustainably.
... In this study, the SWAT model was used in simulating the runoff change process in the JRB from 1951 to 1996, with the years 1951-1960, 1961-1980, and 1981-1996 being used as the warm-up, calibration, and verification periods, respectively, and the SUFI-2 algorithm of SWAT-CUP was used for parameter optimization [44]. In addition, the correlation coefficient (R 2 ), Nash-Sutcliffe efficiency coefficient (NS), and relative error (RE) were used to verify the accuracy of the simulation results and evaluate the applicability of the SWAT model in the JRB [45,46]: ...
In the context of global warming and intensified human activities, the quantitative assessment of the combined effects of land use/cover change (LUCC) and climate change on the hydrological cycle is crucial. This study was based on the simulation results of future climate and LUCC in the Jinghe River Basin (JRB) using the GFDL–ESM2M and CA–Markov combined with the SWAT models to simulate the runoff changes under different scenarios. The results revealed that the future annual precipitation and average temperature in the JRB are on the increase, and the future LUCC changes are mainly reflected in the increase in forest and urban lands and decrease in farmlands. Changes in runoff in the JRB are dominated by precipitation, and the frequency of extreme events increases with the increase in the concentration of CO2 emissions. Under four climate scenarios, the contribution of future climate change to runoff changes in the JRB is −8.06%, −27.30%, −8.12%, and +1.10%, respectively, whereas the influence of future LUCC changes is smaller, ranging from 1.14–1.64%. In response to the future risk of increasing water-resources stress in the JRB, the results of this study can provide a scientific basis for ecological protection and water-resources management and development.
... From these examples two aspects of PSA can be identified: OAT has been widely used and proven useful approach; PSA has mostly been carried out for some typical hydrological models (Huang et al., 2020). Li et al. (2021) indicate that optimizing the sensitivity parameters can greatly reduce the computational costs of SWAT streamflow simulations while ensuring their accuracy. However, it is difficult to calibrate the parameters of hydrological models in poorly gauged or ungauged basins. ...
Study Region: Nine mountain catchments (5 in northern and 4 in southern China).Study Focus: Flash floods in small mountain catchments are some of the most frequent and most destructive natural disasters of China. Mixed runoff hydrological model is a popular method for flood warning and forecasting. However, the key factors affecting runoff components in hydrological model parameter calibration need to be further studied. In this study, four different overland flow routing techniques are implemented into a semi-distributed mixed runoff hydrological model system (MRHMS), which are the triangle unit hydrograph (TUH), geomorphological instantaneous unit hydrograph (GIUH), geomorphoclimatic instantaneous unit hydrograph (GcIUH) and kinematic wave (KW). MRHMS-KW is the previously developed spatiotemporally-mixed runoff (SKBY) model.New hydrological insights for the region: The study finds that sensitivity parameters of the MRHMS in northern and southern China are related to their main runoff components. Different parameter orders in automatic calibration of MRHMS result in different model performances, and the suggested parameter orders are consistent with the sensitivity of parameters in northern and southern China. Soil moisture and primary runoff components in northern and southern China differ significantly. Only MRHMS-GcIUH can achieve similar performance in northern and southern catchments. MRHMS-GcIUH therefore enhances our understanding of runoff formation mechanisms in the heterogeneous hydroclimatic landscape of China and is a promising tool for a nationwide flood prediction and flood warning system.
... Based on several studies [35,39,40], we considered 24 commonly used 7 of 23 parameters, along with their previously established reference ranges, for the parameter sensitivity analysis. For Sobol's quantitative SA, we specify 100 seed for each parameter, then 5000 samples were randomly generated from 24 tunable parameter ranges [41,42]. We can obtain the 1st order sensitivity results of each parameter after 5000 times SWAT+ model simulation run. ...
Multisource meteorological re-analyses are the most reliable forcing data for driving hydrological models to simulate streamflow. We aimed to assess the different hydrological responses through hydrological modeling in the upper Lancang-Mekong River Basin (LMRB) using the two gridded meteorological datasets, climate forecast system reanalysis (CFSR) and the China Meteorological Assimilation Driving Datasets for the SWAT model (CMADS). We selected the Pearson’s correlation coefficient (R), percent bias (PBIAS), and root mean square error (RMSE) indices to compare the six meteorological variables of the two datasets. The spatial distributions of the statistical indicators in the CFSR and CMADS, namely, the R, PBIAS, and RMSE values, were different. Furthermore, the soil and water assessment tool plus (SWAT+) model was used to do hydrological modeling based on CFSR and CMADS meteorological re-analyses in the upper LMRB. Different meteorological datasets resulted in significant differences of hydrological responses, which reflected by different sensitive parameters and their optimal value. These different calibrated optimal values of sensitive parameters further lead to the different simulated water balance components between CFSR- and CMADS-based SWAT+ model. These findings can help in a better understanding of the strengths and weaknesses of different meteorological re-analysis datasets and the roles on the hydrological modeling.
... Attada et al. (2018) and Reddy et al. (2020) assessed the effectiveness of three LSMs at simulating the main climatic variables such as precipitation during the Indian summer monsoon season. In addition, some uncertainty qualification methods such as parameter optimization (Li et al., 2021;Zhang et al., 2020) and ensemble forecasting Liu et al., 2016) have been used widely to improve the performance of LSMs. ...
In 2020, the Yangtze-Huai river valley (YHRV) experienced the highest record-breaking Meiyu season since 1961, which was mainly characterized by the longest duration of precipitation lasting from early-June to mid-July, with frequent heavy rainstorms that caused severe flooding and deaths in China. Many studies have investigated the causes of this Meiyu season and its evolution, but the accuracy of precipitation simulations has received little attention. It is important to provide more accurate precipitation forecasts to help prevent and reduce flood disasters, thereby facilitating the maintenance of a healthy and sustainable earth ecosystem. In this study, we determined the optimal scheme among seven land surface model (LSMs) schemes in the Weather Research and Forecasting model for simulating the precipitation in the Meiyu season during 2020 over the YHRV region. We also investigated the mechanisms in the different LSMs that might affect precipitation simulations in terms of water and energy cycling. The results showed that the simulated amounts of precipitation were higher under all LSMs than the observations. The main differences occurred in rainstorm areas (>12 mm/day), and the differences in low rainfall areas were not significant (<8 mm/day). Among all of the LSMs, the Simplified Simple Biosphere (SSiB) model obtained the best performance, with the lowest root mean square error and the highest correlation. The SSiB model even outperformed the Bayesian model averaging result. Finally, some factors responsible for the differences modeling results were investigated to understand the related physical mechanism.
... The former parameter is a function of soil permeability, land use, and antecedent soil water conditions, thereby making it very important for the accurate estimation of surface runoff. This finding is consistent with the results of many other studies around the world [46][47][48]. For instance, Nossent et al. [49] performed research investigating the sensitivity of the SWAT's major parameters using Sobel's sensitivity analysis. ...
The water yield produced at the outlet of a sub-basin is the combination of multiple interacting land uses. In the majority of previous research, while accounting for the effect of land use and land cover (LULC) on water yield, the hydrologic components of a watershed have been attributed to the dominant land use class within that sub-basin. We adopted an approach to investigate the interaction effect of LULC on water yield (WYLD) using the Johnson–Neyman (JN) method. The soil and water assessment tool (SWAT) model was employed in the Urmia Lake Basin (ULB) to estimate the WYLD following successful calibration and validation of the model by stream flow. It was found that in each sub-basin, the effect of the soil class on the WYLD was statistically significant only when the area of rangeland was less than 717 ha and when the area of agricultural lands was less than 633 ha. On the other hand, the trend of stream flow was assessed over 70 years at two stations in the Urmia Lake Basin (ULB) using the Bayesian Estimator of Abrupt change, Seasonal change, and Trend (BEAST). The year 1991 turned out to be the most likely change point in both stations. A significant decrease in Urmia Lake’s water level started in 1995, which indicated that part of this shrinkage was most likely caused by water inflow reduction over a 4-year time delay. Besides identifying the most probable seasonal and trend change points, this method has the additional capability to analyze the uncertainty of estimated points, which was lacking in earlier methods.
... showed similar results that in areas where saturation excess runoff dominates, daily discharge is the function of daily rainfall [79][80][81]. ...
A proper understanding of hydrological processes is vital for water resource assessment, management, and conservation at a local, national, and global scale. The role of hydrological models is critically important in rarely studied ungauged catchments including of Kobo-Golina, in the Danakil basin of Ethiopia. The main objective of this research is to model the hydrology of the Kobo-Golina catchment using the completely restructured SWAT (SWAT+) model. Validated reanalysis river flow from the Global Flood Awareness System (GloFAS) and actual evapotranspiration (AET) from Moderate Resolution Imaging Spectroradiometer (MODIS) were used for single and multi-variable calibration. It is found that the multi-variable calibration scenario reasonably attained the minimum satisfactory performance limit for both variables (NSE = 0.67, R 2 = 0.68, PBias = −9.68%, and RSR = 0.57 for calibration of GloFAS flow; and NSE = 0.56, R 2 = 0.63, RSR = 0.66, PBias = 3.86 for calibration of MODIS AET). The model simulation showed that evapotranspiration accounts for 47% of the input water while surface runoff, lateral flow, and groundwater recharge account for 30%, 1.53%, and 21.4%, respectively. The simulated mean annual streamflow at the Basin outlet is 10.6 m 3 /s. The monthly low flow occurs in June with a median flow of 1.43 m 3 /s and a coefficient of dispersion of 0.67. High flows occur in August, with a median flow of 16.55 m 3 /s and a coefficient of dispersion of 1.55. The spatial distribution of simulated runoff was depicted as being higher in the floodplains and along the riparian/drainage lines, whereas upland areas showed lower runoff. The maximum monthly recharge occurs in September with a recharge value of 78.2 mm. The findings of the study suggested that both surface water harvesting and groundwater exploitation can be sought in floodplain areas while conserving the uplands.
... SUFI-2 is a semi-automated approach, used to perform parameterization, sensitivity analysis, uncertainty analysis, calibration and validation of hydrologic parameters [71]. Sensitivity analysis in particular is necessary to understand which particular input parameter has a great impact on the model outflow [72]. In order to reduce the computing effort, the parameters chosen for the sensitivity and uncertainty analysis were chosen as the most sensitive following preliminary tests [73]. ...
