Journal of Hydrologic Engineering

Vertical velocity variance data were derived from remotely sensed mixed layer turbulence measurements at the Atmospheric Boundary Layer Experiments (ABLE) facility in Butler County, Kansas. These measurements and associated data were provided by a collection of instruments that included two 915 MHz wind profilers, two radio acoustic sounding systems, and two eddy correlation devices. The data from these devices were available through the Atmospheric Boundary Layer Experiment (ABLE) database operated by Argonne National Laboratory. A signal processing procedure outlined by Angevine et al. was adapted and further built upon to derive vertical velocity variance, $\overline{w^{\prime 2}}\text{,}$ from 915 MHz wind profiler measurements in the mixed layer. The proposed procedure consisted of the application of a height-dependent signal-to-noise ratio (SNR) filter, removal of outliers plus and minus two standard deviations about the mean on the spectral width squared, and removal of the effects of beam broadening and vertical shearing of horizontal winds. The scatter associated with $\overline{w^{\prime 2}}$ was mainly affected by the choice of SNR filter cutoff values. Several different sets of cutoff values were considered, and the optimal one was selected which reduced the overall scatter on $\overline{w^{\prime 2}}$ and yet retained a sufficient number of data points to average. A similarity relationship of $\overline{w^{\prime 2}}$ versus height was established for the mixed layer on the basis of the available data. A strong link between the SNR and growth/decay phases of turbulence was identified. Thus, the mid to late afternoon hours, when strong surface heating occurred, were observed to produce the highest quality signals.

Flood warning systems in urban areas have greatly improved in recent years with the advent of geographical information systems (GIS), radar-based rainfall estimation using next generation radar (NEXRAD), and delivery systems on the internet. The issue of flood warning and alert was brought to the forefront of public attention in Houston, Texas after the devastation resulting from the Tropical Storm Allison flood of June, 2001. This flood was reported by the National Oceanic and Atmospheric Administration as the most damaging urban flood in U.S. history, and certainly the most devastating event to impact Houston and Harris County with over $5 billion in total damages, and about 50,000 damaged structures. Over $1.5 billion in damages to the Texas Medical Center (TMC) alone were reported, with additional major impacts to the local community. While more traditional gage-based alert systems often provide flood notification to many people, the combination of GIS, NEXRAD, and the internet allows for the more efficient accumulation of rainfall data in real-time. This data can then be used in hydrologic models to estimate peak flows in a basin. This paper will illustrate the design, operation, and performance of an advanced flood warning system, developed for the TMC, which utilizes NEXRAD radar for hydrologic prediction in the Brays Bayou watershed of Houston, Texas. It will specifically document and review its use during the T.S. Allison storm event and other historical events.

Droughts are destructive climatic extreme events that may cause significant damage both in natural environments and in human lives. Drought forecasting plays an important role in the control and management of water resources systems. In this study, a conjunction model is presented to forecast droughts. The proposed conjunction model is based on dyadic wavelet transforms and neural networks. Neural networks have shown great ability in modeling and forecasting nonlinear and nonstationary time series in a water resources engineering, and wavelet transforms provide useful decompositions of an original time series. The wavelet-transformed data aid in improving the model performance by capturing helpful information on various resolution levels. Neural networks are used to forecast decomposed subsignals in various resolution levels and reconstruct forecasted subsignals. The model was applied to forecast droughts in the Conchos River Basin in Mexico, which is the most important tributary of the Lower Rio Grande/Bravo. The performance of the conjunction model was measured using various forecast skill criteria. The results indicate that the conjunction model significantly improves the ability of neural networks to forecast the indexed regional drought.

