A two-stage sensitivity analysis for parameter identification and calibration of a physically-based distributed model in a river basin

Hydrological models demand large numbers 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 the sensitivity analysis simplified the calibration procedure of SHETRAN for the study area.     

Transferability of hydrological model parameter spaces in the estimation of runoff in ungauged catchments

Abstract

In this study, transferability options of the Hydrologiska Byråns Vattenbalansavdelning (HBV) hydrological model parameter (MP) spaces were investigated to estimate ungauged catchment runoff. Three approaches were applied in the study: MP space transfer from single, neighbouring and all potential donor catchments. The model performance was evaluated by a jackknife procedure, where one catchment at a time was treated as if ungauged, and behavioural MP sets from candidate donor catchments were used to estimate the “ungauged” runoff. The results showed that ungauged catchment runoff estimation could not be guaranteed by transferring MP sets from a single physiographically nearest donor catchment. Integrating MP sets typically from one to six donor catchments supplemented the lack of effective MP sets and improved the model performance at the ungauged catchments. In addition, the analysis results revealed that the model performance converged to an average performance when the MP sets of all potential donor catchments were integrated.     

Development of distributed time-variant gain model for nonlinear hydrological systems

Hydrological science is a branch of the earth sci-ences.To study the complexities of hydrological pro-cesses and the associated environmental problems,a systematic approach is desired.For instance,Dooge published Linear Theory of Hydrological System[1]in1973.Singh(1988)published Hydrological Systems[2],with its Chinese version[3]being translated by the Yel-low River Conservancy Commission in2000.Ge(1999)carried out systematic studies on the hydro-logical linear system theory,and produced …     

Global sensitivity analyses for a complex hydrological model applied in an Alpine watershed

The increasing complexity of hydrological models results in a large number of parameters to be estimated. In order to better understand how these complex models work, efficient screening methods are required in order to identify the most important parameters. This is of particular importance for models that are used within an operational real‐time forecasting chain such as HQsim. The objectives of this investigation are to (i) identify the most sensitive parameters of the complex HQsim model applied in the Alpine Lech catchment and (ii) compare model parameter sensitivity rankings attained from three global sensitivity analysis techniques. The techniques presented are the (i) regional sensitivity analysis, (ii) Morris analysis and (iii) state‐dependent parameter modelling. The results indicate that parameters affecting snow melt as well as processes in the unsaturated soil zone reveal high significance in the analysed catchment. The snow melt parameters show clear temporal patterns in the sensitivity whereas most of the parameters affecting processes in the unsaturated soil zone do not vary in importance across the year. Overall, the maximum degree day factor (meltfunc_max) has been identified to play a key role within the HQsim model. Although the parameter sensitivity rankings are equivalent between methods for a number of parameters, for several key parameters differing results were obtained. An uncertainty analysis demonstrates that a parameter ranking attained from only one method is subjected to large uncertainty. Copyright © 2012 John Wiley & Sons, Ltd.     

Development of a distributed hydrological model to facilitate watershed management

To facilitate precise and cost-effective watershed management, a simple yet spatially and temporally distributed hydrological model (DHM-WM) was developed. The DHM-WM is based on the Mishra-Singh version of the curve number method, with several modifications: The spatial distribution of soil moisture was considered in moisture updating; the travel time of surface runoff was calculated on a grid cell basis for routing; a simple tile flow module was included as an option. The DHM-WM was tested on a tile-drained agricultural watershed in Indiana, USA. The model with the tile flow module performed well in the study area, providing a balanced water budget and reasonable flow partitioning. The daily coefficient of determination and Nash-Sutcliffe coefficient were 0.58 and 0.56, for the calibration period, and 0.63 and 0.62 for the validation period. The DHM-WM also provides detailed information about the source areas of flow components, the travel time and pathways of surface runoff.

