scispace - formally typeset
Search or ask a question

Showing papers in "Water Resources Research in 1996"


Journal ArticleDOI
TL;DR: This paper addresses the problem of evaluating the predictive uncertainty of TOPMODEL using the Bayesian Generalised Likelihood Uncertainty Estimation (GLUE) methodology in an application to the small Ringelbach research catchment in the Vosges, France.
Abstract: This paper addresses the problem of evaluating the predictive uncertainty of TOPMODEL using the Bayesian Generalised Likelihood Uncertainty Estimation (GLUE) methodology in an application to the small Ringelbach research catchment in the Vosges, France. The wide range of parameter sets giving acceptable simulations is demonstrated, and uncertainty bands are presented based on different likelihood measures. It is shown how the distributions of predicted discharges are non-Gaussian and vary in shape through time and with discharge. Updating of the likelihood weights using Bayes equation is demonstrated after each year of record and it is shown how the additional data can be evaluated in terms of the way they constrain the uncertainty bands.

807 citations


Journal ArticleDOI
TL;DR: The motivation for this work comes from a desire to preserve the dependence structure of the time series while bootstrapping (resampling it with replacement), and the method is data driven and is preferred where the investigator is uncomfortable with prior assumptions.
Abstract: A nonparametric method for resampling scalar or vector-valued time series is introduced. Multivariate nearest neighbor probability density estimation provides the basis for the resampling scheme developed. The motivation for this work comes from a desire to preserve the dependence structure of the time series while bootstrapping (resampling it with replacement). The method is data driven and is preferred where the investigator is uncomfortable with prior assumptions as to the form (e.g., linear or nonlinear) of dependence and the form of the probability density function (e.g., Gaussian). Such prior assumptions are often made in an ad hoc manner for analyzing hydrologic data. Connections of the nearest neighbor bootstrap to Markov processes as well as its utility in a general Monte Carlo setting are discussed. Applications to resampling monthly streamflow and some synthetic data are presented. The method is shown to be effective with time series generated by linear and nonlinear autoregressive models. The utility of the method for resampling monthly streamflow sequences with asymmetric and bimodal marginal probability densities is also demonstrated.

713 citations


Journal ArticleDOI
TL;DR: In this paper, the water retention model developed by Kosugi was modified to be compatible with Mualem's model in order to derive an analytical expression for the relative hydraulic conductivity Kr.
Abstract: The soil water retention model developed by Kosugi was modified to be compatible with Mualem's model in order to derive an analytical expression for the relative hydraulic conductivity Kr. The modified water retention model is to be derived by applying a lognormal distribution law to the soil pore radius distribution function. Parameters of this retention model have physical significance on the water content (θ)– capillary pressure (ψ) curve and are related directly to the statistics of the pore radius distribution. The accuracy of the resulting combined water-retention-hydraulic-conductivity model is verified for observed data sets for six soils. Results showed that the proposed model produces acceptable matches with observed water retention curves and adequate predictions of hydraulic conductivities in five out of six cases. The θ − ψ and Kr− ψ (or Kr − θ) curves generated by this model are generally similar to those generated by van Genuchten's model.

602 citations


Journal ArticleDOI
TL;DR: A theoretical framework for model validation, based on the methodology originally proposed by Klemes, is presented and it is concluded that all models performed equally well when at least 1 year's data were available for calibration, while the distributed models performed marginally better for cases where no calibration was allowed.
Abstract: A theoretical framework for model validation, based on the methodology originally proposed by Klemes [1985, 1986], is presented. It includes a hierarchial validation testing scheme for model application to runoff prediction in gauged and ungauged catchments subject to stationary and nonstationary climate conditions. A case study on validation and intercomparison of three different models on three catchments in Zimbabwe is described. The three models represent a lumped conceptual modeling system (NAM), a distributed physically based system (MIKE SHE), and an intermediate approach (WATBAL). It is concluded that all models performed equally well when at least 1 year's data were available for calibration, while the distributed models performed marginally better for cases where no calibration was allowed.

584 citations


Journal ArticleDOI
TL;DR: In this paper, the authors present a functional formulation of the groundwater flow inverse problem that is sufficiently general to accommodate most commonly used inverse algorithms, including the Gaussian maximum a posteriori (GAP) algorithm.
Abstract: This paper presents a functional formulation of the groundwaterflow inverse problem that is sufficiently general to accommodate most commonly used inverse algorithms. Unknown hydrogeological properties are assumed to be spatial functions that can be represented in terms of a (possibly infinite) basis function expansion with random coefficients. The unknown parameter function is related to the measurements used for estimation by a ''forward operator'' which describes the measurement process. In the particular case considered here, the parameter of interest is the large-scale log hydraulic conductivity, the measurements are point values of log conductivity and piezometric head, and the forward operator is derived from an upscaled groundwaterflow equation. The inverse algorithm seeks the ''most probable'' or maximum a posteriori estimate of the unknown parameter function. When the measurement errors and parameter function are Gaussian and independent, the maximum a posteriori estimate may be obtained by minimizing a least squares performance index which can be partitioned into goodness-of- fit and prior terms. When the parameter is a stationary random function the prior portion of the performance index is equivalent to a regularization term which imposes a smoothness constraint on the estimate. This constraint tends to make the problem well- posed by limiting the range of admissible solutions. The Gaussian maximum a posteriori problem may be solved with variational methods, using functional generalizations of Gauss-Newton or gradient-based search techniques. Several popular groundwater inverse algorithms are either special cases of, or variants on, the functional maximum a posteriori algorithm. These algorithms differ primarily with respect to the way they describe spatial variability and the type of search technique they use (linear versus nonlinear). The accuracy of estimates produced by both linear and nonlinear inverse algorithms may be measured in terms of a Bayesian extension of the Cramer-Rao lower bound on the estimation error covariance. This bound suggests how parameter identifiability can be improved by modifying the problem structure and adding new measurements.