Persistence is the most important property in any hydrologic design concerning the storage capacity of reservoirs, average return periods, failure risks, and drought properties. Its consideration in analytical derivations of design criteria presents difficulties, especially in autocorrelated hydrologic processes, and for this reason, most often the analytical expressions are obtained on the basis of non-persistent (independent) cases. Although the conventional autocorrelation coefficients and function are used in many hydrological design problems, the very definition of the autocorrelation function requires that the underlying hydrologic process generating mechanism abide with normal (Gaussian) probability distribution function in addition to other restrictive assumptions. Since almost all of the analytical stochastic approaches are based on the normality assumption, it is necessary to transform non-Gaussian distributions to the normal distribution to use analytical expressions. During the transformation process, the very genuine persistence property of the basic hydrologic variables is not preserved, even when the statistical parameters such as the average, standard deviation, skewness coefficient, kurtosis, etc., are maintained in the transformed data. This shortcoming in the autocorrelation function is by-passed by the introduction and use of an autorun function, which is probability distribution free and a robust parameter. Its basis is the conditional probability statement which does not require any assumptions in practical applications. Various practical and simple hydrological design quantities are developed on the basis of the autorun coefficients without considering the conventional autocorrelation structure. The application of methodology is presented for John’s river in Florida.

A discretization of the continuous Kalinin-Milyukov-Nash-cascade is performed with the help of state-space analysis for hydrological forecasting of streamflow. A sample-data system approach is used during the discretization and results in discretely coincident values with the continuous model. The sample-data system uses input values measured instantly in time and assumes linear changes in the value of the input variable between discrete data measurements. Such description gives a generalized formulation of the pulse-data system approach often used in system engineering and discrete-time analysis of hydrological systems. An example is given to demonstrate that the approach results in improved forecasts of stream-discharge values when compared with the more traditional pulse-data system approach.

The feasibility of training an artificial neural network (ANN) for accurately predicting transient water levels in a complex multilayered ground-water system under variable state, pumping, and climate conditions is demonstrated. Using real-world data, an ANN was developed for a public supply wellfield and ground-water monitoring network located near Tampa Bay, Florida. The ANN was trained to predict transient water levels at 12 monitoring well locations screened in different aquifers in response to changing pumping and climate conditions. The trained ANN was then validated with ten sequential seven-day periods, and the results were compared against both measured and numerically simulated ground-water levels. The absolute mean error between the ANN predicted and the measured water levels is 0.16 m, compared to the 0.85 m absolute mean error achieved with the calibrated numerical model at the same locations over the same time period. The ANN also more closely reproduced the dynamic water level responses to pumping and climate conditions. The practical implication is that if ANN technology can achieve superior ground-water level predictions, it can be used to improve management strategies for a wide range of ground-water problems, from water quantity to water quality issues. It can also serve as a powerful sensitivity analysis tool for quantifying interrelationships between different variables, fostering a better understanding of the hydrogeologic system, and improving future modeling endeavors. And while physical-based numerical modeling retains some advantages over the ANN technology, both approaches may be used in a complementary fashion to achieve sound decision-making for complicated ground-water management problems.

The determination of aquifer diffusivity from water level observations in various wells close to a stream, in response to a linear increase in the stream stage, is discussed. It is noted that an earlier analytical solution proposed for this problem is incorrect. A recently proposed explicit expression for diffusivity, based on this erroneous solution, is analyzed and more accurate expressions are derived. The correct analytical solution is listed and new approximate expressions are suggested for explicit evaluation of diffusivity from water level measurements.

Estimation of the design flood flow for hydraulic structures is often performed adjusting probabilistic models to daily mean flow series. This may cause an underdesign of the structure capacity with possible risks of failure, because instantaneous peak flows may be considerably larger than daily averages. Estimation of the instantaneous peak flows is often needed for simulation of the reservoir operation based on the hydraulic characteristics of the structures and the operation rules. Because there is often a lack of instantaneous flow data at a given site of interest, the peak flow has to be estimated. This paper develops a new method to estimate the instantaneous peak flow from mean daily flow data. This methodology was successfully applied to a series of flow information from gauging stations in Brazil, with important improvements compared to the traditional methods available in the literature.