The study of plumes occurring at the mouth of small rivers of temporal flow is a challenging task due to the lack of sedimentological and flow data of appropriate spatiotemporal scales. The present contribution examined the case of a typical un-gauged intermittent Mediterranean stream located in Northern Crete (Xiropotamos river). The SWAT (soil and water assessment tool) model was used to simulate and reproduce the hydrological behavior of the adjacent intermittent (Giofyros) river discharging at the same beach, the basin of which has the same geomorphological and hydrological characteristics. The output of the calibrated SWAT model was used to simulate daily flow data for the year 2014. The results were then considered together with the results of the RGB analysis of optical datasets of high spatio-temporal resolution for the same period, derived from a beach optical monitoring system (BOMS). The RGB analysis of the optical (TIMEX) imagery was shown to be a useful technique to identify and classify coastal plumes by using the spatio-temporal variability of pixel properties. The technique was also shown to be useful for the (qualitative) validation of the SWAT output and could be further improved by the collection of ‘ground truth’ data.
... There are many methods for the analysis of parameter sensitivity and parameter optimization (e.g., Markov chain Monte Carlo (MCMC), generalized likelihood uncertainty estimation (GLUE), shuffled complex evolution (SCE), Bayesian forecasting system (BFS)). However, evaluation of the interactions between parameters and their impacts on runoff components is still insufficient (Bo et al., 2020;Li et al., 2021a). ...
The coupled development of soil and vegetation leads to a close interaction between their attributes and impacts the sustainability of eco-hydrology at different scales. In this study, a distributed hydrological model of a watershed was created with the Soil and Water Assessment Tool (SWAT) in a representative tributary watershed for investigating such effects. The results quantify the intensity and interval of the relationship and the impacts on hydrological composition between major model parameters. Among the examined interactions, SCS runoff curve number (CN2) and soil bulk density (BD) show the strongest interaction and effects on surface runoff, lateral flow, percolation, groundwater flow, and soil water content. The interaction between CN2 and BD highlights the importance of the soil surface and topsoil for runoff generation processes. In addition, the soil-vegetation interactions show clear seasonal effects due to impacts from the changes in land use and precipitation patterns, which influence the river discharge and flow variability more significantly at the sub-basin scale than at the watershed scale. The insight into the interactions and hydrological effects of soil and vegetation may help improve the spatial planning for ecological sustainability and hydrological extrema mitigation with a more reliable reflection of the spatial heterogeneity.
... Eight parameters were identified by the SWAT-CUP with the SUFI-2 method as the most sensitive parameters for auto-calibration. Sensitivity analysis is a key step in understanding how parameters affect the SWAT model performances, and to decrease the number of parameters when a calibration process is required (Li et al., 2021a). The selected sensitive parameters were determined for the calibration and validation processes based on their p-value (Abbaspour et al., 2007) as shown in Table A1. ...
Drought may lead to severe and diverse impacts on agriculture, economy, and society across different regions and periods, posing predictive and adaptive challenges. In recent years, severe droughts have affected >60 million people in the Mekong River Basin (MRB). Additionally, El Niño Southern Oscillation (ENSO) episodes had distinct influences on the occurrence and intensity of drought variability in the regions. Understanding the spatiotemporal characteristics of droughts across the MRB is critical to improving management and mitigation actions. This study aims to investigate spatiotemporal drought characteristics in the MRB and their teleconnection with the ENSO. Three multiple drought indices, including the Standardized Precipitation Index (SPI) for meteorological drought, Standardized Soil Water Index (SSWI) for agricultural drought, and Standardized Runoff Index (SRI) for hydrological drought were calculated to quantify drought events, drought frequency, and drought severity. The overall patterns showed more events and larger intensity were identified by the SPI than those by the SRI or the SSWI, while the higher frequency was observed by the SRI. The Middle Mekong basins seem to experience more drought events, while higher levels of frequency and intensity of droughts were observed in the Upper Mekong Basin. The correlation analysis between ENSO index and precipitation suggested that the strongest ENSO events in Dec-Jan-Feb may result in developments of meteorological drought in Mar-Apr-May, and further led to hydrological and agricultural drought in Apr-May-Jun. Such ENSO effects had significant influences on drought variabilities in southern MRB and were insignificant in the north. The multiple drought indices show skills in identifying spatial and temporal drought characteristics from meteorological, agricultural, and hydrological perspectives, and potential for drought outlook further considering their ENSO teleconnections. The results can be applied to the development of drought monitoring methods and adaptive strategies to mitigate drought impacts through scientific and quantitative assessments.
... Currently, SWAT Calibration and Uncertainty Programs (SWAT-CUP) is mostly to determine the sensitive parameters and the optimal value. CN2 is mostly recognized as the most important parameter in SWAT streamflow simulations [61,62] because SWAT uses the soil conservation service streamflow curve method to calculate the surface streamflow. In addition, the surface streamflow in most watersheds plays a dominant role in the total annual streamflow. ...
The Grain for Green project (GGP), initialized by the Chinese government in 1999, has achieved substantial achievements accompanied by a decrease in surface runoff on the Loess Plateau, but the impacts of large-scale afforestation on regional water resources are uncertain. Hence, the objective of this study was to explore the impact of land use change on generalized water resources and ecological water stress using the blue and green water concepts, taking the Yanhe River Basin as the case study. The Soil and Water Assessment Tool (SWAT) was applied to quantify the green water and blue water, which are defined as generalized water resources. The ecological water requirement of vegetation (forest and grass), agricultural water footprint and virtual water flow are considered regional water requirements. The land use types of 1980 (Scenario I) and 2017 (Scenario II) were entered into the SWAT model while keeping the other parameters constant to isolate the influence of land use changes. The results show that the average annual differences in blue, green and generalized water resources were -72.08 million m ³ , 24.34 million m ³ , and -47.74 million m ³ , respectively, when the simulation results of Scenario II were subtracted from those of Scenario I, which shows that land use change caused by the GGP led to a decrease in blue and generalized water resources and an increase in green water resources. Surface runoff in Scenario I was more than that in Scenario II in all of the years of the study period from 1980–2017, and green water storage in Scenario I was more than that in Scenario II in all of the years of the study period except in 1998; although lateral flow in Scenario I was less than that in Scenario II except in 2000 and 2015, as was groundwater runoff in 1992, 2000 and 2015, and green water flow in 1998. Blue water flow, green water storage and green water flow in Scenario II were less than those in Scenario I in the whole basin, 12.89 percent of the basin and 99.21 percent of the basin, respectively. The total water footprint increased from 1995 to 2010 because the forest water footprint increased significantly in this period, although the agricultural water footprint and grass water footprint decreased. The ecological water stress index values had no obvious temporal change trends in either land use scenario, but the ecological water stress index in Scenario II was greater than that in Scenario I, which illustrates that the GGP led to an increase in ecological water stress from the perspective of generalized water resources.
... The maximum canopy storage (CANMX) parameter, which we calibrated in the initial stage and excluded from further calibration, was sensitive for all models. CANMX represents the maximum amount of precipitation that can be trapped in the canopy when the canopy is fully developed and plays an important role in the hydrological cycle of humid temperate regions (Arnold et al., 2013;Li et al., 2021). Our results showed a calibrated value of CANMX ranging from 51 mm (gauged model) to 69 mm (IMERG model), indicating that the vegetation canopies in the catchment retain more precipitation and allow for higher vegetation evapotranspiration . ...
Study region
The Porijõgi catchment in Estonia, Northern Europe.
Study focus
The lack of adequate precipitation gauges has long been a major obstacle for hydrological modeling. To date, the global availability of satellite and reanalysis precipitation (SRP) products at an increasingly high spatiotemporal resolution has made their use in hydrological applications possible. This study focuses on a comprehensive evaluation of the hydrological applicability of five SRP products. The evaluations were carried out in two parts: (1) evaluating the quality of satellite and reanalysis-based precipitation products relative to gauge observations, and (2) comparing gauge-simulated streamflow with SRP-based simulations using the SWAT model.
New hydrological insights for the region
Our results show moderate variation in the detection capability of SRP products and in their influence on streamflow simulations. None of the products performed well with daily-scale predictions. The IMERG, ERA5, and PERSIANN-CDR products showed the best detection capability for the monthly precipitation and demonstrated reliable performance to simulate the monthly streamflow (0.65 < Nash–Sutcliffe efficiency coefficient < 0.75). However, the SM2RASC and CMORPH-CRT products tended to underestimate the gauged precipitation and provided unsatisfactory performance in streamflow simulation. Overall, our findings suggest that SRP products can be a priori alternative sources of precipitation data suitable for hydrological applications in poorly gauged areas.
... Calibration can usually be defined as approximations of system parameters to closely fit observed data and hydrological behaviour (Abbaspour et al. 2007). Parameters were selected based on streamflow driving parameters (Li et al. 2021;Goudarzi et al. 2021). The developed model was processed from 1998 to 2016 and the first 3 years were given as warmup time. ...
A vital challenge of assessing global water resources is to achieve the optimal set of parameters in any hydrological model by simulating streamflow. Simulating single-site hydrological features for a large catchment may not account for regional variation, resulting in unmet watershed needs. The objective of this study was to assess the single-site calibration (SSC) and multi-site calibration (MSC) approaches using the hydrological Soil and Water Assessment Tool (SWAT) model on the Bharathpuzha watershed of India. The multi-site method entails splitting a large catchment into smaller ones and applying MSC criteria to the entire catchment. Monthly streamflow simulations were conducted to verify the model’s performance using the coefficient of determination (R²), Nash–Sutcliffe Efficiency (NSE), percent of bias (PBIAS) and Kling Gupta Efficiency (KGE). Results show that single-site approach values are more meaningful than multi-site. In SSC, the parameter determined by optimizing the model parameters at four different stations delivers a better result than in MSC, and MSC has less uncertainty with lower PBIAS and r-factor values. This case study was assumed to give experience with single and multi-site calibration in a large catchment and reveal the positives and negatives of SSC and MSC estimated parameters.
... The Soil and Water Assessment Tool (SWAT) is a physically based and semi distributed model that can simulate hydrological processes, soil erosion, and nonpoint source pollution in a watershed (Li et al. 2013). The SWAT model has a complex structure and many parameters, and most of these parameters need to be calibrated with the help of measured data (Li et al. 2021). Theoretically, parameter calibration mainly depends on the objective function, constraints, and measured data, but there is likely to be "equifinality for different parameters" due to uncertainty (Lee et al. 2012). ...