564 citations


Journal ArticleDOI
TL;DR: In this article, a case study is presented in which ANN methods are used to forecast salinity in the River Murray at Murray Bridge (South Australia) 14 days in advance, and the results obtained were most promising.
Abstract: This paper presents the use of artificial neural networks (ANNs) as a viable means of forecasting water quality parameters. A review of ANNs is given, and a case study is presented in which ANN methods are used to forecast salinity in the River Murray at Murray Bridge (South Australia) 14 days in advance. It is estimated that high salinity levels in the Murray cause $US 22 million damage per year to water users in Adelaide. Previous studies have shown that the average salinity of the water supplied to Adelaide could be reduced by about 10% if pumping from the Murray were to be scheduled in an optimal manner. This requires forecasts of salinity several weeks in advance. The results obtained were most promising. The average absolute percentage errors of the independent 14-day forecasts for four different years of data varied from 5.3% to 7.0%. The average absolute percentage error obtained as part of a real-time forecasting simulation for 1991 was 6.5%.

547 citations


Journal ArticleDOI
TL;DR: A number of methods have been developed to interpolate between data values and use geologic, hydrogeologic, and geophysical information to create images of aquifer properties.
Abstract: Numerical models that solve governing equations for subsurface fluid flow and transport are commonly applied to analyze quantitatively the effects of heterogeneity. These models require maps of spatially variable hydraulic properties. Because complete three-dimensional information about hydraulic properties is never obtainable, numerous methods have been developed to interpolate between data values and use geologic, hydrogeologic, and geophysical information to create images of aquifer properties. Image creation approaches fall into three general categories: structure-imitating, process-imitating, and descriptive. Structure-imitating methods rely on one or more of the following to constrain the geometry of spatial patterns in geologic media: correlated random fields, probabilistic rules, and deterministic constraints developed from facies relations. Structure-imitating methods include spatial statistical algorithms and geologically based sedimentation pattern-matching approaches. Process-imitating models include aquifer model calibration methods and geologic process models. Aquifer model calibration methods use governing equations for subsurface fluid flow and transport to relate hydraulic properties to heads and solute information through history and steady state data matching. Geologic process models combine fundamental laws of conservation of mass and momentum with sediment transport equations to simulate spatial patterns in grain size distributions. At the sedimentary basin scale, multiprocess models include thermomechanical mechanisms of basin subsidence. Descriptive methods couple geologic observations with facies relations to divide an aquifer into zones of characteristic hydraulic properties. All approaches are capable of reproducing heterogeneity over a range of scales and considering some types of geologic information. Some approaches are strictly spatial while some are linked to the time evolution of sedimentation. Some approaches can be conditioned on measurements. Recent advances aimed at infusing geologic information into images of the subsurface include extracting more information from sedimentological facies models, incorporating qualitative geologic information into random field generators and simulating depositional processes. Classes of research missing from the literature include multiprocess models that incorporate diagenesis and three-dimensional surface water flow, hybrid methods that combine features of existing approaches, and approaches that can make use of all available geologic, geophysical, and hydrologic data.

521 citations


Journal ArticleDOI
TL;DR: The improved genetic algorithm (GA) formulation for pipe network optimization has been developed and found a solution for the New York tunriels problem which is the lowest-cost feasible discrete size solution yet presented in the literature.
Abstract: An improved genetic algorithm (GA) formulation for pipe network optimization has been developed. The new GA uses variable power scaling of the fitness function. The exponent introduced into the fitness function is increased in magnitude as the GA computer run proceeds. In addition to the more commonly used bitwise mutation operator, an adjacency or creeping mutation operator is introduced. Finally, Gray codes rather than binary codes are used to represent the set of decision variables which make up i the pipe network design. Results are presented comparing the performance of the traditional or simple GA formulation and the improved GA formulation for the New York City tunnels problem. The case study results indicate the improved GA performs significantly better than the simple GA. In addition, the improved GA performs better than previously used traditional optimization methods such as linear, dynamic, and nonlinear programming methods and an enumerative search method. The improved GA found a solution for the New York tunriels problem which is the lowest-cost feasible discrete size solution yet presented in the literature.

