Journal of Hydrologic Engineering

Application of jet erosion tests (JETs) to study in situ erodibility is gaining popularity. New versions of the JET (original JET versus mini-JET) and new data analysis techniques have introduced questions regarding their operation and data collection procedures. One of the major issues regarding JETs is the high degree of variability of the erodibility parameters (i.e., erodibility coefficient, $k_d$, and critical shear stress, ${\tau }_c$). This variability has been attributed to heterogeneity in different soil properties under natural field conditions, but limited research has quantified variability under controlled laboratory conditions, especially for the newer mini-JET. This study uniquely conducted 20 mini-JETs under controlled laboratory conditions on each of two soil types of contrasting texture. Mini-JETs were conducted in situ on streambanks of these same soils in previous research. The laboratory mini-JETs predicted similar values of most parameters with much less variability than in the field. Three to five mini-JETs conducted in the laboratory estimated erodibility parameters with a precision of 25% at a 95% confidence level. Laboratory JETs on disturbed, remolded samples provided baseline estimates of in situ erodibility parameters. Additional mini-JETs were conducted at three different head settings on the two soil types. The influence of the head setting was dependent on the soil type, solution technique, and detachment model. In general, variability in derived erodibility parameters increased at larger head settings especially for the less-erodible soil. Existing mini-JET data were resampled to evaluate the effect of the initial time interval and termination time interval of data collection on derived erodibility parameters. Both initial and termination time intervals were most influential at larger head settings. When the applied pressure head is sufficient to create scouring upon test initiation, an initial time interval of at least 30 s and a termination time interval of at least 300 s is recommended for less-erodible soils.

This paper describes a method for estimating runoff hydrographs for dam safety evaluations in which the time of concentration is considered dependent on the magnitude of the excess rainfall intensity. Current practice for hydrologic routing assumes the time of concentration to be invariant and a function of basin characteristics. The proposed method addresses the limitations arising from the assumed linearity in the unit hydrograph by showing that the time of concentration is dependent on the excess rainfall intensity through a power relationship. The proposed method uses a variable flow velocity approach to compute travel time and construct time-area curves for a range of excess rainfall intensities. From the derived relationships between time of concentration, flow, and excess rainfall intensity, an explicit nonlinear convolution is performed to compute runoff hydrographs. In its current implementation, the method assumes a uniform distribution of rainfall over the entire watershed. This method is applied to a $282\text{-}{\mathrm{km}}^2$ gauged watershed in Northern California and computed hydrographs are compared with measured runoff hydrographs for six large storms. The proposed method is then compared with constant time of concentration predictions. It is shown that using the traditional invariant time of concentration would require different times of concentration for different storms to match measured discharges. By incorporating a variable flow velocity field and a variable time of concentration, surface hydrologic predictions would reproduce more closely measured discharge. The time of concentration is not a property of the physical characteristics of a watershed alone but a variable that depends also on the intensity and temporal distribution of precipitation. Because the method explicitly accounts for the nonlinear effects arising from variations in rainfall intensity, it is best suited for the evaluation of extreme events, which are characterized by large precipitation intensity pulses. It is shown that current hydrologic models based on a constant time of concentration may underestimate peak discharge associated with the extreme events used in dam safety evaluations.

In snow-dominated regions surface air temperature is expected to have a substantial effect on the magnitude of a flood during a storm event. It is risky to estimate the design flood based only on the maximum precipitation while excluding other atmospheric variables like temperature and radiation. To overcome this problem, a methodology to estimate the maximum flood is proposed based on a physically based hydrologic model with input from physically maximized storm events by means of a numerical atmospheric model. As a case study, the probable maximum floods are simulated for the Upper Feather River watershed, the Yuba River watershed, and the American River watershed that are located in a mountainous region in Northern California, from the most severe 60 historical precipitation events during 1951–2010 for each watershed. The results show that this methodology can explain the underlying physical causes for the occurrence of maximum precipitation. It also shows that the maximum precipitation, determined in terms of the maximum areal average precipitation depth over a selected watershed during a selected storm duration, does not necessarily produce the maximum flood over that watershed.