The phenomenon of "equifinality for different parameters" limits the link between parameters and catchment characteristics; however, solving the equifinality problem is a major challenge in the development, generalization, and application of a model. This study focused on the Yanhe River Watershed to investigate the time-varying characteristics of sensitivity and identifiability of SWAT (Soil and Water Assessment Tool) runoff and sediment parameters based on the Sobol' and generalized likelihood uncertainty estimation methods. The results indicate that (i) the nondominated sorting genetic algorithm-II has good adaptability and reliability in parameter calibration of the SWAT model in the Yanhe River Watershed. The evaluation indicators (Nash-Sutcliffe efficiency, R2, and percent bias) of monthly runoff and sediment in the Ganguyi hydrological station were all satisfactory per the SWAT model during the calibration and validation periods. (ii) The interaction between runoff and sediment parameters is a crucial reason for parameter sensitivity, which has obvious time-varying characteristics and is largely dependent on precipitation in the Yanhe River Watershed. Temporal and spatial variability of precipitation should be considered in the detailed analysis of parameter identifiability, and watershed managers should not ignore changes in the runoff process when regulating sediment. (iii) Only a relatively small number of parameters can be identified in the runoff and sediment simulation process of the Yanhe River Watershed, such as CN2 (initial soil conservation service runoff curve number for moisture condition II), CH_K2 (effective hydraulic conductivity in main channel alluvium), ALPHA_BF (baseflow alpha factor), USLE_C (cover and management factor), USLE_P (support practice factor), and USLE_K (soil erodibility factor), due to high surface runoff, reduced lag time, reduced low flows, increased peak flows, and channel erosion, respectively. More importantly, there is a strong positive correlation between parameter identifiability and parameter sensitivity. Both are effective methods of parameter diagnosis, but the identifiability of parameters is not equivalent to its sensitivity. Our results strongly suggest that a detailed parameter sensitivity and identifiability analysis is a critical step in improving hydrological model performance to reduce the risk of "equifinality for different parameters" while articulating all relevant hydrological processes.
The accurate and reliable simulation and prediction of runoff in the Beas River Basin are becoming more and more important due to the increased uncertainty posed by climate change, which is making it difficult to manage water resources efficiently. In order to minimize the effect of flash floods, estimating the accurate peak flow is essential. It can be challenging to comprehend and anticipate peak flow due to natural streamflow variance as well as the streamflow management offered by dams and reservoirs. Which makes it difficult to mimic hydrologic behavior on a daily scale with reliable accuracy. This study evaluated the efficacy of physics-aided machine learning (ML) based regression models for modeling streamflow in combination with process-based hydrological SWAT (soil and water assessment tool). Performance of eight machine learning (ML) models including linear regression (LR), multi-layer perceptron (MLP), light gradient-boosting machine (LGBM), extreme gradient boosting (XGBoost), kernel ridge (KR), elastic net (EN), Bayesian ridge (BR), and gradient boosting (GB) have been analyzed and compared with the calibrated-SWAT (cSWAT) model. The Nash–Sutcliffe efficiency (NSE), root mean square error (RMSE) and coefficient of determination (R2) were used to assess the effectiveness of both models. Results showed that the uncalibrated SWAT in combination with ML regression models (cSWAT-ML) performed well and found comparable to calibrated SWAT (cSWAT), though, few ML regression models in combination with uncalibrated SWAT (uSWAT-MLmodels models) performed superior. cSWAT model performed well with R2 values of 0.73, RMSE value of 276.92 m3/s and NSE value of 0.72. In uSWAT-ML, EN and BR have obtained better results with R2 values of 0.89 and 0.89, NSE values of 0.87 and 0.87, and RMSE values of 158.31 m3/s and 159.48 m3/s. The approached uSWAT-ML models have effectively predicted the peak stream flow rates with models BR and EN have predicted with better results of R2 value of 0.71 each. This study’s findings highlight the potential of all the eight ML models as promising techniques for predicting the peak flow discharge values when uncalibrated process-based models are combined.
Complex hydrological processes can be modeled using the Soil and Water Assessment Tool (SWAT). However, poor quality inputs, lack of observation and high number of input parameters make the modeling challenging and complicate the calibration process. The main aim of this study is to assess the parameters uncertainty and suitability of discharge and remotely sensed ET data to calibrate the SWAT model and explore the possibility of using MODIS satellite imagery for direct calibration in ungauged study areas. Different techniques are adopted to acquire accurate land use map and remotely sensed evapotranspiration derived from Sentinel-2 and MODIS satellite imagery. Sobol’ sensitivity analysis is used to effectively identify the key parameters and provides crucial information on how they influence the model outputs. To address model equifinality and uncertainty, the model of the Ilam Dam Basin in Iran is calibrated and validated using a multi-criteria performance evaluation approach that considers all performance criteria concurrently. The discharge predictions at the Sarjoy station showed good discharge predictions during 12 years of calibration (NSE = 0.6, KGE = 0.79, PBIAS = 3.7 and \({R}^{2}\)= 0.64) and 5 years of validation (NSE = 0.5, KGE = 0.76, PBIAS = -0.3 and \({R}^{2}\) = 0.58). ET predictions at the selected HRU also demonstrated acceptable results for calibration (NSE = 0.37, KGE = 0.69, PBIAS = -0.9 and \({R}^{2}\) = 0.47) and validation periods (NSE = 0.47, KGE = 0.73, PBIAS = 0.4 and \({R}^{2}\) = 0.54). The sensitivity analysis shows that while key parameters in both the discharge and ET models exhibited similarity, their varying levels of influence are notable. Furthermore, the direct calibration of the SWAT model yields acceptable performance criteria (NSE = 0.43, KGE = 0.66, PBIAS = -18 and \({R}^{2}\) = 0.61), making it a valuable approach in data-scarce regions.
Many hydrological applications employ conceptual-lumped models to support water resource management techniques. This study aims to evaluate the workability of applying a daily time-step conceptual-lumped model, HYdrological MODel (HYMOD), to the Headwaters Benue River Basin (HBRB) for future water resource management. This study combines both local and global sensitivity analysis (SA) approaches to focus on which model parameters most influence the model output. It also identifies how well the model parameters are defined in the model structure using six performance criteria to predict model uncertainty and improve model performance. The results showed that both SA approaches gave similar results in terms of sensitive parameters to the model output, which are also well-identified parameters in the model structure. The more precisely the model parameters are constrained in the small range, the smaller the model uncertainties, and therefore the better the model performance. The best simulation with regard to the measured streamflow lies within the narrow band of model uncertainty prediction for the behavioral parameter sets. This highlights that the simulated discharges agree with the observations satisfactorily, indicating the good performance of the hydrological model and the feasibility of using the HYMOD to estimate long time-series of river discharges in the study area.
HIGHLIGHTS
Local and global sensitivity analysis (SA) approaches were used for SA and parameter identifiability.;
Both approaches gave similar results in terms of sensitive parameters for the model output.;
A group of sensitive parameters depends on the selected objective criterion.;
Precisely identified parameters reduce the model uncertainties and enhance the model performance.;
Sensitive, well-defined parameters and model performance increase with catchment size.;
Hydro-sedimentological models make it possible to understand the dynamics of water and sediment production in watersheds, if properly calibrated. The objective of this study is to analyze the effect of Curve Number (CN) and Green & Ampt (GA) methods and of seasonal calibration of the Soil and Water Assessment Tool (SWAT) model for estimating flow and sediment production in an agricultural basin. This research presented an original application with hourly suspended sediment concentration (SSC) generated by Artificial Neural Networks (ANNs) for using to the SWAT model calibration. The study was applied in the Taboão basin (77.5 km²), with data from 2008 to 2018. The best Nash–Sutcliffe (NS) coefficients were obtained using the combination of wet years for calibration and the GA method, both for daily flow (NScalibration 0.74 and NSvalidation 0.68) and for daily sediment production (NScalibration 0.83 and NSvalidation 0.77). The CN method did not result in satisfactory values already in the calibration for daily flow (NScalibration 0.39). The results showed that it is possible to apply the SWAT model for hydrosedimentological prediction in the Taboão basin, with good efficiency, using the GA method and calibration with wet periods.
Best management practices (BMPs) have been widely adopted to mitigate diffuse source pollutants, and the simulated processes of its pollutant reduction effectiveness suffer from manifold uncertainties, such as watershed model parameters and climate change. We presented a novel Bayesian modeling framework for BMPs planning, integrating process-based watershed modeling and Bayesian optimization algorithm to reveal the impact of multiple uncertainties. The proposed framework was applied to a BMPs planning case study in the Erhai watershed, the seventh-largest freshwater lake in China. Firstly, priority management areas (PMAs) were identified for BMPs siting using a simulation-optimization approach. Bayesian networks were subsequently embedded to reveal the multiple uncertainty sources in the optimal planning and the reliability level (RL) is introduced to represent the probability to meet the water quality target with BMPs implementation. The results suggest that ENS of discharge and nutrients concentration simulation by LSPC are both greater than 0.5, which displays satisfactory performance. The identified PMAs account for 0.8% of the total watershed areas while contribute to more than 15% of nutrient loadings reduction. The analysis of multiple uncertainty sources reveals that precipitation is the most influential source of uncertainties in BMP effectiveness. The construction of hedgerows plays an important role in the nutrient reduction. With the improvement of the reliability levels, the cost increases sharply, indicating that the implementation of BMPs has a marginal utility. The study addressed the urgent need for effective and efficient BMPs planning by identifying PMAs and addressing multi-source uncertainties.
Since the implementation of landscape conservation of the green heart area in the Changsha-Zhuzhou-Xiangtan Metropolitan Region, the landscape structure and pattern have changed significantly. The ecosystem service functions in the area have been improved, but the status of ecohydrological and water quality and service functions (EHWQSFs) is still unclear. To clarify the status of EHWQSFs and their driving factors influenced by landscape conservation, this study analysed landscape changes using remote sensing image data from 1998, 2008, and 2018 and the changes and their spatial characteristics using the Soil and Water Assessment Tool (SWAT) and spatial analysis methods. The results showed that the dominant land types in the area were forestland and cropland from 1998 to 2018; the area of forestland and construction land expanded and that of cropland decreased year by year; the annual average surface runoff volume rose, and the annual average actual evapotranspiration and soil water content fell from 1998 to 2008 and rose from 2008 to 2018; and all pollutant indicators decreased significantly after 2008. The areas with higher surface runoff were mainly concentrated in the central and southern regions, those with higher evapotranspiration were in the northwestern and southwestern regions, those with higher soil water content were in the northern region, and those with higher sediment and nitrogen and phosphorus pollutant contents were in the central and southeastern regions. The results showed that land use, land cover and meteorological factors were the most significant drivers on EHWQSFs and illustrated that EHWQSFs in the area decreased after 1998. There was a significant improvement after 2008 and the area currently has a good status. This study not only provides insights into land use, land cover and meteorological factors that have significant impacts on EHWQSFs but also highlights that the landscape conservation of the area can improve ecosystem service functions.