507 citations


Journal ArticleDOI
TL;DR: This paper quantified long-term changes in streamflows associated with clear-cutting and road construction and examined alternative hydrologic mechanisms to explain stream hydrograph changes in the Cascades Range, western Oregon.
Abstract: This study quantified long-term changes in streamflows associated with clear- cutting and road construction and examined alternative hydrologic mechanisms to explain stream hydrograph changes in the Cascades Range, western Oregon. We examined differences in paired peak discharges for 150 to 375 storm events for five basin pairs, using 34-year records from two pairs of 60-to-101-ha experimental basins in the H. J. Andrews Experimental Forest, and 50-to-55-year records from three pairs of adjacent basins ranging from 60 to 600 km2. Forest harvesting has increased peak discharges by as much as 50% in small basins and 100% in large basins over the past 50 years. These increases are attributable to changes both in flow routing due to roads and in water balance due to treatment effects and vegetation succession.

489 citations


Journal ArticleDOI
TL;DR: In this article, the authors evaluated the reliability of the stream tracer approach to characterize hyporheic exchange in St. Kevin Gulch, a Rocky Mountain stream in Colorado contaminated by acid mine drainage.
Abstract: Stream water was locally recharged into shallow groundwater flow paths that returned to the stream (hyporheic exchange) in St. Kevin Gulch, a Rocky Mountain stream in Colorado contaminated by acid mine drainage. Two approaches were used to characterize hyporheic exchange: sub-reach-scale measurement of hydraulic heads and hydraulic conductivity to compute streambed fluxes (hydrometric approach) and reachscale modeling of in-stream solute tracer injections to determine characteristic length and timescales of exchange with storage zones (stream tracer approach). Subsurface data were the standard of comparison used to evaluate the reliability of the stream tracer approach to characterize hyporheic exchange. The reach-averaged hyporheic exchange flux (1.5 mL s−1 m−1), determined by hydrometric methods, was largest when stream base flow was low (10 L s−1); hyporheic exchange persisted when base flow was 10-fold higher, decreasing by approximately 30%. Reliability of the stream tracer approach to detect hyporheic exchange was assessed using first-order uncertainty analysis that considered model parameter sensitivity. The stream tracer approach did not reliably characterize hyporheic exchange at high base flow: the model was apparently more sensitive to exchange with surface water storage zones than with the hyporheic zone. At low base flow the stream tracer approach reliably characterized exchange between the stream and gravel streambed (timescale of hours) but was relatively insensitive to slower exchange with deeper alluvium (timescale of tens of hours) that was detected by subsurface measurements. The stream tracer approach was therefore not equally sensitive to all timescales of hyporheic exchange. We conclude that while the stream tracer approach is an efficient means to characterize surface-subsurface exchange, future studies will need to more routinely consider decreasing sensitivities of tracer methods at higher base flow and a potential bias toward characterizing only a fast component of hyporheic exchange. Stream tracer models with multiple rate constants to consider both fast exchange with streambed gravel and slower exchange with deeper alluvium appear to be warranted.

428 citations


Journal ArticleDOI
TL;DR: In this paper, a network model was developed to test the functional relationship between capillary pressure, saturation, and interfacial area, and the results indicated that a smooth, though complex, functional relationship exists at the continuum scale.
Abstract: The constitutive relationships required for the parameterization of multiphase flow and transport problems are of critical importance to hydrologic modeling. Recently, a hypothesis has been developed that predicts a functional relationship between capillary pressure, saturation, and interfacial area. A network model was developed to test this hypothesis. Microscale physical processes were simulated and volume averaging was used to derive the macroscopic measures of saturation andfluid-fluid interfacial area per volume of porous media. Results indicate that a smooth, though complex, functional relationship exists at the continuum scale. These results have direct relevance to constitutive theory and the modeling of nonaqueous phase liquid dissolution processes.

Journal ArticleDOI
TL;DR: In this article, the authors examined biases in WSR-88D hourly precipitation accumulation estimates from analyses of more than 1 year of weather surveillance radar data and rain gage data from the southern plains.
Abstract: Systematic biases in WSR-88D (Weather Surveillance Radar–1988 Doppler) hourly precipitation accumulation estimates are characterized from analyses of more than 1 year of WSR-88D data and rain gage data from the southern plains. Biases are examined in three contexts: (1) biases that arise from the range-dependent sampling of the WSR-88D, (2) systematic differences in radar rainfall estimates from two radars observing the same area, and (3) systematic differences between radar and rain gage estimates of rainfall. Range-dependent biases affect hourly rainfall accumulations products over much of the area covered by the WSR-88D. Significant underestimation of rainfall occurs within 40 km range of the radar due to bias in reflectivity observations at the higher elevation angles used for rainfall estimation close to the radar. Bright band and anomalous propagation (AP) lead to systematic overestimation of rainfall at intermediate range. Beyond 150 km in spring-summer and beyond 100 km in winter-fall, underestimation of precipitation is pronounced due to incomplete beam filling and overshooting of precipitation. Radar-radar intercomparison studies suggest that radar calibration is a significant problem at some sites. Anomalous propagation during clear-air conditions, a major problem with previous National Weather Service network radars, has been largely eliminated by the WSR-88D processing. AP remains a problem for cases in which AP returns are embedded in rain. Radar–rain gage intercomparison analyses indicate systematic underestimation by the WSR-88D relative to rain gages for paired gage-radar rainfall estimates. Analyses of spatial coverage of heavy rainfall, however, illustrate fundamental advantages of radar over rain gage networks for rainfall estimation.

