Journal of Hydrologic Engineering

The forecasting of future rainfall is one way to extend the lead time of flood forecasting. This study proposes a rainfall forecast model based on a gray model, in which model parameters are estimated by using a fuzzy goal regression method. A one-time step forecasting technique is introduced to improve forecasting. This improvement allows for forecasting further ahead of the event in question. Ten historical storm events are considered to calibrate and verify the proposed model. Simulation results indicate that the proposed model exhibits reasonable rainfall mass curves and allows for rainfall hyetograph forecasting 1–3 h in advance.

A simple and efficient formulation is developed to describe the effects of wind speed and direction on ground level precipitation intercepted by steep mountain slopes. A 3D rainfall field is reproduced from measurements of (vertical) precipitation intensity, wind speed, and wind direction. The Marshall-Palmer drop size distribution is used to express the volumetric precipitation water content as a function of measured precipitation intensity, and raindrops are assumed to move horizontally with the measured wind speed and direction. Land topography is described using digital elevation model data, and local contributions to ground level precipitation are calculated as the interception of the obtained 3D rainfall field by horizontal and vertical surfaces that constitute the elemental land surface systems. The developed formulation is tested at the Acquabona and Fiames mountain slopes, located in the Northern Italian Dolomites, near the town of Cortina d'Ampezzo, where debris flow phenomena often occur. Simulation results are corroborated, although in an indirect and approximate manner, by field estimates of debris flow volumes delivered by the considered mountain slopes in response to monitored storm events. Although more accurate and comprehensive validation is needed, the developed formulation appears to constitute a useful diagnostic tool for providing interpretation of storm-flow hydrographs delivered by steep mountain slopes in response to storm precipitations affected by wind.

A model is presented that can simulate infiltration from rainstorms on areas exhibiting random variation in saturated hydraulic conductivity K_S. Heterogeneity in the capillary drive (or length scale) parameter G can be treated as well. The method is based on a point infiltration model that includes the Green-Ampt or Smith-Parlange infiltration functions. The runoff area is characterized as an ensemble of infiltrating points or flow path strips that provides runoff to a receiving channel. The model is developed by simulation of a large ensemble using Latin hypercube sampling. The infiltration expression is responsive to a changing rainfall rate r and is easily characterized using the basic infiltration parameters K_S and G, plus a third parameter based only on the coefficient of variation of K_S or G. Areal heterogeneity causes a rainfall-dependent change in the areal effective value for K_S, called K_e(r). The infiltration expression contains rainfall rate as a variable, and observed storms with temporal rainfall patterns may easily be treated. Moreover, the new expression eliminates the explicit concept of ponding time as a separate calculation. The effect of heterogeneous infiltration parameters is demonstrated using several field cases.

Modeling guidelines of parsimony (simplicity of assumptions), accuracy, modesty, and testability are discussed in terms of urban snowmelt induced runoff. Heterogeneity and variability at the 10–100-m spatial and <1 day timescales of snow distribution, energy availability, and surface permeability in urban environments are outlined. The snow accumulation and melt routines of three drainage models that have been applied to urban settings are reviewed; two of these, MouseNAM and SWMM, were designed for urbanized catchments; the third, HBV, is a regional-scale model for rural catchments. The surface hydrology of the models have varying degrees of complexity. All contain a temperature index for melt—this method is shown to be theoretically unsound without modification for urban simulations. There are discordant scales between processes important for snowmelt runoff generation and those modeled. Literature on model development, validation, and application is lacking. Urban snow data are also sparse, which limits the possibility of assessing or improving model performance.

Egypt depends on the Nile River for all of its water resources. Using the streamflows' history, the large fluctuations of the Nile flood cause the best predictions to be unsatisfactory. The purpose of this paper is to stochastically forecast the Nile summer runoff one-season-ahead using, as inputs, an El $\mathrm{Ni}ño-\mathrm{southern}$ oscillation (ENSO) sea surface temperatures (SSTs) signal in the East Pacific and SSTs in the South Indian Ocean. Causality between inputs and outputs is established, and a multiple-input transfer function with noise (TFN) model is built for forecasting purposes. The model explains 63% of the variability of the Nile flood with relatively stable parameters. The mean of absolute percentage error of forecasts is 6% calculated on a data set that was not used in the parameter estimation. The model is parsimonious, and its behavior agrees with the most recent studies in climatology. The forecasting ability of the model is high for extreme floods and severe drought years, except when the South Atlantic Ocean displays a strong warm signal opposite to the El $\mathrm{Ni}ño-\mathrm{southern}$ oscillation cold signal.

A simple method for transmuting popular synthetic unit hydrographs (SUHs), such as those of Snyder, the Soil Conservation Service (SCS), and Gray, into the Gamma distribution has been proposed, the calculations for which can be performed on a calculator. It gives a smooth shape to an SUH, the area under which is guaranteed to be unity. Previous methods of SUH are shown to be the special forms of the Gamma distribution. The SUHs obtained using these methods are observed to be in serious error. The peak rate coefficient C_p in Snyder's method and the peak rate factor D_f in the SCS method have imparted physical and conceptual meaning. The possible unified conceptual interpretation of the popular SUHs in line with the conceptual models of instantaneous unit hydrograph (IUH) has been presented. The method can do away with the calculations for W₅₀, W₇₅ and the exceptionally long tail of UH as is usually obtained using Snyder's method for fast responding catchments. It opens up the possibility for a variable dimensionless UH in the SCS method and gives a unified basis for the three popular methods in particular and the other methods in general. Application of the method has been illustrated through problems.