In hydrology research, how to quantitatively analyze the spatial and temporal changes of water resources in the ecosystem of a watershed is one of the focuses of attention. Especially in the strategically important Dongjiang River Basin, which is the source of water for three world-class cities of Hong Kong, Shenzhen and Guangzhou. Hence, it is important to quantitatively analyze the spatial and temporal variation of water resources at the basin and ecosystem scale of Dongjiang river basin. The Soil and Water Assessment Tool (SWAT) was used in this study to analyze the spatial and temporal dynamics of typical ecosystem water resources in the Dongjiang River basin over the last 50 years. The objectives of the study were to: (1) identify the distribution of water resources in typical ecosystems; (2) determine the quantity of water in the forest ecosystem; and (3) characterize the spatial and temporal changes. The findings revealed that: (1) the average water resources in the Dongjiang River Basin from 1971 to 2020 were 37.944 Bm³, and there were obvious changes in the basin's total quantity of water resources. The most important climatic factors influencing water resources were precipitation and temperature. (2) From 1971 to 2020, the total water resources of typical Donjiang River Basin ecosystems exhibited a declining trend. The total average water resources of the forest ecosystem were 27.532 Bm³, which first increased and then decreased from upstream to downstream. The total quantity of water resources in the grassland ecosystem increases with the amount of water per unit area. (3) The change in total water resources has been decreasing over the last decade, which is consistent with the change in the water resources of the forest ecosystem in the Dongjiang River Basin. Changes in water resources were greater in the lower and middle reaches than in the upper reaches. Changes in land use were a significant factor in the evolution of water resources in typical ecosystems. This study can provide theoretical support for the comprehensive management of typical ecosystems and water resources in the Dongjiang River Basin.
Climate change is one of the biggest challenges facing humanity that affects the sciences related to nature and environment. Changes in factors affecting climate change affect the hydrological responses of watersheds and this phenomenon affects the quantity and quality of water resources such as lakes, reservoirs and dams. Today, the development and progress of computer and computing technologies has had a significant impact on hydrological modeling. In recent years, conceptual and physical models based on the characteristics of watersheds for hydrological systems have attracted the attention of researchers. Due to the emergence of such developments, there will be complications in the process of hydrological modeling. Correct management forecasts in the field of water resources due to the lack of water resources, is one of the topics of interest in the institutions in charge of water resources management, which should use high-precision simulator models for long-term estimates and planning. Considering the crisis and shortage of water resources in Iran, the sub-catchment upstream of Zayandeh Rood Dam was selected as a case study in this research, and the hydrological processes in the sub-basins of Zayandeh Rood Dam were simulated and the inflow to the reservoir of Zayandeh Rood Dam was estimated. For this purpose, the water and soil assessment model SWAT has been used to simulate the rainfall-runoff process and finally to estimate the inflow to the reservoir. Based on the simulated runoff from the three sub-basins of Buin-Damneh, Qala Shahrokh-Chelgerd and Chadegan-Ceshmeh, the model showed acceptable results for the simulation in the three sub-basins mentioned by the model, and it was estimated from the algebraic sum of the three inflows that the mentioned error coefficients are equal to 0.86 has been obtained. This result shows the high accuracy of the models in the simulation
Assessing ecosystem services (ESs) is vital for the environment development. This study explored the variation characteristics of ESs under the restoring vegetation since the Grain for Green Project (GFGP) in Jinghe Watershed, which is a typical watershed experiencing severe soil erosion in the Loess Plateau of China. Herein, the fractional vegetation cover (FVC) is selected as an indicator of the changes in vegetation restoration from 2000 to 2020. Simulations of ESs are based on the Soil and Water Assessment Tool (SWAT), Carnegie Ames-Stanford Approach (CASA) model, Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST). This paper analyzed the spatiotemporal heterogeneity in vegetation coverage and the response of vegetation coverage to ESs in the Jinghe Watershed with Mann-Kendall test, Pearson correlation coeffificient R and the gray relational analysis (GRA). The results showed that: (1) Since the implementation of the GFGP, the area of forests and grassland has increased significantly. The proportion of vegetation restoration area was 95.15 %; (2) At the whole watershed scale, vegetation restoration and soil conservation (SC), water yied (WY), net primary production (NPP) and habitat quality (HQ) were synergistic relationships, while focusing on the sub-watershed scale, vegetation restoration and SC and HQ were transformed into trade-off relationships in some sub-watershed. (3) Vegetation restoration in the Jinghe Watershed had positive effects on SC, WY, NPP, and HQ. The order of relevance with vegetation restoration was: SC > HQ > NPP > WY, indicating that vegetation restoration in the Jinghe Watershed has the highest closeness to SC. This study showed that since the implementation of the GFGP, vegetation coverage in the Jinghe Watershed has significantly restored, which has promoted the improvement of ESs.
Global sensitivity analysis (GSA) has long been recognized as an indispensable tool for model analysis. GSA has been extensively used for model simplification, identifiability analysis, and diagnostic tests. Nevertheless, computationally efficient methodologies are needed for GSA, not only to reduce the computational overhead, but also to improve the quality and robustness of the results. This is especially the case for process‐based hydrologic models, as their simulation time typically exceeds the computational resources available for a comprehensive GSA. To overcome this computational barrier, we propose a data‐driven method called VISCOUS, variance‐based sensitivity analysis using copulas. VISCOUS uses Gaussian mixture copulas to approximate the joint probability density function of a given set of input‐output pairs for estimating the variance‐based sensitivity indices. Our method identifies dominant hydrologic factors by recycling existing input‐output data, and thus can deal with arbitrary sample sets drawn from the input‐output space. We used two hydrologic models of increasing complexity (HBV and VIC) to assess the performance of VISCOUS. Our results confirm that VISCOUS and the conventional variance‐based method can detect similar important and unimportant factors. Furthermore, the VISCOUS method can substantially reduce the computational cost required for sensitivity analysis. Our proposed method is particularly useful for process‐based models with many uncertain parameters, large domain size, and high spatial and temporal resolution.
Change of flow
Anthropogenic influence on climate has changed temperatures, precipitation, atmospheric circulation, and many other related physical processes, but has it changed river flow as well? Gudmundsson et al. analyzed thousands of time series of river flows and hydrological extremes across the globe and compared them with model simulations of the terrestrial water cycle (see the Perspective by Hall and Perdigão). They found that the observed trends can only be explained if the effects of climate change are included. Their analysis shows that human influence on climate has affected the magnitude of low, mean, and high river flows on a global scale.
Science , this issue p. 1159 ; see also p. 1096
Terrestrial water storage (TWS) modulates the hydrological cycle and is a key determinant of water availability and an indicator of drought. While historical TWS variations have been increasingly studied, future changes in TWS and the linkages to droughts remain unexamined. Here, using ensemble hydrological simulations, we show that climate change could reduce TWS in many regions, especially those in the Southern Hemisphere. Strong inter-ensemble agreement indicates high confidence in the projected changes that are driven primarily by climate forcing rather than land and water management activities. Declines in TWS translate to increases in future droughts. By the late twenty-first century, the global land area and population in extreme-to-exceptional TWS drought could more than double, each increasing from 3% during 1976–2005 to 7% and 8%, respectively. Our findings highlight the importance of climate change mitigation to avoid adverse TWS impacts and increased droughts, and the need for improved water resource management and adaptation. Projections of terrestrial water storage (TWS)—the sum of all continental water—are key to water resource and drought estimates. A hydrological model ensemble predicts climate warming will more than double the land area and population exposed to extreme TWS drought by the late twenty-first century.
Land models are increasingly used in terrestrial hydrology due to their process-oriented representation of water and energy fluxes. A priori specification of the grid size of the land models is typically defined based on the spatial resolution of forcing data, the modeling objectives, the available geospatial information, and computational resources. The variability of the inputs, soil types, vegetation covers, and forcing is masked or aggregated based on the a priori grid size. In this study, we propose an alternative vector-based implementation to directly configure a land model using unique combinations of land cover types, soil types, and other desired geographical features that have hydrological significance, such as elevation zone, slope, and aspect. The main contributions of this paper are to (1) implement the vector-based spatial configuration using the Variable Infiltration Capacity (VIC) model; (2) illustrate how the spatial configuration of the model affects simulations of basin-average quantities (i.e., streamflow) as well as the spatial variability of internal processes (snow water equivalent, SWE, and evapotranspiration, ET); and (3) describe the work and challenges ahead to improve the spatial structure of land models. Our results show that a model configuration with a lower number of computational units, once calibrated, may have similar accuracy to model configurations with more computational units. However, the different calibrated parameter sets produce a range of, sometimes contradicting, internal states and fluxes. To better address the shortcomings of the current generation of land models, we encourage the land model community to adopt flexible spatial configurations to improve model representations of fluxes and states at the scale of interest.
The choice of hydrological model structure, i.e. a model’s selection of states and fluxes and the equations used to describe them, strongly controls model performance and real-ism. This work investigates differences in performance of 36 lumped conceptual model structures calibrated to and evaluated on daily streamflow data in 559 catchments across the United States. Model performance is compared against a benchmark that accounts for the seasonality of flows in each catchment. We find that our model ensemble struggles to beat the benchmark in snow-dominated catchments. In most other catchments model structure equifinality (i.e. cases where different models achieve similar high efficiency scores) can be very high. We find no relation between the number of model parameters and performance during either calibration or evaluation periods, nor evidence of increased risk of overfitting for models with more parameters. Instead, the choice of model parametrization (i.e. which equations are used and how parameters are used within them) dictates the model’s strengths and weaknesses. Results suggest that certain model structures are inherently better suited for certain objective functions and thus for certain study purposes. We find no clear relationships between the catchments where any model performs well and descriptors of those catchments’ geology, topography, soil and vegetation characteristics. In-stead, model suitability seems to relate strongest to the streamflow regime each catchment generates and we have formulated several tentative hypotheses that relate commonalities in model structure to similarities in model performance. Modeling results are made publicly available for further investigation.