Journal Article
TL;DR: In this paper, a parametric water balance model based on statistical averaging of the main hydrological processes was developed to improve the representation of runoff relative to the simple bucket without introducing the full complexity of the Sacramento model.
Abstract: A parametric water balance model was developed based on statistical averaging of the main hydrological processes. The model has a two-layer structure with both a physical and statistical basis for the model parameters. It was developed to fill a need for models with a small number of parameters and of intermediate complexity between a one-parameter simple bucket and more complex hydrologically oriented models with many parameters such as the Sacramento model. The focus was to improve the representation of runoff relative to the simple bucket without introducing the full complexity of the Sacramento model. The model was designed to operate over a range of time steps to facilitate coupling to an atmospheric model. The model can be used for catchment scale simulations in hydrological applications and for simple representation of runoff in coupled atmospheric/hydrological models. An important role for the simple water balance (SWB) model is to assist in understanding how much complexity in representing land surface processes is needed and can be supported with available data to estimate model parameters. The model is tested using rainfall, runoff, and surface meteorological data for three catchments from different climate regimes. Model performance is compared to performance of a simple bucket model, the Sacramento model, and the Oregon State University land surface model. Finally, a series of tests were conducted to evaluate the sensitivity of SWB performance when it is operated at time steps different from the time step for which it was calibrated.

Journal ArticleDOI
TL;DR: In this paper, a new method is proposed to simulate groundwater age directly, by use of an advection-dispersion transport equation with a distributed zero-order source of unit strength, corresponding to the rate of aging.
Abstract: A new method is proposed to simulate groundwater age directly, by use of an advection-dispersion transport equation with a distributed zero-order source of unit (1) strength, corresponding to the rate of aging. The dependent variable in the governing equation is the mean age, a mass-weighted average age. The governing equation is derived from residence-time-distribution concepts for the case of steady flow. For the more general case of transient flow, a transient governing equation for age is derived from mass-conservation principles applied to conceptual "age mass." The age mass is the product of the water mass and its age, and age mass is assumed to be conserved during mixing. Boundary conditions include zero age mass flux across all noflow and inflow boundaries and no age mass dispersive flux across outflow boundaries. For transient-flow conditions, the initial distribution of age must be known. The solution of the governing transport equation yields the spatial distribution of the mean groundwater age and includes diffusion, dispersion, mixing, and exchange processes that typically are considered only through tracer-specific solute transport simulation. Traditional methods have relied on advective transport to predict point values of groundwater travel time and age. The proposed method retains the simplicity and tracer-independence of advection-only models, but incorporates the effects of dispersion and mixing on volume-averaged age. Example simulations of age in two idealized regional aquifer systems, one homogeneous and the other layered, demonstrate the agreement between the proposed method and traditional particle-tracking approaches and illustrate use of the proposed method to determine the effects of diffusion, dispersion, and mixing on groundwater age.

Journal ArticleDOI
TL;DR: In this article, Hornberger et al. used the Regional Hydro-Ecological Simulation System to simulate water, carbon, and N dynamics and found that significant spatial and temporal variability in the release of nitrate-nitrogen from catchments in a sugar maple forest in central Ontario was observed.
Abstract: During the past decade, significant spatial and temporal variability in the release of nitrate-nitrogen (N) from catchments in a sugar maple forest in central Ontario was observed. To explain this variability, we tested the flushing hypothesis [Hornberger et al., 1994], where, when the soil saturation deficit is high, N accumulates in the upper layers of the soil and, as the soil saturation deficit decreases, the formation of a saturated subsurface layer flushes N from the upper layers of the soil into the stream. We used the Regional Hydro-Ecological Simulation System to simulate water, carbon, and N dynamics. A N flushing index was modeled as S/S30, the ratio of the current day saturation deficit to the previous 30-day average saturation deficit. A N source index was modeled as the ratio of N supply/demand. The relationship between the simulated N indices and the observed release of N indicated two mechanisms for the release of N from catchments: (1) a N flushing mechanism, where the N-enriched upper layer of the soil is flushed, after a period of low demand for N by the forest (e.g., during spring snowmelt and autumn stormflow, the water table rising into previously unsaturated parts of a N-enriched soil profile) or after a period of high demand for N by the forest (e.g., during summer droughts, the water table rising into previously saturated parts of a N-impoverished soil profile following a period of enhanced rates of nitrification); and (2) a N draining mechanism, where spring snowmelt recharge of the groundwater translocates N from the upper layer of the soil into deeper hydrological flow pathways that are released slowly over the year.

