Journal of Hydrologic Engineering

The finite-difference model MODFLOW was used to model the flow of meteoric groundwater discharge (MGWD) into the Indian River Lagoon (IRL). The model was calibrated by four methods: (1) estimation of the root-mean-squared error (RMSE); (2) estimation of the Nash-Sutcliffe efficiency (NSE) index; (3) testing the null hypothesis by the two-sided test; and (4) visual comparisons. The statistics for the first three methods were obtained by comparing model-predicted and measured freshwater hydraulic head nodal values at 14 measurement points. Visual comparisons were made by comparing model-predicted and measured freshwater hydraulic head equipotential lines below the entire IRL over a domain approximately 2 km wide. Model validation confirmed that calibration by visual comparison was very accurate. The annual MGWD values obtained by the statistical calibration techniques were compared with the visual calibration results. Acceptable calibration with RMSE values ranging from 0.18–0.31 m underpredicted or overpredicted the annual MGWD by 28–213%. The annual MGWD values were off by 105% even when the two-sided test did not reject the null hypothesis. Acceptable calibration by the NSE statistic yielded excellent results at one transect but underestimated the MGWD by 39% at the other transect. Negative NSE index values always correlated with poor calibrations, although positive values did not always correlate with good calibration. Sensitivity analyses showed that the vertical hydraulic conductivity, $K_v$, was the most important factor governing MGWD into the IRL because it was orders of magnitude smaller than the lateral hydraulic conductivity. The predominant $K_v$ value was $0.015\mathrm{m}/\mathrm{day}$, but in some regions $K_v$ was only $0.00015\mathrm{m}/\mathrm{day}$, indicating that clogging can become much more significant near estuarine beds compared to river beds, perhaps due to negligible currents near estuarine beds. Results also showed that use of a single anisotropy ratio value to represent the entire domain could lead to significant errors.

Infiltration is the process of water movement from the ground surface into the soil and is an important process in the hydrological cycle. In this study, a mathematical equation of the upward-advancing wet front is derived from the Green and Ampt infiltration model, and a new method is proposed to estimate the values of soil hydraulic parameters such as unsaturated hydraulic conductivity ($K$) and the capillary pressure head (${\Psi }_f$). In the proposed method, these parameters are estimated from the advancing wet front data or infiltration test data under constant head. Experiments in which water flow is entered into the unsaturated soil column from beneath under constant head and the advancing wet front is recorded with the time are conducted on the five uniform sandy soils for different values of the saturation ratio (Sr). Results indicated that hydraulic parameters of unsaturated soils are not constant values. With increasing soil water content, $K$ is decreased from $1.69K_s$ at $\mathrm{Sr}=0\%$ to $K_s$ at $\mathrm{Sr}=100\%$. But the value of ${\Psi }_f$ is first decreased and then increased until reaching zero at $\mathrm{Sr}=100\%$.

The availability of new technologies for the measurement of surface elevation has addressed the lack of high-resolution elevation data, which has led to an increase in the attraction of automated procedures based on digital elevation models (DEMs) for hydrological applications, including the delineation of floodplains. In particular, the exposure to flooding may be delineated quite well by adopting a modified topographic index ($T{I}_m$) computed from a DEM. The comparison of $T{I}_m$ and flood inundation maps (obtained from hydraulic simulations) shows that the portion of a basin exposed to flood inundation is generally characterized by a $T{I}_m$ higher than a given threshold, $\tau$ (e.g., equal to 2.89 for DEMs with cell size of 20 m). This allows the development of a simple procedure for the identification of flood-prone areas that requires only two parameters for the calibration: the threshold $\tau$ and the exponent of $T{I}_m$. Because the modified topographic index is sensitive to the spatial resolution of the DEM, the optimal scale of representation for the performance of the method is investigated. The procedure is tested on the Arno River Basin by using the existing documentation of flood inundations produced by the Arno River Basin Authority for calibration and validation. This approach is applied on 11 subcatchments with areas ranging from 489–$6,929{\mathrm{km}}^2$ utilizing DEMs of different resolutions with cell sizes ranging from 20–720 m. Results show that the proposed procedure may help in the delineation of flood-prone areas, especially in basins with marked topography. The method is sensitive to the DEM resolution, but a cell size of $\sim 100\mathrm{m}$ is sufficient for good performance for the catchments investigated here. The procedure is also tested by adopting DEMs from different sources, such as the shuttle radar topography mission (SRTM) DEM, ASTER global DEM (GDEM), and national elevation data. This experiment highlights the reliability of the SRTM DEM for the delineation of flood-prone areas. A useful relationship between model parameters and the reference scale of the DEM was also obtained, providing a strategy for the application of this method in different contexts. The use of the modified topographic index should not be considered as an alternative to standard hydrological-hydraulic simulations for flood mapping, but it may represent a useful and rapid tool for a preliminary delineation of flooding areas in ungauged basins and in areas where expensive and time-consuming hydrological-hydraulic simulations are not affordable or economically convenient.