This research aimed to appraise the performance of the Soil and Water Assessment Tool (SWAT) model in sediment flow simulation and also to investigate the uncertainty of the model in the watershed areas of arid and semi-arid regions. In this survey, we used the Sequential Uncertainty Fitting ver.2 (SUFI-2) algorithm to assess the uncertainty and calibrate the model. Different factors of water resources are simulated, and we consider the crop yield and water quality at the Hydrological Response Units (HRU) level. Besides, to quantify the water resources, we implemented monthly time intervals. Also, we used monthly time intervals to quantify the water resources. The results showed that in a 3-year validation period (2007–2010), the P-factor and the r-factor were 0.28 and 0.38 respectively, while in the 7-year calibration period (2000–2006), these two factors were 0.29 and 0.39, respectively. The findings of this study proved that in the validation period, statistical indicators of model evaluation comprise R2, bR2, and the Nash–Sutcliffe efficiency (NSE) coefficients were 0.85, 0.23, and 0.47, respectively, while in the calibration period, these coefficients were 0.46, 0.14, and 0.37, respectively. The results of uncertainty and calibration analysis were acceptable, but in the validation phase, the model has been more applicable and useful. These results show the acceptable efficiency of the SWAT model in simulating the sediment load of the study area. To assess the sensitivities of 22 input parameters, we used SWAT Calibration Uncertainties Program (SWAT-CUP) and achieved three of the most sensitive parameters comprising CN2, SOL_BD, and USLE_P. In contrast, the parameters with the least sensitivity were SLSUBBSN, GW_DELAY, and ESCO. We can use the calibrated model as inputs for SWAT, to assess the impact of climate change on soil erosion.
Model parameter calibration is a fundamentally important stage that must be completed before applying a model to address practical problems. In this study, we describe an automatic calibration framework that combines sensitivity analysis (SA) and an adaptive surrogate modeling‐based optimization (ASMO) algorithm. We use this framework to calibrate catchment‐specific sensitive parameters for streamflow simulation in the variable infiltration capacity (VIC) model with a 0.25° spatial resolution over 10 major river basins of China from 1960 to 1979. We found that three parameters—the infiltration parameter (B) and two of the soil depth parameters (D1, D2)—are highly sensitive in most basins, while other parameter sensitivities are strongly related to the dynamic environment of the basin. Compared with directly calibrating the seven parameters recommended for the default calibration procedure, our framework not only reduced the computing time by two thirds through opting out of insensitive parameters (type I error) but also improved the Nash‐Sutcliffe model efficiency coefficient (NSE) for optimized results when it identified a missing sensitive parameter (type II error) in the case study river basins. Results show that the SA‐based ASMO framework is an effective and efficient model‐optimization technique for matching simulated streamflow with observations across China. The NSE for monthly streamflow ranged from 0.75 to 0.97 and from 0.71 to 0.97 during the validation and calibration periods, respectively. The calibrated parameters can be applied directly in streamflow simulations across China, and the proposed calibration framework holds important implications for relevant simulation studies in other regions.
Saturated hydraulic conductivity (Ks) is a fundamental soil property that regulates the fate of water in soils. Its measurement, however, is cumbersome and instead pedotransfer functions (PTFs) are routinely used to estimate it. Despite much progress over the years, the performance of current generic PTFs estimating Ks remains poor. Using machine learning, high‐performance computing, and a large database of over 18,000 soils, we developed new PTFs to predict Ks. We compared the performances of four machine learning algorithms and different predictor sets. We evaluated the relative importance of soil properties in explaining Ks. PTF models based on boosted regression tree algorithm produced the best models with root‐mean‐squared log‐transformed error in ranges of 0.4 to 0.3 (log10(cm/day)). The 10th percentile particle diameter (d10) was found to be the most important predictor followed by clay content, bulk density (ρb), and organic carbon content (C). The sensitivity of Ks to soil structure was investigated using ρb and C as proxies for soil structure. An inverse relationship was observed between ρb and Ks, with the highest sensitivity at around 1.8 g/cm3 for most textural classes. Soil C showed a complex relationship with Ks with an overall positive relation for fine‐textured and midtextured soils but an inverse relation for coarse‐textured soils. This study sought to maximize the extraction of information from a large database to develop generic machine learning‐based PTFs for estimating Ks. Models developed here have been made publicly available and can be readily used to predict Ks.
Hydrological models demand large number of input parameters, which are to be optimally identified for better simulation of various hydrological processes. Identifying the most relevant parameters and their values using efficient sensitivity analysis methods helps to better understand model performance. In this study, the physically-based, distributed model SHETRAN is used for hydrological simulation on the Netravathi River Basin in south India and the most important parameters are identified using the Morris screening method. Further, the influence of a particular model parameter on streamflow is quantified using local sensitivity analysis and optimal parameters are obtained for calibration of the SHETRAN model. The results demonstrate the capability of two-stage sensitivity analysis, combining qualitative and quantitative methods in the initial screening-out of insignificant model parameters, identifying parameter interactions and quantifying the contribution of each model parameter to the streamflow. The results of sensitivity analysis simplified the calibration procedure of SHETRAN for the study area.
The main objective of this study was to calibrate and validate the eco-hydrological model Soil and Water Assessment Tool (SWAT) with satellite-based actual evapotranspiration (AET) data from the Global Land Evaporation Amsterdam Model (GLEAM_v3.0a) and from the Moderate Resolution Imaging Spectroradiometer Global Evaporation (MOD16) for the Ogun River Basin (20 292 km²) located in southwestern Nigeria. Three potential evapotranspiration (PET) equations (Hargreaves, Priestley–Taylor and Penman–Monteith) were used for the SWAT simulation of AET. The reference simulations were the three AET variables simulated with SWAT before model calibration took place. The sequential uncertainty fitting technique (SUFI-2) was used for the SWAT model sensitivity analysis, calibration, validation and uncertainty analysis. The GLEAM_v3.0a and MOD16 products were subsequently used to calibrate the three SWAT-simulated AET variables, thereby obtaining six calibrations–validations at a monthly timescale. The model performance for the three SWAT model runs was evaluated for each of the 53 subbasins against the GLEAM_v3.0a and MOD16 products, which enabled the best model run with the highest-performing satellite-based AET product to be chosen. A verification of the simulated AET variable was carried out by (i) comparing the simulated AET of the calibrated model to GLEAM_v3.0b AET, which is a product that has different forcing data than the version of GLEAM used for the calibration, and (ii) assessing the long-term average annual and average monthly water balances at the outlet of the watershed. Overall, the SWAT model, composed of the Hargreaves PET equation and calibrated using the GLEAM_v3.0a data (GS1), performed well for the simulation of AET and provided a good level of confidence for using the SWAT model as a decision support tool. The 95 % uncertainty of the SWAT-simulated variable bracketed most of the satellite-based AET data in each subbasin. A validation of the simulated soil moisture dynamics for GS1 was carried out using satellite-retrieved soil moisture data, which revealed good agreement. The SWAT model (GS1) also captured the seasonal variability of the water balance components at the outlet of the watershed.
This study demonstrated the potential to use remotely sensed evapotranspiration data for hydrological model calibration and validation in a sparsely gauged large river basin with reasonable accuracy. The novelty of the study is the use of these freely available satellite-derived AET datasets to effectively calibrate and validate an eco-hydrological model for a data-scarce catchment.
The spatial‐temporal changes in terrestrial water storage (TWS) over the Tibetan Plateau (TP) and six selected basins during 2003–2014 were analyzed by applying the Gravity Recovery and Climate Experiment data and the extended Variable Infiltration Capacity‐glacier model, including the upstream of Yangtze (UYA), Yellow (UYE), Brahmaputra (UB), and Indus river basins and the Inner TP and the Qaidam Basin. The possible causes of TWS changes were investigated from the perspective of surface water balance and TWS components through multisource data and the Variable Infiltration Capacity‐glacier model. There was a strong spatial heterogeneity in changes of Gravity Recovery and Climate Experiment TWS in the TP—with apparent mass accumulation in central and northern TP and a sharp decreasing trend in southern and northwestern TP. The TWS changes in the TP were mostly attributed to variations in precipitation and evapotranspiration from the perspective of land‐surface water balance. Precipitation played a dominant role on the TWS accumulation in the UYA and UYE, while evapotranspiration had a more important role than precipitation in TWS depletion in the UB. From the perspective of TWS components, the TWS increase in the UYA and UYE was mainly caused by an increase in soil moisture, whereas the decrease in TWS in the UB was mostly due to glacier mass loss. TWS was accumulating from March through August in southeastern TP while from November to April/May in northwestern TP. The seasonal variations of TWS are highly modulated by the large‐scale climate system, atmospheric moisture flux, and precipitation regime over the TP.
Quantifying and reducing uncertainties in physics-based hydrological model parameters will improve model reliability for hydrological forecasting. We present an uncertainty quantification framework that combines the strengths of stepwise sensitivity analysis and adaptive surrogate-based multi-objective optimization to facilitate practical assessment and reduction of model parametric uncertainties. Framework performance was tested using the distributed hydrological model Coupled Routing and Excess Storage (CREST) for daily streamflow simulation over ten watersheds. By identifying sensitive parameters stepwisely, we reduced the number of parameters requiring calibration from twelve to seven, thus limiting the dimensionality of calibration problem. By updating surrogate models adaptively, we found the optimal sets of sensitive parameters with the surrogate-based multiobjective optimization. The calibrated CREST was able to satisfactorily simulate observed streamflow for all watersheds, improving one minus Nash-Sutcliffe efficiency (1−NSE) by 65–90% and percentage absolute relative bias (|RB|) by 60–95% compared to the default. The validation result demonstrated that the calibrated CREST was also able to reproduce observed streamflow outside the calibration period, improving 1−NSE by 40–85% and |RB| by 35–90% compared to the default. Overall, this uncertainty quantification framework is effective for assessment and reduction of model parametric uncertainties, the results of which improve model simulations and enhance understanding of model behaviors.
In this study, a semi-distributed hydrologic model Soil and Water Assessment Tool (SWAT) has been employed for the Karnali River basin, Nepal to test its applicability for hydrological simulation. Further, model was evaluated to carry out the water balance study of the basin and to determine the snowmelt contribution in the river flow. Snowmelt Runoff Model (SRM) was also used to compare the snowmelt runoff simulated from the SWAT model. The statistical results show that performance of the SWAT model in the Karnali River basin is quite good (p-factor = 0.88 and 0.88, for daily calibration and validation, respectively; r-factor = 0.76 and 0.71, for daily calibration and validation, respectively). Baseflow alpha factor (ALPHA_BF) was found most sensitive parameter for the flow simulation. The study revealed that the average annual runoff volume available at the basin outlet is about 47.16 billion cubic metre out of which about 12% of runoff volume is contributed by the snowmelt runoff. About 25% of annual precipitation seems to be lost as evapotranspiration. The results revealed that both the models, SWAT and SRM, can be efficiently applied in the mountainous river basins of Nepal for planning and management of water resources.