Journal ArticleDOI
TL;DR: In this article, two alternative forms of topographic index and soil-topographic index are developed based on parabolic and linear transmissivity functions, leading to more frequently observed second-order hyperbolic and exponential recession curves, respectively.
Abstract: Preliminary studies of the application of TOPMODEL to the 36-ha Ringelbach catchment suggested that the original form of exponential transmissivity function leading to the In (a/tan β) topographic index and first-order hyperbolic base flow recession curve is not appropriate to this catchment. Two alternative forms of topographic index and soil-topographic index are developed based on parabolic and linear transmissivity functions, leading to the more frequently observed second-order hyperbolic and exponential recession curves, respectively. It is shown how these can be used in the same way as the original to relate catchment average water table depths to local water table depths so that patterns of saturation can be evaluated. Two companion [Ambroise et al., this issue; Freer et al., this issue] papers show how the new parabolic index is used in the prediction of Ringelbach discharges, and how the limitations of the model are reflected in the estimated predictive uncertainties using the Generalised Likelihood Uncertainty Estimation (GLUE) approach.

Journal ArticleDOI
TL;DR: In this article, the random walk method for simulating solute transport in porous media is typically based on the assumption that the velocity and velocity-dependent dispersion tensor vary smoothly in space.
Abstract: The random-walk method for simulating solute transport in porous media is typically based on the assumption that the velocity and velocity-dependent dispersion tensor vary smoothly in space. However, in cases where sharp interfaces separate materials with contrasting hydraulic properties, these quantities may be discontinuous. Normally, velocities are interpolated to arbitrary particle locations when finite difference or finite element methods are used to solve the flow equation. The use of interpolation schemes that preserve discontinuities in velocity at material contacts can result in a random-walk model that does not locally conserve mass unless a correction is applied at these contacts. Test simulations of random-walk particle tracking with and without special treatment of material contacts demonstrate the problem. Techniques for resolving the problem, including interpolation schemes and a reflection principle, are reviewed and tested. Results from simulations of transport in porous media with discontinuities in the dispersion tensor show which methods satisfy continuity. Simulations of transport in two-dimensional heterogeneous porous media demonstrate the potentially significant effect of using a nonconservative model to compute spatial moments and breakthrough of a solute plume.

Journal ArticleDOI
TL;DR: In this article, a programmable data logger signals a pumping sampler to collect suspended sediment concentration (SSC) specimens at specific turbidity thresholds to estimate sediment loads during runoff events.
Abstract: For estimating suspended sediment concentration (SSC) in rivers, turbidity is generally a much better predictor than water discharge. Although it is now possible to collect continuous turbidity data even at remote sites, sediment sampling and load estimation are still conventionally based on discharge. With frequent calibration the relation of turbidity to SSC could be used to estimate suspended loads more efficiently. In the proposed system a programmable data logger signals a pumping sampler to collect SSC specimens at specific turbidity thresholds. Sampling of dense field records of SSC and turbidity is simulated to investigate the feasibility and efficiency of turbidity-controlled sampling for estimating sediment loads during runoff events. Measurements of SSC and turbidity were collected at 10-min intervals from five storm events in a small mountainous watershed that exports predominantly fine sediment. In the simulations, samples containing a mean of 4 to 11 specimens, depending on storm magnitude, were selected from each storm's record, and event loads were estimated by predicting SSC from regressions on turbidity. Using simple linear regression, the five loads were estimated with root mean square errors between 1.9 and 7.7%, compared to errors of 8.8 to 23.2% for sediment rating curve estimates based on the same samples. An estimator for the variance of the load estimate is imprecise for small sample sizes and sensitive to violations in regression model assumptions. The sampling method has potential for estimating the load of any water quality constituent that has a better correlate, measurable in situ, than discharge.

Journal ArticleDOI
TL;DR: In this paper, a parameter estimation procedure which combines the Levenberg-Marquardt nonlinear parameter optimization method involving weighted least squares, with a quasi-three-dimensional numerical model which solves the variably saturatedflow equation is described.
Abstract: Tension disc infiltrometers are becoming increasingly popular devices for in situ measurement of the unsaturated hydraulic properties of soil Tension infiltration data are generally used to evaluate the parametersKs and ain Gardner's exponential model of the unsaturated hydraulic conductivity Either two measurements using different disc diameters or measurements with a single disc but using multiple pressure heads are then used In this paper we describe a parameter estimation procedure which combines the Levenberg-Marquardt nonlinear parameter optimization method involving weighted least squares, with a quasi-three-dimensional numerical model which solves the variably saturatedflow equation By numerical inversion of Richards' equation the unknown parameters in van Genuchten's model of the unsaturated soil-hydraulic properties are estimated from observed cumulative infiltration data during transient waterflow Additional measurements of the pressure head or water content, as well as a penalty function for constraining the unknown parameters to remain in some feasible region (Bayesian estimation), can be optionally included into the parameter estimation procedure The problem of optimal sampling design, that is, selecting the best points in space and time for making measurements, is addressed by studying the sensitivity of the objective function to changes in the optimized hydraulic parameters We calculate objective functions based on available cumulative infiltration, pressure head, and water content measurements and also on several combinations of these data The behavior of the objective function in three-dimensional parameter space is evaluated by means of a series of two-dimensional response surfaces The utility of the parameter estimation procedure is demonstrated using numerically generated data The sensitivity of the procedure to different initial estimates of the model parameters is also discussed