EDITOR A. Castellarin; ASSOCIATE EDITOR F.-J. Chang     

A distributed monthly hydrological model for integrating spatial variations of basin topography and rainfall

Hydrological models at a monthly time‐scale are important tools for hydrological analysis, such as in impact assessment of climate change and regional water resources planning. Traditionally, monthly models adopt a conceptual, lumped‐parameter approach and cannot account for spatial variations of basin characteristics and climatic inputs. A large requirement for data often severely limits the utility of physically based, distributed‐parameter models. Based on the variable‐source‐area concept, we considered basin topography and rainfall to be two major factors whose spatial variations play a dominant role in runoff generation and developed a monthly model that is able to account for their influences in the spatial and temporal dynamics of water balance. As a hybrid of the Xinanjiang model and TOPMODEL, the new model is constructed by innovatively making use of the highly acclaimed simulation techniques in the two existing models. A major contribution of this model development study is to adopt the technique of implicit representation of soil moisture characteristics in the Xinanjiang model and use the TOPMODEL concept to integrate terrain variations into runoff simulation. Specifically, the TOPMODEL topographic index ln(a/tanβ) is converted into an index of relative difficulty in runoff generation (IRDG) and then the cumulative frequency distribution of IRDG is used to substitute the parabolic curve, which represents the spatial variation of soil storage capacity in the Xinanjiang model. Digital elevation model data play a key role in the modelling procedures on a geographical information system platform, including basin segmentation, estimation of rainfall for each sub‐basin and computation of terrain characteristics. Other monthly data for model calibration and validation are rainfall, pan evaporation and runoff. The new model has only three parameters to be estimated, i.e. watershed‐average field capacity WM, pan coefficient η and runoff generation coefficient α. Sensitivity analysis demonstrates that runoff is least sensitive to WM and, therefore, it can be determined by a prior estimation based on the climate and soil properties of the study basin. The other two parameters can be determined using optimization methods. Model testing was carried out in a number of nested sub‐basins of two watersheds (Yuanjiang River and Dongjiang River) in the humid region in central and southern China. Simulation results show that the model is capable of describing spatial and temporal variations of water balance components, including soil moisture content, evapotranspiration and runoff, over the watershed. With a minimal requirement for input data and parameterization, this terrain‐based distributed model is a valuable contribution to the ever‐advancing technology of hydrological modelling. Copyright © 2006 John Wiley & Sons, Ltd.     

Observational and predictive uncertainties for multiple variables in a spatially distributed hydrological model

In this study, uncertainty in model input data (precipitation) and parameters is propagated through a physically based, spatially distributed hydrological model based on the MIKE SHE code. Precipitation uncertainty is accounted for using an ensemble of daily rainfall fields that incorporate four different sources of uncertainty, whereas parameter uncertainty is considered using Latin hypercube sampling. Model predictive uncertainty is assessed for multiple simulated hydrological variables (discharge, groundwater head, evapotranspiration, and soil moisture). Utilizing an extensive set of observational data, effective observational uncertainties for each hydrological variable are assessed. Considering not only model predictive uncertainty but also effective observational uncertainty leads to a notable increase in the number of instances, for which model simulation and observations are in good agreement (e.g., 47% vs. 91% for discharge and 0% vs. 98% for soil moisture). Effective observational uncertainty is in several cases larger than model predictive uncertainty. We conclude that the use of precipitation uncertainty with a realistic spatio‐temporal correlation structure, analyses of multiple variables with different spatial support, and the consideration of observational uncertainty are crucial for adequately evaluating the performance of physically based, spatially distributed hydrological models.     

Testing the applicability of a kinematic wave-based distributed hydrological model in two climatically contrasting catchments

Abstract

Steep mountainous areas account for 70% of all river catchments in Japan. To predict river discharge for the mountainous catchments, many studies have applied distributed hydrological models based on a kinematic wave approximation with surface and subsurface flow components (DHM-KWSS). These models reproduce observed river discharge of catchments in Japan well; however, the applicability of a DHM-KWSS to catchments with different geographical and climatic conditions has not been sufficiently examined. This research applied a DHM-KWSS to two river basins that have different climatic conditions from basins in Japan to examine the transferability of the DHM-KWSS model structure. Our results show that the DHM-KWSS model structure explained flow regimes for a wet river basin as well as a large flood event in an arid basin; however, it was unable to explain long-term flow regimes for the arid basin case study.     

