Journal of Hydrologic Engineering

Automatic calibration of complex models, which includes continuous hydrograph models, requires sophisticated calibration methods. Also, these model-independent-parameter estimators require complex objective functions to ensure that the final parameter values reflect the hydrologic flow components (surface runoff, interflow, and baseflow) that the models are designed to represent. A multicomponent objective function should include components to represent each of the important physical processes represented in the model. The goal of this investigation was to develop a weighted multicomponent objective function and a method that can be used to provide estimates of the weights. For best calibration accuracy, the weights of the objective function should reflect the flow proportions of the streamflow record. A properly weighted objective function will enhance the accuracy of a watershed model such as the Hydrological Simulation Program–FORTRAN (HSPF). Methods of weighting were first made using data from eight hypothetical watersheds that had varying proportions of flow components (surface runoff, interflow, and baseflow). Then, the methods were tested with data from two watersheds from Delaware and Maryland. The components of the objective function include baseflow separation and autoregression equations to represent hydrologic processes. The results indicated that the calibration accuracy greatly improved when the objective function weights were set based on flow proportions. Also, a conceptual baseflow separation was superior to hydrograph separation based on a Laplace transform filter. Two weighted multicomponent objective functions were compared and shown to provide different levels of accuracy, which suggests that the structure of the objective function is important.

A new class of time series models, referred to in this paper as the “periodic transfer function-noise (PTFN) model,” has been developed through an extension of conventional nonperiodic (or constant parameter) transfer function-noise (TFN) models. The proposed PTFN model is very flexible, as its form or order and parameter values of both the dynamic and noise components may vary depending on the season of the year. It is shown that Box et al.’s modeling techniques for TFN models can be applied to PTFN models as well. The model has been applied for monthly forecasting of the Ganges River flow using monthly rainfall data of northern India as the predictor. The results are encouraging and suggest that the PTFN class of models has the potential to be useful in capturing the seasonally varying dynamic relationship between a dependent time series and one or more independent time series where each series is interyear stationary but within-year nonstationary.

Hydrologic (i.e., flow volumes) and sediment loading calibration and validation were completed for a New York City water supply watershed, the Cannonsville Reservoir basin, located in upstate New York. An unusually large amount of data exist for this watershed, with more than 20,000 data points consisting of climate, flow, and sediment measurements for calibration and validation. Hydrology results using the Soil and Water Assessment Tool (SWAT) displayed a Nash–Sutcliffe measure (NS) of 0.74 at the main monitoring station that drains 80% of the $\mathrm{1,200}{\mathrm{km}}^2$ basin. Sediment results at this station came within 1% of average loading estimated from data for calibration (with one large storm and snowmelt event removed). Two of the three smaller drainage areas displayed calibrated hydrology NS values above 0.7. Adjustments in parameter values for calibration reflected northeastern conditions related to snowmelt events and fragipan soils. Validation of the hydrology and sediment results was based on an independent data set with no changes in parameter values. These validation results for hydrology showed an $\mathrm{NS}=0.76$ and an 8% difference in average flow volumes. Certain limitations of SWAT are discussed, including the model’s approach to snowmelt, sediment erosion, and sediment transport.

The state of Texas recently implemented a water-availability modeling (WAM) system to support planning and regulatory activities. River basin hydrology is represented in the WAM system by sequences of historical monthly naturalized streamflows and net reservoir evaporation rates. This paper describes a case study investigation of the potential effects of climate-change on assessments of water-supply capabilities and focuses on whether and how climate change considerations should be incorporated in the WAM system. A modeling approach was adopted to explore the impacts of climate change on hydrologic and institutional water availability for the numerous water users who depend on supplies provided by the $\mathrm{118,000}{\mathrm{km}}^2$ Brazos River Basin. Analyses of historical naturalized streamflows indicate hidden but significant multiple-year cycles but no long-term trends during the twentieth century. A climate model and watershed hydrology model are used to adjust the WAM system hydrology to reflect anomalous climate during 2040–2060. The future climate scenario generally results in decreased mean streamflows and greater variability. However, the effects on water availability vary significantly in different regions of the river basin and among water users.

In this study, fuzzy sets are utilized to interpret uncertainty in aquifer parameters in the solution of groundwater optimization problems. For this purpose, an optimization model is developed for the design of groundwater remediation systems with multiple uncertain parameters and multiple candidate pumping wells. The uncertain parameters selected include hydraulic conductivity and longitudinal and transverse dispersion coefficients. A genetic algorithm embedded with fuzzy vertex algebra is used to solve the optimization model. This approach is an extension of an earlier method proposed by the writers that is more suitable for large scale applications with multiple uncertain parameters. The numerical experiments are conducted to demonstrate the effectiveness of the procedures discussed in this study. The approach presented in this study provides guidance for the interpretation of uncertain parameters in groundwater optimization problems using fuzzy sets. The computational results show that the combined use of genetic algorithms and fuzzy vertex analysis yields an efficient method for the solution of optimization problems with multiple uncertain aquifer parameters.

Synthetic hydrologic time series can be used to quantify the uncertainty of a water resources system. Conventional parametric models, such as autoregressive moving average or Markovian models, assume that the variable under consideration is Gaussian. This assumption, however, is a shortcoming of parametric models and motivates the development of nonparametric approaches. Nonparametric models based on a kernel function have an innate low-order structure and are restricted to highly persistent variables. This study presented a seminonparametric (SNP) model that takes advantage of both parametric and nonparametric models to generate monthly precipitation and temperature in the Conchos River Basin in Mexico. By adopting a consistent and robust scheme from the Markovian model and a nonparametric mechanism to generate a distribution-free random component, the SNP model reliably reproduced sample properties such as mean, variance, correlation, and multimodality in the probability density function.

A modified temperature index snowmelt model (SDSM-MTI), based on both near-surface soil temperature $\left(T_g\right)$ and air temperature $\left(T_a\right)$ , was proposed and successfully tested at the Paddle River Basin (PRB) of Alberta. By using a weighted average of $T_g$ and $T_a$ and introducing a melt-rate adjustment factor $M_{rf}$ , SDSM-MTI could simulate more accurate snowmelt runoff, snow water equivalent, and snow depth at PRB than the standard temperature index approach operated either under fixed or seasonally variable melt factor $\left(M_r\right)$ independently calibrated with $T_a$ only. This is partly because at PRB the primary energy fluxes responsible for snowmelt correlate more closely with $T_g$ than $T_a$ , especially at a daily time step, and partly because when $T_g<0°\mathrm{C}$ , $M_{rf}$ of SDSM-MTI is much less than one that reduces $M_r$ to a very small value, and hence we can more effectively control the timing of major snowmelt for PRB, which usually happens only when $T_g⩾0°\mathrm{C}$ .

The source/sink terms in groundwater flow are singularities that require special treatment in the numerical solution of a mathematical model. Except for the case of a point source, these singularities are not consistently defined in the governing partial differential equation or treated rigorously in the numerical solution of the governing equations. In this paper, we propose a consistent notation for handling the source/sink terms within the governing equation using the familiar Dirac- $\delta$ function notation for three-dimensional applications. We also present the domain integrals of these terms, similar to the ones that occur frequently in finite element solution of the mathematical model.