Journal ArticleDOI
TL;DR: Using 94 years of monthly Palmer Drought Severity Index (PDSI) data for 344 climate divisions, the authors investigates the hydroclimatic response in the United States to the extreme phases of the Southern Oscillation (El Nino and La Nina).
Abstract: Using 94 years of monthly Palmer Drought Severity Index (PDSI) data for 344 climate divisions, this study investigates the hydroclimatic response in the United States to the extreme phases of the Southern Oscillation (El Nino and La Nina). Several regions of coherent response to El Nino-Southern Oscillation (ENSO) are identified. The strongest relationship between El Nino and extreme drought years is found in the Pacific Northwest. A strong relationship is also noticed in the southern United States, where dry conditions occur consistently during La Nina events. Next, the conditional response in PDSI is evaluated based on the extreme phases of the Southern Oscillation. The PDSI results were compared to similar analyses on 41 years of station temperature, precipitation, and streamflow data. A consistent response is seen in the other hydroclimatic variables, though the most filtered response is seen in PDSI data and streamflow data. The major contribution to the understanding of the ENSO–United States climate relationship is the evaluation of the general form of drought )nd comparison of these results to the fundamental hydrologic processes (precipitation, temperature, and streamflow).

Journal ArticleDOI
TL;DR: In this article, a fully automated method uses Landsat Thematic Mapper data to map snow cover in the Sierra Nevada and make quantitative estimates of the fractional snow-covered area within each pixel.
Abstract: A fully automated method uses Landsat Thematic Mapper data to map snow cover in the Sierra Nevada and make quantitative estimates of the fractional snow-covered area within each pixel. We model winter and spring reference scenes as linear mixtures of image end member spectra to produce the response variables for tree-based regression and classification models. Decision trees identify cloud cover and fractional snow-covered area. We test the algorithm on a different Thematic Mapper scene and verify with high- resolution, large-format, color aerial photography. The accuracy of the automated classification of Thematic Mapper data equals that obtainable from the aerial photographs but is faster, cheaper, and covers a vastly larger area. The mapping method is insensitive to the choice of lithologic or vegetation end members, the water equivalent of the snow pack, snow grain size, or local illumination angle.

Journal ArticleDOI
TL;DR: In this paper, the authors compared five methods (Thornthwaite, Hamon, Jensen-Haise, Turc, and Penman) for estimating potential evaporation for a reference surface (PEr) were compared to four methods (Priestley-Taylor, McNaughton-Black, Penman-Monteith, and Shuttleworth-Wallace), for estimating surface-dependent potential eviction (PEs) using three cover types at each of seven locations from Fairbanks, Alaska, to San Juan, Puerto Rico.
Abstract: Five methods (Thornthwaite, Hamon, Jensen-Haise, Turc, and Penman) for estimating potential evaporation for a reference surface (PEr) were compared to four methods (Priestley-Taylor, McNaughton-Black, Penman-Monteith, and Shuttleworth-Wallace) for estimating surface-dependent potential evaporation (PEs) using three cover types at each of seven locations from Fairbanks, Alaska, to San Juan, Puerto Rico. For annual PE the PEs methods generally agreed with the PEr methods, but for many locations, differences among methods were hundreds of millimeters per year. No methods were consistently low or high. Three of the PEs methods depend strongly on maximum leaf conductance, for which Korner [1994] provided satisfactory values by cover type. Potential interception ∥PEi∥ can only be estimated appropriately for all cover types by the Shuttleworth-Wallace method. Use of 5-day or monthly input data did not greatly degrade results, so use of monthly data to generate PE estimates appears warranted in global water balance models.

Journal ArticleDOI
TL;DR: In this article, the authors used neural networks to model the drying water retention curve (WRC) of 204 sandy soil samples from particle-size distribution (PSD), soil organic matter content (SOM), and bulk density (BD).
Abstract: We used neural networks (NNs) to model the drying water retention curve (WRC) of 204 sandy soil samples from particle-size distribution (PSD), soil organic matter content (SOM), and bulk density (BD). Neural networks can relate multiple model input data to multiple model output data without the need of an a priori model concept. In this way a high performance black-box model is created, which is very useful in a data exploration effort to assess the maximum obtainable prediction accuracy. We used a series of NN models with an increasing parametrization of input and output variables to get a better interpretability of model results. In the first two models we used the nine PSD fractions, BD, and SOM as input, while we predicted the nine points of the water retention curve. These NNs had 12 input and 9 output variables, predicting WRCs with an average root-mean-square residual (RMSR) water content of 0.020 cm3 cm−3. After a few intermediary models with increasing parametrization of PSD and WRC using (adapted) van Genuchten [1980] equations we arrived at a final NN model that used six input variables to predict three van Genuchten [1980] parameters resulting in a RMSR of 0.024 cm3 cm−3. We found saturated and residual water contents to be unrelated to the PSD, BD, or SOM, therefore the saturated water content was considered to be an independent input variable, while the residual water content was set to zero. Sensitivity analyses showed that the PSD had a major influence on the shape of the WRC, while BD and SOM were less important. On the basis of these sensitivity analyses we established more explicit equations that demonstrated similarity relations between PSD and WRC and incorporated effects of SOM and BD in an empirical way. Despite the fact that we considered a large number of linear and nonlinear variants these equations had a weaker performance (RMSR: 0.029 cm3 cm−3) than the NN models, proving the modeling power of that technique.