Coupling of a simultaneous heat and water model with a distributed hydrological model and evaluation of the combined model in a cold region watershed

Snow and frozen soil prevail in cold regions worldwide, and the integration of these processes is crucial in hydrological models. In this study, a combined model was developed by fully coupling a simultaneous heat and water model with a geomorphologically based distributed hydrological model. The combined model simulates vertical and lateral water transfer as well as vertical heat fluxes and is capable of representing the effects of frozen soil and snowmelt on hydrological processes in cold regions. This model was evaluated by using in situ observations in the Binggou watershed, an experimental watershed for cold region hydrology of the Watershed Allied Telemetry Experimental Research Project. Results showed that the model was able to predict soil freezing and thawing, unfrozen soil water content, and snow depth reasonably well. The simulated hydrograph was in good agreement with the in situ observation. The Nash–Sutcliffe coefficient of daily discharge was 0.744 for the entire simulation period, 0.472 from April to June, and 0.711 from June to November. This model can improve our understanding of hydrological processes in cold regions and assess the impacts of global warming on hydrological cycles and water resources. Copyright © 2012 John Wiley & Sons, Ltd.     

Assessment on the function of reservoirs for flood control during typhoon seasons based on a distributed hydrological model

A grid‐based distributed hydrological model, PDTank model, is used to simulate hydrological processes in the upper Tone River catchment. The Tone River catchment often suffers from heavy rainfall events during the typhoon seasons. The reservoirs located in the catchment play an important role in flood regulation. Through the coupling of the PDTank model and a reservoir module that combines the storage function and operation function, the PDTank model is used for flood forecasting in this study. By comparing the hydrographs simulated using gauging and radar rainfall data, it is found that the spatial variability of rainfall is an important factor for flood simulation and the accuracy of the hydrographs simulated using radar rainfall data is slightly improved. The simulation of the typhoon flood event numbered No. 9 shows that the reservoirs in the catchment attenuate the peak flood discharge by 423·3 m³/s and validates the potential applicability of the distributed hydrological model on the assessment of function of reservoirs for flood control during typhoon seasons. Copyright © 2011 John Wiley & Sons, Ltd.     

A distributed parameter model of a frame pin‐supported wall structure

Many reinforced‐concrete frames collapse via a soft‐story mechanism during severe earthquakes. Such collapses are mainly attributed to concentrated deformation in a soft story. Deformation control is thus important in preventing collapse. The frame pin‐supported wall structure is a type of rocking structure that releases constraints at the bottom of the wall. Previous research has obtained good results for the deformation control of this type of structure. However, the interior forces and strength demands of the pin‐supported wall have not been systematically explored. In this paper, a distributed parameter model is developed to investigate the strength demand of the wall in a frame pin‐supported wall structure. In the model, the pin‐supported wall is simplified as a bending beam and the frame is simplified as a shear beam. The two beams are joined by distributed shear connectors, so that the shear force can be transferred at any location on the interface. The model can be solved using differential equations based on equilibrium and compatibility. The accuracy of the model is verified using SAP2000 (Computers and Structures Inc., Berkeley, CA, USA). Displacement distribution of the structure and distributions of the moment and shear force within the pin‐supported wall are obtained for two typical external force profiles. It is found that the pin‐supported wall can effectively reduce the drift concentration factor. Distributions of the displacement, moment, and shear force are closely correlated with the relative stiffness of the wall and frame. Finally, recommendations on the stiffness and strength of a pin‐supported wall are made. Copyright © 2015 John Wiley & Sons, Ltd.     

Investigating model performance and parameter sensitivity for runoff simulation across multiple events for a large green roof