Solute transport in aquifers is strongly influenced by the spatial distribution of subsurface hydraulic conductivity ($K$), while limited drilling typically results in lack of data characterizing both the $K$ and the in situ fluid-flow boundary conditions (BC). To characterize such environments, this paper presents an efficient direct inverse method to simultaneously identify an aquifer’s $K$ pattern, values, and flow field. The method ensures fluid-flow continuity using local approximate solutions of the governing flow equation conditioned to limited measurements, while the physics of flow are enforced, making the inverse problem well-posed. A single system of equations is assembled and solved, from which parameters and BC can be simultaneously estimated. For problems with irregular and regular $K$ distributions, inversion is demonstrated for different measurement types, qualities, and quantities. When measurement error is increased, the estimated $K$ pattern is largely insensitive to the error, although the inverted flow field suffers greater inaccuracy. Local conductivity and Darcy flux measurements are found to have similar information content, although subtle differences exist in the inversion outcomes when long-term contaminant release is simulated. Local conductivity measurements lead to better identification of conductivity pattern, values, and hydraulic head field; Darcy flux measurements lead to more-accurate estimation of the velocity field and thus improved transport predictions. Overall, velocity field estimated by the direct inverse method based on hydraulic measurements can lead to reasonable predictions of contaminant migration under unknown aquifer BC.

Precipitation data are critical for developing a fundamental understanding of hydrologic processes and reducing flood losses, but the acquisition of reliable and representative precipitation data is extremely challenging in high mountain environments, such as the Himalayas. Remotely sensed and ground-based observations are the two primary sources of precipitation data, but can have important biases, particularly in areas of complex and steep topography. Remotely sensed [Tropical Rainfall Measuring Mission (TRMM)] and ground-based precipitation measurements in the Sutlej River basin, India, were compared to better understand the biases and uncertainties associated with variable storm processes and precipitation sampling programs in high mountain environments. Results highlight how extreme rainfall values were the highest across the Lower Himalayas where synoptic weather systems are more prevalent, whereas extreme convective intensities were the highest across the Inner Himalayas, where orography triggers convective precipitation events. In addition, deviations between extreme precipitation values measured by TRMM and ground-based observations reversed with increasing elevations, with TRMM measurements underestimating precipitation relative to ground-based observations for lower elevation sites and overestimating precipitation at the higher elevation sites. Finally, large subgrid differences in the variance of the extreme rainfall distributions were found within a single TRMM pixel, which appear to be driven in large part by local topographic relief, underscoring the complexities of bias correction of gridded rainfall measurements using ground-based stations. Recommendations are made for advancements needed to apply remotely sensed precipitation data for improving understanding of physical processes and forecasting of floods in mountain areas.

Previous studies have shown that the shape parameter of the generalized extreme-value (GEV) distribution, a distribution commonly applied in precipitation frequency analysis, is difficult to estimate from short records because of sampling bias. This bias is confirmed in the present study using annual precipitation maxima data from NOAA Atlas 14, which covers much of the United States and its territories. The bias depends on record length as well as on the magnitude of the population shape parameter. Correcting sample shape parameter estimates for bias yields extreme precipitation quantiles that are more conservative (i.e., higher) than would be obtained without a correction. For example, in a test case presented here, the bias correction increased the 1- and 24-h 10,000-year quantiles by 8 and 10%, respectively. This paper presents an expression and coefficients that may be used to develop approximately unbiased GEV shape parameter estimates for regions covered by NOAA Atlas 14.

Flow time-series data are crucial for water resource and flood management. In catchments where flow observations are not available or data are of poor quality, regionalization methods are needed to generate flow time series. Explicit attention to data quality and uncertainty can give insight into the value of information, model predictive capability, and priorities for new and better data. While such analyses are becoming more common in the literature, there is a particular lack of knowledge about suitable modeling approaches and data priorities in tropical, monsoon-dominated regions. To address this, a case study of 44 subcatchments of the upper Ping catchment in Thailand is reported. Three regionalized rainfall-runoff indices— rainfall-runoff coefficient, base flow index, and rainfall-runoff elasticity—are used to condition the rainfall-runoff model using a Bayesian approach. The model performance is tested at daily, monthly, and seasonal timescales in terms of accuracy, reliability, and sharpness in predicting observed flows. The regionalized model was considered to be imprecise at the daily timescale and thus unsuitable for flood studies. The model is more suitable for supporting water resources planning at the monthly and seasonal timescales, with recognition of its tendency to underestimate flows at the start and end of the monsoon. The model showed poor reliability in a few catchments, implying that assumptions made in the Bayes approach were not useful in these cases and a different approach to uncertainty analysis may be justified along with further efforts to improve the rainfall-runoff model and data sets used.