Journal ArticleDOI
TL;DR: In this article, a power law correlation is found between satellite-derived effective width We and discharge Q, where We is the water surface area within a braided reach divided by the reach length.
Abstract: Analysis of 41 ERS 1 synthetic aperture radar images and simultaneous ground measurements of discharge for three large braided rivers indicates that the area of active flow on braided river floodplains is primarily a function of discharge. A power law correlation is found between satellite-derived effective width We and discharge Q, where We is the water surface area within a braided reach divided by the reach length. Synthetic values of We and Q generated from a cellular automata model of stream braiding display a similar power law correlation. Power functions that are fit through plots of We and Q represent satellite-derived rating curves that can subsequently be used to estimate instantaneous river discharge from space, with errors ranging from tens to hundreds of cubic meters per second. For ungaged rivers, changes in relative discharge can be determined from satellite data alone to determine the shape and timing of annual flows in glacierized basins. Absolute discharge can probably be estimated within a factor of 2. More accurate estimates will require either (1) one or more ground measurements of discharge acquired simultaneously with a satellite image acquisition, or (2) successful parameterization of known morphologic controls such as total sinuosity 5;P, valley slope, bank material and stability, and braid channel hydraulic geometry. Values of total sinuosity 5;P derived from satellite imagery and field measurements from two rivers of braid channel width, depth, velocity, water surface slope, and bed material grain size indicate that while the shape of satellite-derived We-Q rating curves may be influenced by all of these variables, the sensitivity of flow area to changing discharge is most dependent upon the degree of braiding. Efforts to monitor river discharge from space will be most successful for intensely braided rivers with high values of total sinuosity. Subsampling of existing daily discharge records from the Iskut River suggests that satellite return times of about 1 week are sufficient for approximating the shape and timing of the seasonal hydrograph in large, glacierized basins. Although errors are large, the presented technique represents the only currently available way to estimate discharge in ungauged braided rivers.

Journal ArticleDOI
TL;DR: In this paper, a computer model for the dynamics of meandering rivers has been used to study the interplay between the migrating river and the changing sedimentary environment created by the meandering river itself.
Abstract: A computer model for the dynamics of meandering rivers has been used to study the interplay between the migrating river and the changing sedimentary environment created by the meandering river itself. The model is based on the theory of Ikeda et al. [1981] and is closely related to that proposed by Howard [1983]. Coarser sands, which are often associated with high credibility, are deposited in the point bars formed when the river migrates away from its former bank. Fine-grained material eventually fills the oxbow lakes created by cutoff processes and forms erosion-resistant plugs. In the simulations, geometric forms of individual meanders observed in different natural sedimentary environments have been reproduced by changing the credibility of the corresponding sedimentary materials, such as point bar deposits, flood plain deposits, and oxbow lake deposits. The simulations indicate that the typical meander wavelength is determined mainly by hydraulic factors such as the flow in the channel and the inclination of the underlying flood plain and is independent of the difference in the credibilities of sedimentary deposits. The computational approach permits exploration of long-term changes in the floodplain geology, mediated by the meandering river. As an initial demonstration, the formation of meander belts is investigated using the model. The results suggest that a meander belt will be formed by a river's own cutoff loops only if the characteristic time of deposition and solidification of an oxbow lake is longer than the typical time that it takes the river to migrate downstream over the distance of a meander-loop wavelength.

Journal ArticleDOI
TL;DR: In this article, a detailed analysis of Sr isotopic composition with a mass balance at Cone Pond Watershed, New Hampshire, was coupled to further constrain estimates of these processes, and the calcium weathering rate was less than half of that determined by mass balance assuming steady state conditions.
Abstract: Depletion of Ca in forests and its effects on forest health are poorly quantified. Depletion has been difficult to document due to limitations in determining rates at which Ca becomes available for ecosystem processes through weathering, and difficulty in determining changes in ecosystem storage. We coupled a detailed analysis of Sr isotopic composition with a mass balance at Cone Pond Watershed, New Hampshire, in order to further constrain estimates of these processes. Strontium acted as an analog for Ca in most processes except translocation of nutrients within forest vegetation. Variability in mineralogic and Sr isotopic composition of bedrock and soils complicated assessment of the 87Sr/86Sr ratio released to solution by weathering reactions. By conducting a mass balance on atmospherically derived Ca, it is possible to distinguish Ca weathering losses from Ca leached from ecosystem pools. The calcium weathering rate estimated by this method was less than half of that determined by mass balance assuming steady state conditions.

Journal ArticleDOI
TL;DR: In this paper, three optimization methods, shuffle complex evolution method (SCE-UA), multiple start Simplex (MSX), and local Simplex, were compared for the same CRR-catchment cases.
Abstract: From 32 CRR-catchment cases (combinations from four conceptual rainfall-runoff models (CRR) and eight catchments) calibrated with either two or three optimization methods, (1) the shuffle complex evolution method (SCE-UA), (2) the multiple start Simplex (MSX), and (3) the local Simplex, it seems that all three methods produced parameter sets of comparable, local-optimum quality. Even with comparable performance among the models, some parameter values derived by the three optimization methods for the same CRR-catchment cases are surprisingly different from each other. In addition, parameter sets of SCE-UA or MSX, which often produce marginally better results than the local Simplex at the calibration stage, could end up with worse results at the validation stage. Apparently, given the inherent limitations of calibration data, model inadequacies, and identifiability problems, it is impossible to achieve global convergence in the parameter search. However, other than those for dry catchments such as Ihimbu or Bird Creek, the parameter sets obtained are generally feasible. Both SCE-UA and the local Simplex are viable optimization tools, while MSX is inefficient computationally. SCE-UA can complete the parameter search in one run, while the local Simplex often requires multirun operations to get good results.