Green roofs are a form of green infrastructure aimed at retaining or slowing the movement of precipitation as stormwater runoff to sewer systems. To determine total runoff versus retention from green roofs, researchers and practitioners alike employ hydrologic models that are calibrated to one or more observed events. However, questions still remain regarding how event size may impact parameter sensitivity, how best to constrain initial soil moisture (ISM), and whether limited observations (i.e., a single event) can be used within a calibration-validation framework. We explored these questions by applying the storm water management model to simulate a large green roof located in Syracuse, NY. We found that model performance was very high (e.g., Nash Sutcliffe efficiency index > 0.8 and Kling-Gupta efficiency index > 0.8) for many events. We initially compared model performance across two parameterizations of ISM. For some events, we found similar performance when ISM was varied versus set to zero; for others, varying ISM yielded higher performance as well as greater water balance closure. Within a calibration-validation framework, we found that calibrating to larger events tended to produce moderate to high performance for other non-calibration events. However, very small storms were notoriously difficult to simulate, regardless of calibration event size, as these events are likely fully retained on the roof. Using regional sensitivity analysis, we confirmed that only a subset of model parameters was sensitive across 16 events. Interestingly, many parameters were sensitive regardless of event size, though some parameters were more sensitive when simulating smaller events. This emphasizes that storm size likely influences parameter sensitivity. Overall, we show that while calibrating to a single event can achieve high performance, exploring simulations across multiple events can yield important insight regarding the hydrologic performance of green roofs that can be used to guide the gathering of in situ properties and observations for refining model frameworks.     

Higher mode effects in frame pin‐supported wall structure by using a distributed parameter model

Frame pin‐supported wall structure is a kind of rocking structure, which releases constraints at the bottom of the wall. The wall is affiliated to the frame and can rotate around the hinge. Previous studies have investigated seismic performance (such as deformation pattern and plastic hinge distribution) of frame pin‐supported wall structure. Strength demand of this system was investigated through static pushover analysis. However, dynamic characteristics, especially higher mode effects, remain to be quantified. As demonstrated in several researches, higher mode effects have non‐negligible effects on seismic response. For this purpose, a distributed model for analyzing higher mode effects in frame pin‐supported wall structure was proposed, where the pin‐supported wall and the frame were simplified as a bending beam and a shear beam, respectively. The model was solved by differential equations derived from equilibrium and compatibility. Displacement and inner force distribution of frame pin‐supported wall structure in higher modes were quantified according to the model. Influence of critical parameters, such as wall stiffness and structure period, was assessed on higher mode effects. It was demonstrated that response in higher modes cannot be neglected in the design of frame pin‐supported wall structure. Capacity design based on the fundamental mode is not conservative, especially in the wall. Furthermore, pin‐supported walls tend to force the frame to vibrate in the rocking mode and suppress higher mode effects in the frame. Copyright © 2016 John Wiley & Sons, Ltd.     

Real-time assimilation of streamflow observations into a hydrological routing model: effects of model structures and updating methods

This paper comparatively assesses the performance of five data assimilation techniques for three-parameter Muskingum routing with a spatially lumped or distributed model structure. The assimilation techniques used include direct insertion (DI), nudging scheme (NS), Kalman filter (KF), ensemble Kalman filter (EnKF) and asynchronous ensemble Kalman filter (AEnKF), which are applied to river reaches in Texas and Louisiana, USA. For both lumped and distributed routing, results from KF, EnKF and AEnKF are sensitive to the error specification. As expected, DI outperformed the other models in the case of lumped modelling, while in distributed routing, KF approaches, particularly AEnKF and EnKF, performed better than DI or nudging, reflecting the benefit of updating distributed states through error covariance modelling in KF approaches. The results of this work would be useful in setting up data assimilation systems that employ increasingly abundant real-time observations using distributed hydrological routing models.     

How to improve the representation of hydrological processes in SWAT for a lowland catchment – temporal analysis of parameter sensitivity and model performance

Model diagnostic analyses help to improve the understanding of hydrological processes and their representation in hydrological models. A detailed temporal analysis detects periods of poor model performance and model components with potential for model improvements, which cannot be found by analysing the whole discharge time series. In this study, we aim to improve the understanding of hydrological processes by investigating the temporal dynamics of parameter sensitivity and of model performance for the Soil and Water Assessment Tool model applied to the Treene lowland catchment in Northern Germany. The temporal analysis shows that the parameter sensitivity varies temporally with high sensitivity for three groundwater parameters (groundwater time delay, baseflow recession constant and aquifer fraction coefficient) and one evaporation parameter (soil evaporation compensation factor). Whereas the soil evaporation compensation factor dominates in baseflow and resaturation periods, groundwater time delay, baseflow recession constant and aquifer fraction coefficient are dominant in the peak and recession phases. The temporal analysis of model performance identifies three clusters with different model performances, which can be related to different phases of the hydrograph. The lowest performance, when comparing six performance measures, is detected for the baseflow cluster. A spatially distributed analysis for six hydrological stations within the Treene catchment shows similar results for all stations. The linkage of periods with poor model performance to the dominant model components in these phases and with the related hydrological processes shows that the groundwater module has the highest potential for improvement. This temporal diagnostic analysis enhances the understanding of the Soil and Water Assessment Tool model and of the dominant hydrological processes in the lowland catchment. Copyright © 2013 John Wiley & Sons, Ltd.     