Inappropriate use of land and poor ecosystem management have accelerated land
degradation and reduced the storage capacity of reservoirs. To mitigate the effect of the increased
sediment yield, it is important to identify erosion-prone areas in a 287 km2 catchment in Ethiopia.
The objectives of this study were to: (1) assess the spatial variability of sediment yield; (2) quantify
the amount of sediment delivered into the reservoir; and (3) prioritize sub-catchments for watershed
management using the Soil and Water Assessment Tool (SWAT). The SWAT model was calibrated
and validated using SUFI-2, GLUE, ParaSol, and PSO SWAT-CUP optimization algorithms. For
most of the SWAT-CUP simulations, the observed and simulated river discharge were not
significantly different at the 95% level of confidence (95PPU), and sources of uncertainties were
captured by bracketing more than 70% of the observed data. This catchment prioritization study
indicated that more than 85% of the sediment was sourced from lowland areas (slope range: 0–8%)
and the variation in sediment yield was more sensitive to the land use and soil type prevailing in
the area regardless of the terrain slope. Contrary to the perception of the upland as an important
source of sediment, the lowland in fact was the most important source of sediment and should be
the focus area for improved land management practice to reduce sediment delivery into storage
reservoirs. The research also showed that lowland erosion-prone areas are typified by extensive
agriculture, which causes significant modification of the landscape. Tillage practice changes the
infiltration and runoff characteristics of the land surface and interaction of shallow groundwater
table and saturation excess runoff, which in turn affects the delivery of water and sediment to the
reservoir and catchment evapotranspiration.
This study assessed the performances of the traditional temperature-index snowmelt runoff model (SRM) and an SRM model with a finer zonation based on aspect and slope (SRM+AS model) in a data-scarce mountain watershed in the Urumqi River Basin, in Northwest China. The proposed SRM+AS model was used to estimate the melt rate with the degree-day factor (DDF) through the division of watershed elevation zones based on aspect and slope. The simulation results of the SRM+AS model were compared with those of the traditional SRM model to identify the improvements of the SRM+AS model’s performance with consideration of topographic features of the watershed. The results show that the performance of the SRM+AS model has improved slightly compared to that of the SRM model. The coefficients of determination increased from 0.73, 0.69, and 0.79 with the SRM model to 0.76, 0.76, and 0.81 with the SRM+AS model during the simulation and validation periods in 2005, 2006, and 2007, respectively. The proposed SRM+AS model that considers aspect and slope can improve the accuracy of snowmelt runoff simulation compared to the traditional SRM model in mountain watersheds in arid regions by proper parameterization, careful input data selection, and data preparation.
Satellite-based and reanalysis quantitative precipitation estimates are attractive for hydrologic prediction or forecasting and reliable water resources management, especially for ungauged regions. This study evaluates three widely used global high-resolution precipitation products (PERSIANN-CDR, TRMM 3B42V7, and NCEP-CFSR) against gauge observations with seven statistical indices over two humid regions in China. Furthermore, the study investigates whether the three precipitation products can be reliably utilized as inputs in SWAT, a semi-distributed hydrological model, to simulate streamflows. Results show that the precipitation estimates derived from TRMM 3B42V7 outperform the other two products with the smallest errors and bias, and highest correlation at monthly scale, which is followed by PERSIANN-CDR and NCEP-CFSR in this rank. However, the superiority of TRMM 3B42V7 in errors, bias and correlations is not warranted at daily scale. PERSIANN-CDR and 3B42V7 present encouraging potential for streamflow prediction at daily and monthly scale respectively over the two humid regions, while the performance of NCEP-CFSR for hydrological applications varies from basin to basin. Simulations forced with 3B42V7 are the best among the three precipitation products in capturing daily measured streamflows, whilst PERSIANN-CDR-driven simulations underestimate high streamflows and high streamflow simulations driven by NCEP-CFSR mostly are overestimated significantly. In terms of extreme events analysis, PERSIANN-CDR often underestimates the extreme precipitation, so do extreme streamflow simulations forced with it. NCEP-CFSR performs just the reverse, compared with PERSIANN-CDR. The performance pattern of TRMM 3B42V7 on extremes is not certain, with coexisting underestimation and overestimation.
Performance measures (PMs) and corresponding performance evaluation criteria (PEC) are important as-pects of calibrating and validating hydrologic and water quality models and should be updated with advances in modeling science. We synthesized PMs and PEC from a previous special collection, performed a meta-analysis of performance data reported in recent peer-reviewed literature for three widely published watershed-scale models (SWAT, HSPF, WARMF), and one field-scale model (ADAPT), and provided guidelines for model performance evaluation. Based on the synthesis, meta-analysis, and personal modeling experiences, we recommend coefficient of determination (R2; in conjunction with gradient and intercept of the corresponding regression line), Nash Sutcliffe efficiency (NSE), index of agreement (d), root mean square error (RMSE; alongside the ratio of RMSE and standard deviation of measured data, RSR), percent bias (PBIAS), and several graphical PMs to evaluate model performance. We recommend that model performance can be judged “satisfactory” for flow simulations if monthly R2 > 0.70 and d > 0.75 for field-scale models, and daily, monthly, or annual R2 > 0.60, NSE > 0.50, and PBIAS ≤±15% for watershed-scale models. Model performance at the watershed scale can be evaluated as “satisfactory” if monthly R2 > 0.40 and NSE > 0.45 and daily, monthly, or annual PBIAS ≤±20% for sediment; monthly R2 > 0.40 and NSE > 0.35 and daily, monthly, or annual PBIAS ≤±30% for phosphorus (P); and monthly R2 > 0.30 and NSE > 0.35 and daily, monthly, or annual PBIAS ≤±30% for nitrogen (N). For RSR, we rec-ommend that previously published PEC be used as detailed in this article. We also recommend that these PEC be used primarily for the four models for which there were adequate data, and used only with caution for other models. These PEC can be adjusted within acceptable bounds based on additional considerations, such as quality and quantity of avail-able measured data, spatial and temporal scales, and project scope and magnitude, and updated based on the framework presented herein. This initial meta-analysis sets the stage for more comprehensive meta-analysis to revise PEC as new PMs and more data become available.
Hydrological models play vital roles in management of water resources. However, the calibration of the hydrological models is a large challenge because of the uncertainty involved in the large number of parameters. In this study, four uncertainty analysis methods, including Generalized Likelihood Uncertainty Estimation (GLUE), Parameter Solution (ParaSol), Particle Swarm Optimization (PSO), and Sequential Uncertainty Fitting (SUFI-2), were employed to perform parameter uncertainty analysis of streamflow simulation in the Srepok River Catchment by using the Soil and Water Assessment Tool (SWAT) model. The four methods were compared in terms of the model prediction uncertainty, the model performance, and the computational efficiency. The results showed that the SUFI-2 method has the advantages in the model calibration and uncertainty analysis. This technique could be run with the smallest of simulation runs to achieve good prediction uncertainty bands and model performance. This technique could be run with the smallest of simulation runs to achieve good prediction uncertainty bands and model performance.
Uncertainty quantification (UQ) of environmental dynamic models requires an efficient way to extract the information about the relationship between input parameter and model output. A uniformly scattered sample set is generally preferred over crude Monte Carlo sampling for its ability to explore the parameter space more effectively and efficiently. This paper compares eight commonly used uniform sampling methods along with the crude Monte Carlo sampling. The efficiency is measured by six uniformity metrics, while the effectiveness is measured by the goodness-of-fit of the surrogate models, and the sensitivity analysis and optimization results. We used two test problems: the Sobol' g-function and the SAC-SMA hydrological model. The results show that among the sampling methods evaluated, the Good Lattice Points (GLP) and Symmetric Latin hypercube (SLH) have the highest uniformity scores, and the Ranked Gram-Schmidt (RGS) de-correlation algorithm can further improve the uniformity of the lattice sample sets. On the other hand, the Quasi-Monte-Carlo (QMC) methods, such as Halton and Sobol' sequences, are not as uniform as their theoretical potential suggests when the number of sample points is low. Further, we found no clear relationship between the sampling methods used and their effectiveness, as the latter is affected by many factors other than the sampling methods, such as the choice of the surrogate modeling methods, sensitivity analysis and optimization methods, and the intrinsic properties of the dynamic models.
A combination of driving forces are increasing pressure on local, national, and regional water supplies needed for irrigation, energy production, industrial uses, domestic purposes, and the environment. In many parts of Europe groundwater quantity, and in particular quality, have come under sever degradation and water levels have decreased resulting in negative environmental impacts. Rapid improvements in the economy of the eastern European block of countries and uncertainties with regard to freshwater availability create challenges for water managers. At the same time, climate change adds a new level of uncertainty with regard to freshwater supplies. In this research we build and calibrate an integrated hydrological model of Europe using the Soil and Water Assessment Tool (SWAT) program. Different components of water resources are simulated and crop yield and water quality are considered at the Hydrological Response Unit (HRU) level. The water resources are quantified at subbasin level with monthly time intervals. Leaching of nitrate into groundwater is also simulated at a finer spatial level (HRU). The use of large-scale, high-resolution water resources models enables consistent and comprehensive examination of integrated system behavior through physically-based, data-driven simulation. In this article we discuss issues with data availability, calibration of large-scale distributed models, and outline procedures for model calibration and uncertainty analysis. The calibrated model and results provide information support to the European Water Framework Directive and lay the basis for further assessment of the impact of climate change on water availability and quality. The approach and methods developed are general and can be applied to any large region around the world.
Proper specification of model parameters is critical to the performance of land surface models (LSMs). Due to high dimensionality and parameter interaction, estimating parameters of an LSM is a challenging task. Sensitivity analysis (SA) is a tool that can screen out the most influential parameters on model outputs. In this study, we conducted parameter screening for six output fluxes for the Common Land Model: sensible heat, latent heat, upward longwave radiation, net radiation, soil temperature and soil moisture. A total of 40 adjustable parameters were considered. Five qualitative SA methods, including local, sum-of-trees, multivariate adaptive regression splines, delta test and Morris methods, were compared. The proper sampling design and sufficient sample size necessary to effectively screen out the sensitive parameters were examined. We found that there are 2-8 sensitive parameters, depending on the output type, and about 400 samples are adequate to reliably identify the most sensitive parameters. We also employed a revised Sobol' sensitivity method to quantify the importance of all parameters. The total effects of the parameters were used to assess the contribution of each parameter to the total variances of the model outputs. The results confirmed that global SA methods can generally identify the most sensitive parameters effectively, while local SA methods result in type I errors (i.e., sensitive parameters labeled as insensitive) or type II errors (i.e., insensitive parameters labeled as sensitive). Finally, we evaluated and confirmed the screening results for their consistency with the physical interpretation of the model parameters.