Journal ArticleDOI
TL;DR: In this paper, the bed shear velocity is estimated using the depth-averaged velocity in the vertically averaged logarithmic velocity profile, using a single near-bed velocity observation.
Abstract: Replicate velocity observations using conventional equipment under typical field conditions are used to evaluate the precision of different methods for estimating local boundary shear stress from velocity measurements. The bed shear velocity u, can be estimated within 3% using the depth-averaged velocity in the vertically averaged logarithmic velocity profile. To be accurate, this method is limited to relatively simple flow geometries which may be expected to have the appropriate velocity structure. Estimates of u, made using a single near-bed velocity observation are less precise by a factor of 3 because of the larger uncertainty associated with a single observation. Accuracy of this method requires appropriate flow conditions only near the bed, so it may be applied in a wider range of flow conditions, including spatially variable flow. Estimates of u, from the slope of the near-bed velocity profile are the least precise and require the most restrictive flow conditions for accuracy but offer the advantage that they may be made without independent knowledge of the bed roughness. different methods for estimating roe should be useful in judging the precision of estimates made under similar conditions. These estimates of precision, together with considerations of convenience and model accuracy, provide the basis for select- ing the most appropriate or advantageous method for different conditions and purposes.

Journal ArticleDOI
TL;DR: In this paper, the authors report results from analysis on two extensive data sets describing sea surface temperature of the Pacific Ocean and the flow of water in the Nile River, and suggest that 25% of the natural variability in the annual flow of the Nile is associated with El Nino oscillations.
Abstract: Natural variability in the annual flow of the Nile River has been the subject of great interest to the civilizations that have historically occupied the banks of that river. Here we report results from analysis on two extensive data sets describing sea surface temperature of the Pacific Ocean, and the flow of water in the Nile River. The analysis suggests that 25% of the natural variability in the annual flow of the Nile is associated with El Nino oscillations. A procedure is developed for using this observed correlation to improve the predictability of the Nile flood. A simple hypothesis is presented to explain physically the occurrence of the Hurst phenomenon in the Nile flow.

Journal ArticleDOI
TL;DR: In this article, the authors used transient tracer methods (chlorofluorocarbons (CFCs) and tritium/helium 3 (3H/3He) to determine the age of water in the Kirkwood-Cohansey aquifer system.
Abstract: Groundwater age dating through the combination of transient tracer methods (chlorofluorocarbons (CFCs) and tritium/helium 3 (3H/3He)) and groundwater flow path analysis is useful for investigating groundwater travel times, flow patterns, and recharge rates, as demonstrated by this study of the homogeneous shallow, unconfined Kirkwood-Cohansey aquifer system in the southern New Jersey coastal plain. Water samples for age dating were collected from three sets of nested observation wells (10 wells) with 1.5-m-long screens located near groundwater divides. Three steady state finite difference groundwater flow models were calibrated by adjusting horizontal and vertical hydraulic conductivities to match measured heads and head differences (range, 0.002–0.23 m) among the nested wells, with a uniform recharge rate of 0.46 m per year and porosities of 0.35 (sand) and 0.45 (silt) that were assumed constant for all model simulations and travel time calculations. The simulated groundwater travel times increase with depth in the aquifer, ranging from about 1.5 to 6.5 years for the shallow wells (screen bottoms 3–4 m below the water table), from about 10 to 25 years for the medium-depth wells (screen bottoms 8–19 m below the water table), and from about 30 to more than 40 years for the deep wells (screen bottoms 24–26 m below the water table). Apparent groundwater ages based on CFC- and 3H/3He-dating techniques and model-based travel times could not be statistically differentiated, and all were strongly correlated with depth. Confinement of 3He was high because of the rapid vertical flow velocity (of the order of 1 m/yr), resulting in clear delineation of groundwater travel times based on the 3H/3He-dating technique. The correspondence between the 3H/3He and CFC ages indicates that dispersion has had a minimal effect on the tracer-based ages of water in this aquifer. Differences between the tracer-based apparent ages for seven of the 10 samples were smaller than the error values. A slight bias toward older apparent ages, found not to be statistically significant, was noted for the 3H/3He-dating technique relative to the CFC-dating technique. This result may be caused by enrichment of local air in CFC-Il and CFC-12 from urban and industrial sources in the northeastern United States and minor contamination from sampling equipment. The demonstrated validity of the combined tracer-dating techniques to determine the age of water in the Kirkwood-Cohansey aquifer system indicates that groundwater flow models can be refined when apparent ages based on 3H/3He- and CFC- dating are used as calibration targets.