Development and application of a spatially‐distributed Arctic hydrological and thermal process model (ARHYTHM)

A process‐based, spatially distributed hydrological model was developed to quantitatively simulate the energy and mass transfer processes and their interactions within arctic regions (arctic hydrological and thermal model, ARHYTHM). The model first determines the flow direction in each element, the channel drainage network and the drainage area based upon the digital elevation data. Then it simulates various physical processes: including snow ablation, subsurface flow, overland flow and channel flow routing, soil thawing and evapotranspiration. The kinematic wave method is used for conducting overland flow and channel flow routing. The subsurface flow is simulated using the Darcian approach. The energy balance scheme was the primary approach used in energy‐related process simulations (snowmelt and evapotranspiration), although there are options to model snowmelt by the degree‐day method and evapotranspiration by the Priestley–Taylor equation. This hydrological model simulates the dynamic interactions of each of these processes and can predict spatially distributed snowmelt, soil moisture and evapotranspiration over a watershed at each time step as well as discharge in any specified channel(s). The model was applied to Imnavait watershed (about 2·2 km²) and the Upper Kuparuk River basin (about 146 km²) in northern Alaska. Simulated results of spatially distributed soil moisture content, discharge at gauging stations, snowpack ablations curves and other results yield reasonable agreement, both spatially and temporally, with available data sets such as SAR imagery‐generated soil moisture data and field measurements of snowpack ablation, and discharge data at selected points. The initial timing of simulated discharge does not compare well with the measured data during snowmelt periods mainly because the effect of snow damming on runoff was not considered in the model. Results from the application of this model demonstrate that spatially distributed models have the potential for improving our understanding of hydrology for certain settings. Finally, a critical component that led to the performance of this modelling is the coupling of the mass and energy processes. Copyright © 2000 John Wiley & Sons, Ltd.     

Isotope hydrological study of mean transit time in the granitic Strengbach catchment (Vosges massif,France): application of the FlowPC model with modified input function

Measurements of ¹⁸O concentrations in precipitation, soil solution, spring and runoff are used to determine water transit time in the small granitic Strengbach catchment (0·8 km²; 883–1146 m above sea level) located in the Vosges Mountains of northeastern France. Water transit times were calculated by applying the exponential, exponential piston and dispersion models of the FlowPC program to isotopic input (rainfall) and output (spring and stream water) data sets during the period 1989–95. The input function of the model was modified compared with the former version of the model and estimated by a deterministic approach based on a simplified hydrological balance. The fit between observed and calculated output data showed marked improvements compared with results obtained using the initial version of the model. An exponential piston version of the model applied to spring water indicates a 38·5 month mean transit time, which suggests that the volume in the aquifer, expressed in water depth, is 2·4 m. A considerable thickness (>45 m) of fractured bedrock may be involved for such a volume of water to be stored in the aquifer. Copyright © 2005 John Wiley & Sons, Ltd.     

A regional model to assess the hydrological sensitivity of medium size catchments to climate variability

Catchment storage capacity is an important factor in the determination of catchment sensitivity to climate variability. Quantification of catchment sensitivity is in turn important in the regional assessment of the effects of possible climate change. In the present paper, an empirical regional model is proposed that quantifies catchment sensitivity as the ratio of present maximum reservoir storage to catchment storage capacity. Catchment storage capacity is defined theoretically using readily available catchment variables. Present maximum reservoir storage in a catchment, as determined from recession analysis, is expressed as a fraction of catchment storage capacity; the fraction defines catchment sensitivity and depends on storage capacity and annual net precipitation. Average annual conditions for present maximum reservoir storage and average annual net precipitation are used to test the developed model. Although the study used data from only 15 catchments in the Upper Loire region in France, the model proved statistically valid. Storage capacity calculated with the model compares favourably with the baseflow index and a storage index defined in previous research. Values of storage capacity are probable with respect to reported water resources in the area. With the model catchment sensitivity can easily be assessed. Flood or drought prone catchments can be identified as well as a catchment's sensitivity to a catchment-type transition (baseflow versus direct flow dominated catchments). © 1998 John Wiley & Sons, Ltd.     