The level of model complexity that can be effectively supported by available information has long been a subject of many studies in hydrologic modeling. In particular, distributed parameter models tend to be regarded as overparameterized because of numerous parameters used to describe spatially heterogeneous hydrologic processes. However, it is not clear how parameters and observations influence the degree of overparameterization, equifinality of parameter values, and uncertainty. This study investigated the impact of the numbers of observations and parameters on calibration quality including equifinality among calibrated parameter values, model performance, and output/parameter uncertainty using the SWAT model. In the experiments, the number of observations was increased by expanding calibration period or by including measurements made at inner points of a watershed. Similarly, additional calibration parameters were included in the order of their sensitivity. Then, unique sets of parameters were calibrated with the same objective function, optimization algorithm, and stopping criteria but the different numbers of observations. The calibration quality was quantified with statistics calculated based on the ‘behavioral’ parameter sets, identified using 1 and 5% cut-off thresholds in a GLUE framework. The study demonstrated that equifinality, model performance, and output/parameter uncertainty were responsive to the numbers of observations and calibration parameters; however the relationship between the numbers, equifinality, and uncertainty was not always conclusive. Model performance improved with increased numbers of calibration parameters and observations, and substantial equifinality did neither necessarily mean bad model performance nor large uncertainty in the model outputs and parameters. This article is protected by copyright. All rights reserved.
The Soil and Water Assessment Tool (SWAT) model has emerged as one of the most widely used water quality watershed- and river basin-scale models worldwide, applied extensively for a broad range of hydrologic and/or environmental problems. The international use of SWAT can be attributed to its flexibility in addressing water resource problems, extensive networking via dozens of training workshops and the several international conferences that have been held during the past decade, comprehensive online documentation and supporting software, and an open source code that can be adapted by model users for specific application needs. The catalyst for this special collection of papers was the 2011 International SWAT Conference & Workshops held in Toledo, Spain, which featured over 160 scientific presentations representing SWAT applications in 37 countries. This special collection presents 22 specific SWAT-related studies, most of which were presented at the 2011 SWAT Conference; it represents SWAT applications on five different continents, with the majority of studies being conducted in Europe and North America. The papers cover a variety of topics, including hydrologic testing at a wide range of watershed scales, transport of pollutants in northern European lowland watersheds, data input and routing method effects on sediment transport, development and testing of potential new model algorithms, and description and testing of supporting software. In this introduction to the special section, we provide a synthesis of these studies within four main categories: (i) hydrologic foundations, (ii) sediment transport and routing analyses, (iii) nutrient and pesticide transport, and (iv) scenario analyses. We conclude with a brief summary of key SWAT research and development needs.
Copyright © by the American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America, Inc.
Expansion of a detailed hydrological modelling from catchment- to national- or continental-scale is of important theoretical and practical significance, yet faces many challenges. We developed a high-resolution, physically-based hydrological model named WEP-CN (Water and Energy transfer Processes in China) for China. Different climatic and hydrological conditions and geological structures as well as their impact on infiltration and runoff were examined and incorporated in WEP-CN. The study area (9.6 million km²) was divided into 19,406 sub-watersheds and 81,687 contour belts. The computation units, model input, and model structure and parameters in WEP-CN were improved to obtain a well-defined simulation area, a more reliable input as well as a detailed description of soil moisture movement in several special vadose zones such as karst development, swelling and frozen soil. Continuous simulations of various natural hydrological processes were conducted for 62 years from 1956 to 2017. We demonstrated the efficacy of our model by comparing simulated and statistical monthly streamflow at 203 hydrological stations across the country. For the validation period of 1981-2000, the Nash-Sutcliff Efficiency (NSE) was found to be larger than 0.7 at 80% of the stations, and the absolute value of relative error (RE) was less than 10% at 95% of the stations. Our result highlights the benefit of incorporating new mechanisms on the special vadose zone water movement and accounts for the impact of elevation change on meteorological and vegetation variables. This paper can serve as a reference for large-scale hydrological simulation with diverse climate, topography and underlying surface conditions.
Appropriate soil property data are important inputs for the development of hydrological models, and as the scale of the study decreases, the relevance of these data increases. Watershed soil surveys require time and money, so readily available data are often used. This study compares the hydrological and nitrate loading prediction performance of the globally available dataset SoilGrids and a dataset prepared from local legacy data by a random forest approach. The study was carried out using the SWAT model for a 33 km 2 watershed in the Czech Republic. Model performance was tested by applying a Latin hypercube-sampled parameter set to both a model using the SoilGrids data as input, and another model based on the local soil dataset. The SoilGrids dataset generally shows finer soil texture and higher organic carbon content than the dataset based on local legacy data. These differences are also reflected in the hydrological process predictions, where the SoilGrids-based model produced less lateral flow than that of the local dataset, with the compensation of higher evapotranspiration. This difference in water balance led to worse nitrate loading estimation by the SoilGrids-based model. The results of this case study suggest that the use of a local dataset is still more appropriate, despite the availability of global data with mid resolution.
Development of robust approaches for calibrating daily rainfall-runoff models to monthly streamflow data is of major practical interest. Such approaches would enable widely used hydrological modelling platforms operating at a daily time step to be applied in practical situations where daily precipitations can be obtained but only monthly streamflow data are available (e.g. predicting inflows into large dams). This study compares the performance of a daily and monthly calibrated hydrological model running at daily time step (GR4J) using a wide range of metrics: fit of the daily and monthly flow duration curve, daily and monthly pattern metrics, and long-term bias. The comparison is undertaken for 508 Australian catchments, two evaluation periods and four objective functions (including sum-of-squared-errors of Box-Cox transformed streamflow and the Kling-Gupta efficiency). Monthly calibration performs similar or better than daily calibration in a majority of sites and periods in terms of bias and fit of the flow duration curve. This result holds even when the flow duration curve is computed at the daily time step, which constitutes a major finding of this study. However, performance of monthly calibration is worse than daily calibration for daily pattern metrics such as Nash-Sutcliffe efficiency in a majority of sites and periods. This performance loss can be reduced significantly by using regionalised values for the flow-timing parameter of GR4J. Similar results are obtained for other pattern metrics and all objective functions. These findings suggest that monthly calibration of rainfall-runoff models to daily-rainfall/monthly-streamflow is a viable alternative to daily calibration when no daily streamflow data are available.
Kangsabati basin located in tropical plateau region faces multiple problems of soil erosion susceptibility (SES), soil fertility deterioration, and sedimentation in reservoirs. Hence, identification of SES zones in thirty-eight sub-basins (SB) for basin prioritization is necessary. The present research addressed the issue by using four multi-criteria decision-making (MCDM) models: VlseKriterijumska optimizacija I Kompromisno Resenje (VIKOR), technique for order preference by similarity to ideal solution (TOPSIS), simple additive weighing (SAW), compound factor (CF). To determine the best fitted method from MCDM for erosion susceptibility (ES), a comparison has been made with Soil and Water Assessment Tool (SWAT), where fifteen morphometric parameters were considered for MCDM, and meteorological data, soil, slope and land use land cover (LULC) were considered for SWAT model. Two validation indices of percentage change and intensity change were used for evaluation and comparison of MCDM results. With SWAT model performance, SWAT calibration and uncertainty analysis programs (CUP) was used for sensitive analysis of SWAT parameters on flow discharge and sediment load simulation. The results showed that 23, 16, 18 SB have high ES; therefore they were given 1 to 3 ranks, whereas 31, 37, 21SB have low ES, hence given 38 to 36 rank as predicted by MCDM methods and SWAT. MCDM validation results depict that VIKOR and CF methods are more acceptable than TOPSIS and SAW. Calibration (flow discharge R² 0.86, NSE 0.75; sediment load R² 0.87, NSE 0.69) and validation (flow discharge R² 0.79, NSE 0.55; sediment load R² 0.79, NSE 0.76) of SWAT model indicated that simulated results are well fitted with observed data. Therefore, VIKOR reflects the significant role of morphometric parameters on ES, whereas SWAT reflects the significant role of LULC, slope, and soil on ES. However, it could be concluded that VIKOR is more effective MCDM method in comparison to SWAT prediction.
Soil and Water Assessment Tool (SWAT) is one of the most widely used semi-distributed hydrological models. Assessment of the uncertainties of SWAT outputs is a common but challenging topic due to the more significant number of parameters. The purpose of this study is to investigate the use of Polynomial Chaos Expansion (PCE) in assessing uncertainty propagation in SWAT under the impact of significant parameter sensitivity. Furthermore, for the first time, a machine learning technique (i.e., artificial neural network, ANN) is integrated with PCE to expand its capability in assessing the uncertainty for probabilistic forecasting of daily flow. The traditional PCE and the proposed PCE-ANN method are applied to a case study in the Guadalupe watershed in Texas, USA to assess the uncertainty propagation in SWAT for flow prediction during the historical and forecasting periods. The results show that PCE provides similar results as the traditional Monte-Carlo (MC) method, with a coefficient of determination (R²) value of 0.99 for the mean flow, during the historical period; while the proposed PCE-ANN method reproduces MC output with a R² value of 0.84 for mean flow during the forecasting period. It is also indicated that PCE and PCE-ANN are as reliable as but much more efficient than MC. PCE takes about 1% of the computational time required by MC; PCE-ANN only takes a few minutes to produce probabilistic forecasting, while MC needs rerunning the model all over again for dozens or hundreds, even thousands, of times. Notably, the development of the PCE-ANN framework is the first attempt to explore PCE’s probabilistic forecasting capability using machine learning. PCE-ANN is a promising uncertainty assessment and probabilistic forecasting technique, as it is more efficient in terms of computation time, and it does not cause loss of essential uncertainty information.
Flood is a major natural hazard with extremely large impact on social-ecological systems. Therefore, developing reliable and efficient tools to identify areas vulnerable to potential flooding is vital for water managers, engineers and decision makers. Moreover, being able to accurately classify the level of hazard is a step forward towards more efficient flood hazard mapping. This study presents a multi-criteria index approach to classify potential flood hazards at the river basin scale. The presented methodology was implemented in the Mashhad Plain basin in North-east Iran, where flood has been a major issue in the last few decades. In the present study, seven factors, selected based on their greater influence towards flooding, were identified and extracted from the basic thematic layers to be used to generate a five-class Flood Hazard Index (FHI) map. The Soil and Water Assessment Tool (SWAT) was used to develop a runoff coefficient map, which was found to be the most influential factor. A sensitivity analysis was performed and the results incorporated to generate a modified Flood Hazard Index (mFHI) map. The accuracy of the proposed method was evaluated against the well-documented flood records in the last 42 years at the study area. The results showed that, for both FHI and mFHI maps, more than 97% of historical flood events have occurred in moderate to very high flood hazard areas. This demonstrates that incorporating hydrological model (such as SWAT) and multi-criteria analysis introduces a robust methodology to generate comprehensive potential flood hazard maps. Moreover, the proposed modified methodology can be used to identify high potential flood hazard zones and work towards more efficient flood management and mitigation strategies.