With reservoir drawdown, the groundwater table in the adjacent aquifer falls down correspondingly. It is useful to calculate the groundwater table variation as a function of time during reservoir drawdown for hydraulic and hydrological purposes. The Boussinesq equation with a moving boundary is applied to analyze the groundwater table variation in the unconfined aquifer during reservoir drawdown. This approach assumes a negligible seepage face. Because the moving boundary condition in the mathematical formulation precludes analytical solutions even for the linearized Boussinesq equation, we have transformed the Boussinesq equation into an advection–diffusion equation to address the negligible seepage face and the moving boundary condition. On the basis of the Laplace transformation, we yield an analytical solution of a fixed boundary problem, which is further simplified to upper and lower polynomial solutions for convenient practical use. The polynomial approximate solutions are satisfactorily compared with a number of numerical simulations of the nonlinear Boussinesq equation. The results indicate that the polynomial solutions match well with the numerical solution, but demonstrate that the replacement of the sloped reservoir–aquifer interface by a vertical interface may cause errors of up to 10% of the height of the reservoir drawdown in the prediction of the groundwater table location. On the basis of the polynomial solutions, a methodology is provided to determine the ratio of hydraulic conductivity to specific yield along with a chart for convenient practical use. The limitation of the present study is that the presented solution tends to underestimate the groundwater table with seepage face neglected for rapid drawdown, high specific yield, low hydraulic conductivity, or mildly sloped interface cases.

Variability in surface area for 44 glaciers in Wyoming’s Wind River Range was estimated through a comparison of historic aerial photography from 1966 to 2006. The total surface area of the 44 glaciers was estimated to be $45.9\pm 1.6{\mathrm{km}}^2$ in 1966 and $28.6\pm 0.4{\mathrm{km}}^2$ in 2006, a decrease of 38%. Glacial surface area was also estimated utilizing resampled aerial photography to assess the relationship between area and measurement scale. Aerial photographs were resampled to resolutions of 10 m, 15 m, 22.5 m, and 30 m to represent other satellite image resolutions used for evaluating glacier boundaries. The results show a linear decrease of total glacier area as resolution decreases. When comparing 1-m resolution to the resampled 30-m resolution for 1966 and 2006 photos, an average area difference of 5% was observed. Additionally, the current research was compared to previous research efforts that utilized Landsat imagery to detect glacier area change from 1985 to 2005. We concluded that fine spatial resolution remotely sensed imagery remains the preferred and most accurate source for measuring glacier characteristics.