Impacts of land use on the hydrological response of tropical Andean catchments

Changes in land use and land cover are major drivers of hydrological alteration in the tropical Andes. However, quantifying their impacts is fraught with difficulties because of the extreme diversity in meteorological boundary conditions, which contrasts strongly with the lack of knowledge about local hydrological processes. Although local studies have reduced data scarcity in certain regions, the complexity of the tropical Andes poses a big challenge to regional hydrological prediction. This study analyses data generated from a participatory monitoring network of 25 headwater catchments covering three of the major Andean biomes (páramo, jalca and puna) and links their hydrological responses to main types of human interventions (cultivation, afforestation and grazing). A paired catchment setup was implemented to evaluate the impacts of change using a ‘trading space‐for‐time’ approach. Catchments were selected based on regional representativeness and contrasting land use types. Precipitation and discharge have been monitored and analysed at high temporal resolution for a time period between 1 and 5 years. The observed catchment responses clearly reflect the extraordinarily wide spectrum of hydrological processes of the tropical Andes. They range from perennially humid páramos in Ecuador and northern Peru with extremely large specific discharge and baseflows, to highly seasonal, flashy catchments in the drier punas of southern Peru and Bolivia. The impacts of land use are similarly diverse and their magnitudes are a function of catchment properties, original and replacement vegetation and management type. Cultivation and afforestation consistently affect the entire range of discharges, particularly low flows. The impacts of grazing are more variable but have the largest effect on the catchment hydrological regulation. Overall, anthropogenic interventions result in increased streamflow variability and significant reductions in catchment regulation capacity and water yield, irrespective of the hydrological properties of the original biome. Copyright © 2016 The Authors. Hydrological Processes. Published by John Wiley & Sons Ltd.     

Assessment of spatial transferability of process‐based hydrological model parameters in two neighbouring catchments in the Himalayan Region

Estimating the hydrological regime of ungauged catchments in the Himalayan region is challenging due to a lack of sufficient monitoring stations. In this paper, the spatial transferability of the model parameters of the process‐oriented J2000 hydrological model was investigated in 2 glaciated subcatchments of the Koshi river basin in eastern Nepal. The catchments have a high degree of similarity with respect to their static landscape features. The model was first calibrated (1986–1991) and validated (1992–1997) in the Dudh Koshi subcatchment. The calibrated and validated model parameters were then transferred to the nearby Tamor catchment (2001–2009). Sensitivity and uncertainty analyses were carried out for both subcatchments to discover the sensitivity range of the parameters in the two catchments. The model represented the overall hydrograph well in both subcatchments, including baseflow, rising and falling limbs; however, the peak flows were underestimated. The efficiency results according to both Nash–Sutcliffe (ENS) and the coefficient of determination (r²) were above 0.84 in both catchments (1986–1997 in Dudh Koshi and 2001–2009 in Tamor). The ranking of the parameters in respect to their sensitivity matched well for both catchments while taking ENS and log Nash–Sutcliffe (LNS) efficiencies into account. However, there were some differences in sensitivity to ENS and LNS for moderately and less‐sensitive parameters, although the majority (13 out of 16 for ENS and 16 out of 16 for LNS) had a sensitivity response in a similar range. The generalized uncertainty likelihood estimation results suggest that the parameter uncertainty are most of the time within the range and the ensemble mean matches very good (ENS: 0.84) with observed discharge. The results indicate that transfer of the J2000 parameters to a neighbouring catchment in the Himalayan region with similar physiographic landscape characteristics is viable. This indicates the possibility of applying a calibrated process‐based J2000 model to other ungauged catchments in the Himalayan region, which could provide important insights into the hydrological system dynamics and provide much needed information to support water resources planning and management.