Various conservation practices have been used to control soil erosion in cropland which makes assessing the effects of soil conservation more important. Northeastern China is an important crop-producing regions, with severe soil erosion. In the study, 110 runoff plot-year observations from seven experimental stations distributed over the region were used to assess the effects of local conservation practices on soil erosion, using a simplified USLE. Under the assumption that there were no significant differences in rainfall erosivity and soil erodibility in the study area under the “no practice scenario”, we described a simplified linear model based slope steepness and soil loss rate. The Jixing watershed was selected as a case study to test the model. The results indicate that the derived model can be used in the NE region of China for conservation assessment. The P factor values, related to main soil conservation practices, were derived from the observed plot-year data. The average values of P factor were around 0.01 to 0.35, which were low enough to justify their conservation effects of implementation through the assessment. Comparing to the conventional slope ridging tillage, the conservation rates of terraces, strip ridging, orchards and agro-forestry in the Jixing Watershed were 0.57, 0.6, 1.07, and 0.62 mm/ha, respectively, obtained by the assessment method in the study case. The paper neglected the seasonal variation which should be grouped to the C factor for the annual data and discussed the limitations of the assessment model for application.
Scale represents an important concept in all scientific disciplines, but the scaling effect related to non-point source (NPS) pollution simulation and sensitivity parameters has not yet been reported. In this study, the sliding window, the Fourier amplitude sensitivity test and the nested watershed idea were used with the Soil and Water Assessment Tool (SWAT) model to explore the temporal and spatial scaling effects of parameter sensitivity in a typical watershed of the Three Gorges Reservoir Region, China. The results indicated that a great scaling effect could be observed at varying spatio-temporal scales, while the scaling effect would be transferred and amplified from the hydrological modelling to the NPS simulation. Soil properties such as SOL_K and SOL_BD were identified as key parameters under smaller spatial and temporal scales, while channel-related parameters in terms of ALPHA_BF, CH_K and CH_N showed greater sensitivity at larger scales. Specifically, some parameters, such as CH_N, USLE_K, USLE_P and ERORGP, were always identified as key sensitive parameters for the sediment and NPS-TP simulations, but some parameters, such as CH_K, showed sensitivity only above a specific spatial scale (778 km² in this study). These results could be used as a reference for studying the scaling effect of model parameter sensitivity and provide important information for model construction and the management of NPS pollution at different scales.
The methodology used to construct tree structured rules is the focus of this monograph. Unlike many other statistical procedures, which moved from pencil and paper to calculators, this text's use of trees was unthinkable before computers. Both the practical and theoretical sides have been developed in the authors' study of tree methods. Classification and Regression Trees reflects these two sides, covering the use of trees as a data analysis method, and in a more mathematical framework, proving some of their fundamental properties.
The soil and water assessment tool (SWAT) was calibrated for the Luvuvhu River catchment, South Africa in order to simulate runoff. The model was executed through QSWAT which is an interface between SWAT and QGIS. Data from four weather stations and four weir stations evenly distributed over the catchment were used. The model was run for a 33-year period of 1983–2015. Sensitivity analysis, calibration and validation were conducted using the sequential uncertainty fitting (SUFI-2) algorithm through its interface with SWAT calibration and uncertainty procedure (SWAT-CUP). The calibration process was conducted for the period 1986 to 2005 while the validation process was from 2006 to 2015. Six model efficiency measures were used, namely: coefficient of determination (R²), Nash–Sutcliffe efficiency (NSE) index, root mean square error (RMSE)-observations standard deviation ratio (RSR), percent bias (PBIAS), probability (P)-factor and correlation coefficient (R)-factor were used. Initial results indicated an over-estimation of low flows with regression slope of less than 0.7. Twelve model parameters were applied for sensitivity analysis with four (ALPHA_BF, CN2, GW_DELAY and SOL_K) found to be more distinguishable and sensitive to streamflow (p < 0.05). The SUFI-2 algorithm through the interface with the SWAT-CUP was capable of capturing the model's behaviour, with calibration results showing an R² of 0.63, NSE index of 0.66, RSR of 0.56 and a positive PBIAS of 16.3 while validation results revealed an R² of 0.52, NSE of 0.48, RSR of 0.72 and PBIAS of 19.90. The model produced P-factor of 0.67 and R-factor of 0.68 during calibration and during validation, 0.69 and 0.53 respectively. Although performance indicators yielded fair and acceptable results, the P-factor was still below the recommended model performance of 70%. Apart from the unacceptable P-factor values, the results obtained in this study demonstrate acceptable model performance during calibration while validation results were still inconclusive. It can be concluded that calibration of the SWAT model yielded acceptable results with exception to validation results. Having said this, the model can be a useful tool for general water resources assessment and not for analysing hydrological extremes in the Luvuvhu River catchment.
In this study, a Bayesian-based multilevel factorial analysis (BMFA) method is developed to assess parameter uncertainties and their effects on hydrological model responses. In BMFA, Differential Evolution Adaptive Metropolis (DREAM) algorithm is employed to approximate the posterior distributions of model parameters with Bayesian inference; factorial analysis (FA) technique is used for measuring the specific variations of hydrological responses in terms of posterior distributions to investigate the individual and interactive effects of parameters on model outputs. BMFA is then applied to a case study of the Jinghe River watershed in the Loess Plateau of China to display its validity and applicability. The uncertainties of four sensitive parameters, including soil conservation service runoff curve number to moisture condition II (CN2), soil hydraulic conductivity (SOL_K), plant available water capacity (SOL_AWC), and soil depth (SOL_Z), are investigated. Results reveal that (i) CN2 has positive effect on peak flow, implying that the concentrated rainfall during rainy season can cause infiltration-excess surface flow, which is an considerable contributor to peak flow in this watershed; (ii) SOL_K has positive effect on average flow, implying that the widely distributed cambisols can lead to medium percolation capacity; (iii) the interaction between SOL_AWC and SOL_Z has noticeable effect on the peak flow and their effects are dependent upon each other, which discloses that soil depth can significant influence the processes of plant uptake of soil water in this watershed. Based on the above findings, the significant parameters and the relationship among uncertain parameters can be specified, such that hydrological model's capability for simulating/predicting water resources of the Jinghe River watershed can be improved.
Global Sensitivity Analysis (GSA) is an essential technique to support the calibration of environmental models by identifying the influential parameters (screening) and ranking them. In this paper, the widely-used variance-based method (Sobol') and the recently proposed moment-independent PAWN method for GSA are applied to the Soil and Water Assessment Tool (SWAT), and compared in terms of ranking and screening results of 26 SWAT parameters. In order to set a threshold for parameter screening, we propose the use of a “dummy parameter”, which has no influence on the model output. The sensitivity index of the dummy parameter is calculated from sampled data, without changing the model equations. We find that Sobol’ and PAWN identify the same 12 influential parameters but rank them differently, and discuss how this result may be related to the limitations of the Sobol’ method when the output distribution is asymmetric.
In this study, a hybrid sequential data assimilation and probabilistic collocation (HSDAPC) approach is proposed for analyzing uncertainty propagation and parameter sensitivity of hydrologic models. In HSDAPC, the posterior probability distributions of model parameters are first estimated through a particle filter method based on streamflow discharge data. A probabilistic collocation method (PCM) is further employed to show uncertainty propagation from model parameters to model outputs. The temporal dynamics of parameter sensitivities are then generated based on the polynomial chaos expansion (PCE) generated by PCM, which can reveal the dominant model components for different catchment conditions. The maximal information coefficient (MIC) is finally employed to characterize the correlation/association between model parameter sensitivity and catchment precipitation, potential evapotranspiration and observed discharge. The proposed method is applied to the Xiangxi River located in the Three Gorges Reservoir area. The results show that: (i) the proposed HSDAPC approach can generate effective 2nd and 3rd PCE models which provide accuracy predictions; (ii) 2nd-order PCE, which can run nearly ten time faster than the hydrologic model, can capably represent the original hydrological model to show the uncertainty propagation in a hydrologic simulation; (iii) the slow (Rs) and quick flows (Rq) in Hymod show significant sensitivities during the simulation periods but the distribution factor (?) shows a least sensitivity to model performance; (iv) the model parameter sensitivities show significant correlation with the catchment hydro-meteorological conditions, especially during the rainy period with MIC values larger than 0.5. Overall, the results in this paper indicate that uncertainty propagation and temporal sensitivities of parameters can be effectively characterized through the proposed HSDAPC approach.
Process-based hydrological models have a long history dating back to the 1960s. Criticized by some as over-parameterized, overly complex, and difficult to use, a more nuanced view is that these tools are necessary in many situations and, in a certain class of problems, they are the most appropriate type of hydrological model. This is especially the case in situations where knowledge of flow paths or distributed state variables and/or preservation of physical constraints is important. Examples of this include: spatiotemporal variability of soil moisture, groundwater flow and runoff generation, sediment and contaminant transport, or when feedbacks among various Earth's system processes or understanding the impacts of climate non-stationarity are of primary concern. These are situations where process-based models excel and other models are unverifiable. This article presents this pragmatic view in the context of existing literature to justify the approach where applicable and necessary. We review how improvements in data availability, computational resources and algorithms have made detailed hydrological simulations a reality. Avenues for the future of process-based hydrological models are presented suggesting their use as virtual laboratories, for design purposes, and with a powerful treatment of uncertainty.
Hydrologic modeling of the rainfall runoff processes for predictions of future flow events requires a modeling system composed of three elements: (1) selection of an appropriate mathematical rainfall-runoffmodel; (2) a suitable calibration system, and (3) the required observations by the model. Over the past 40 years and with the advent of digital computers, hydrologic models of various levels of sophistication have been developed. Progress toward development of more advanced parameter estimation methods for model calibration has also been made. This chapter provides an overview of the recent developments in modeling and parameter estimation methods available for model calibration purposes. © 2013 by World Scientific Publishing Co. Pte. Ltd. All rights reserved.