Soil moisture plays a key role in various fields such as agriculture, meteorology, ecology, hydrology, and geotechnical engineering. A critical problem in the measurement of soil moisture through remote sensing methods is representing the spatial heterogeneity without distortion. In this study, a geostatistical technique, kriging, is used to characterize the spatial structure of the Little Washita soil moisture data measured by passive microwave remote sensing as part of the 1992 Washita study. Second-order statistical metrics verify the adequacy of kriged estimates against the observed soil moisture field. The results show that sampling of a systematic set of points produces a smoother soil moisture field compared to a random sampling of data points. To check the efficacy of kriging in estimating the unknown soil moisture values and related surface fluxes, evaporation is calculated. There appears to be a good correspondence between the evaporation values calculated from the original data and the kriged data. The frequency scaling ratio is used to identify the spatial coherence between the observed and the estimated soil moisture fields.

Water resources systems operation requires drought risk estimates to mitigate possible drought-related damages. Drought risk assessment is complicated by the nonlinear interaction of the atmospheric hydrologic and oceanic systems where highly varied hydrologic system responses to similar drought-forcing phenomena can occur. A methodology capable of assessing drought risk associated with hydroclimatic events by using a simplified climate model is presented. Ensemble mean and standard deviations of hydrologic water storage represent the expected hydrologic system response to the hydroclimatic event. Relative frequency histograms and cumulative distribution functions characterize the range of hydrologic system responses that can occur and are used to obtain the spatially and temporally evolving drought risks. The methodology is presented in a framework suitable for application to resources management. An outline of the approach, description of the simplified climate model used in this study, and an illustrative example using a La $\mathrm{Ni}ña$ type event as the drought-forcing mechanism are given. Simulation results, the methodology, and future directions are discussed.

During urban storms, the overloading of a sewer system or a riverbank overflow produces flows that are routed by the road infrastructure and causes flooding of adjacent built-up areas. The evaluation of the damage inflicted in these built-up areas requires the determination of important hydraulic parameters of the inundation, such as the maximum water depth or the inundation duration. For each independent hydraulic urban cell, these parameters depend upon its intrinsic hydraulic properties (perviousness and storage capacity) and on the characteristics of the flood outside the cell. This paper proposes a quantitative formulation of the concept of cell hydraulic vulnerability based on the cell's hydraulic properties and on the basic characteristics of the flood (peak flow and time to peak). This study is based on observations made in three districts of Ouagadougou, Burkina Faso, and on a 23-year record of rainfall events in this region. The hydraulic vulnerability is discussed according to the district hydraulic properties and to the flood event characteristics. The developed methodology could be used in principle in every city where built-up areas are highly subpartitioned into walled properties.

Because eroded sediments are produced from different sources throughout a basin, it is often advantageous to model sediment delivery processes at basin scale using a spatially distributed approach. In this paper, a sediment delivery distributed (SEDD) model applicable at morphological unit scale, into which a basin is divided, is initially proposed. The model is based on the Universal Soil Loss Equation (USLE), in which different expressions of the erosivity and topographic factors are considered, coupled with a relationship for evaluating the sediment delivery ratio of each morphological unit. Then the SEDD model is calibrated by sediment yield and rainfall and runoff measurements carried out, at annual and event scales, in three small Calabrian experimental basis. At event scale, the analysis showed that a good agreement between measured and calculated basin sediment yields can be obtained using the simple rainfall erosivity factor; the agreement is independent of the selected equations for estimating the topographic factors. The analysis developed at annual scale showed that the model reliability increases from the event scale to the annual scale. Finally, a Monte Carlo technique was used for evaluating the effects of the uncertainty of the model parameters on calculated sediment yield.

Spring runoff prediction in the Red River Valley, southern Manitoba, Canada, is an important issue because of the devastating effect of the flood of 1997 in that area. Increasing the accuracy of the prediction process is a practical necessity. This study looks at the artificial neural networks (ANN) technique and compares it to linear and nonlinear regression techniques. The advantages and disadvantages of the three modeling techniques are discussed. To fill the predictive accuracy evaluation gap left by the mean squared error and the mean relative error, a modified statistic, namely, pooled mean squared error, is developed and explained. The aim of this work is to show the applicability of ANN for runoff prediction and to evaluate their performances by comparing them with traditional techniques. In this study, according to the accuracy of results, the ANN models show superiority in most of the cases. However, in some situations, the performance of the other two techniques was comparable.

The artificial neural networks (ANNs) that try to mimic the functioning of the human brain are a powerful tool for input-output mapping. The setting up of a stage-discharge relation is an important part of the processing of streamflow data. Three-layer feedforward ANNs have been used to model river-rating curves. The results show that the ANN approach is much superior as compared to the conventional curve-fitting approach. The ANN is also able to model a loop-rating curve (hysteresis effect) very well.