This study presents a new method to estimate the Nash model parameters on the basis of the concept of geomorphologic dispersion stemming from spatial heterogeneity of flow paths within a catchment. The proposed method is formulated by including physically meaningful characteristic velocities for channel and hillslope and also takes account of the effect of complex interactions between channel and hillslope hydrological behaviors on catchment responses. We applied the proposed formulas to the Bocheong watershed, an experimental area established under the International Hydrological Programme (IHP) in Korea, with several storm events to assess the individual effect of channel and hillslope on hydrological responses in the study site. The characteristic velocities were estimated by topographic data based on a $20\times 20\mathrm{m}$ digital elevation model (DEM) from a $1:25000$-scaled topographic map and statistical features of the historical events. We then calculated the Nash model parameters by substituting the estimated characteristic velocities into the new formulas proposed in this study. The sensitivity analysis results of the Nash model parameters and instantaneous unit hydrograph (IUH) shape of the characteristic velocities indicated that the scale parameter of the Nash model was more sensitive to the hillslope velocity than the channel velocity, whereas all of the Nash model parameters were determined by the relative difference between the hillslope and channel characteristic velocities. The rising limb of IUH and time to peak were dependent mainly on the channel velocity, whereas the recession limb of the IUH reacted very sensitively to the variation of the hillslope velocity. In addition, the skewness of the IUH varied with the ratio of characteristic velocities. Finally, the IUHs, estimated from regional analysis of the characteristic velocities, led to acceptable and constant hydrological responses for the catchment scale. If the improved regional analysis, modeled on hydrodynamic (or physical) approaches, is performed, the proposed formulas for the Nash model in this study can be a useful tool to simulate rainfall-runoff processes in ungauged basins.

The second-generation radar-based flood alert system (FAS2) has been operational for more than a decade, and provides accurate advanced warning to the Texas Medical Center (TMC) in the Brays Bayou watershed of southwest Houston. Over the past nine years, FAS has generally predicted floods with average differences of 0.88 h in time of peak and 3.6% in peak flows and a $R^2$ value of 0.90 for the overall performance. After years of urbanization in the watershed, the hydrologic model as a core engine in the system has been updated with recent routing information and loss rate factors to provide more reliable prediction information. Recently, the system has been upgraded into its third generation (FAS3) by incorporating a hydraulic prediction tool—FloodPlain Map Library (FPML)—so that mapped warning information in geographic information systems (GIS) and Google Maps can be provided to emergency personnel in real time. In addition to briefly reviewing the theory and design of the flood warning system for Brays Bayou, this paper focuses on recent improvements with a summary of the flood alert system’s performance over the past years, including three recent major events: Hurricane Ike on September 13, 2008, a smaller event on April 27–28, 2009, and a complex event on July 2, 2010.

Hydrodynamic modeling is usually necessary for predicting water deliveries to marshes from source reservoirs. A novel approach is developed that decouples the delivery structure hydraulics from the marsh hydrodynamics, allowing these components to be analyzed both separately and in combination. This approach is applied to assess the effectiveness of incorporating spreader canals into water delivery systems in Everglades National Park. The results show that Manning’s $n$ in the marsh can be reasonably approximated as a function of $VR$, in which $V$ is the flow velocity and $R$ is the hydraulic radius; spreader canals can provide substantial percentage increases in water deliveries compared to the smaller structure tailwater pools, and spreader canal outflows can be linear functions of the length of the spreader canal.

The Food and Agricultural Organization of the United Nations (FAO)-56 version of Penman-Monteith (PMF-56) model has been established as a standard for calculating reference evapotranspiration ($E{T}_o$). An important constraint of application of the PMF-56 model is the requirement of solar radiation, wind speed, air temperature, and humidity data, which may not be available for a given location, especially in developing countries. The Hargreaves (HG) and Priestley-Taylor (P-T) equations are simple equations that require few weather data inputs, although regional calibration of the equations is needed for acceptable performance before applying them for $E{T}_o$ estimation. In this study, the HG and P-T equations were calibrated on the basis of the PMF-56 method in arid and cold climates of Iran using data from 12 stations during 1994–2005. After calibration of the HG equation, the average value of the adjusted HG coefficient for arid climate was 0.0031, which is about 34% higher than the original value (0.0023). Similarly, the average value of the new HG coefficient for cold climate was 0.0028, which is about 22% higher than the original value. The results showed that the original P-T coefficient of 1.26 was very low for the climatic regions, and the new P-T coefficients of 1.82 and 2.14 have the best fit as compared with the PMF-56 method in cold and arid climates, respectively. Overall, calibration of the HG and P-T equations resulted in improvements of the equations by reducing the errors of the $E{T}_o$ estimates.