Regional flood frequency (RFF) equations are statistical characterizations that, in their simplest form, show how peak discharge quantiles scale with drainage area. They remain separated from the physical processes that occur during flood-generating rainfall-runoff events. A physical understanding of RFF equations is a long-standing unsolved problem in hydrology. This study addresses this problem using a scaling framework. Two hypotheses are introduced that collectively state that scaling slopes of event peak discharges are on average equivalent to the mean scaling slope of annual peak quantiles; the later quantity refers to an average taken over a set of exceedance probabilities under consideration. To test these hypotheses, a nested mixed-effects linear (NMEL) model was developed that characterizes event-to-event variability in scaling relationships between stream discharge peaks and drainage area. The model leads to scaling relationships for event peak discharge quantiles and annual peak quantiles that are the basis for RFF equations. Because event-based scaling relationships can be connected to physical processes, the model provides a way of investigating whether quantile-based scaling relationships can be connected as well. The model was tested against data for 148 rainfall-runoff events from the $21{\mathrm{km}}^2$ Goodwin Creek Experimental Watershed (GCEW) located in the state of Mississippi. Test results support the NMEL model and the two hypotheses, but raise new questions that need to be addressed in future research.

Inappropriate artificial activities can escalate the frequency, the intensity, and the magnitude of floods and aggravate their aftermath. Additionally, urban sprawl is responsible for increasing the manifestation of flood events in coastal areas. In this study, the effects of diachronically formulated artificial stream-channel modifications and registered urban expansion on flood hazards were quantified. Initially, land-use changes were monitored over time and a detailed field topographic survey was performed. The peak discharge for a 50-year return period storm event was found to be equal to $42.3{\mathrm{m}}^3/\mathrm{s}$, as computed by the rational method. A steady flow analysis with a mixed flow regime was then performed so as to compute the water velocity and surface runoff elevation at 49 discrete cross sections for the years of 1945, 1981, and 2009, whereas the inundation extent was found equal to 12,208, 10,519, and $\mathrm{9,530}{\mathrm{m}}^2$ for the same years by implementing another model. Although the water level and inundation extent were respectively reduced by 5.5 and 9.2% in 2009 compared with that in 1981, more houses had been constructed on the riverbed, and thus the flood hazard was greater in 2009 than in 1981. Despite the fact that the flood area was larger by 16.1% in 1945 compared with 1981, the flood hazards were diminished in 1945 because the households were settled more than 78 m away from the stream centerline. Thus, in 1945, the stream would have successfully sustained flood events even with a 50-year recurrence period.

Air temperature is often used to estimate stream water temperature by developing regression models. A method using modified sine functions and sinusoidal wave functions (MSSWF) was proposed for estimating hourly water temperatures in rivers. Estimates of daily maximum and minimum water temperatures from sine functions (SFs) were corrected using linear regressions as functions of deviations of estimates of daily maximum and minimum air temperatures from SFs, and then modified sinusoidal wave functions (SWFs) were used to estimate hourly water temperatures. Excellent agreement was found between observed and estimated hourly water temperatures using MSSWF models developed for 13 rivers in Alabama. The average Nash-Sutcliffe efficiency (NSE) for the MSSWF models developed for individual rivers is 0.94. The distance of a river monitoring station from the weather station has very little effect on the performance of individual MSSWF models when the distance is less than approximately 300 km. A lumped MSSWF model was also developed by combining water temperature data from all 13 rivers. The average NSE of 0.87 from the lumped MSSWF model indicates that the model can be applied to other rivers in Alabama that lack water temperature observations. The procedure to develop a MSSWF model was applied to four rivers in other regions.

By convention, design discharge estimates are often expressed in terms of annual exceedance probability (AEP) or average recurrence interval (ARI). This paper describes the limitations of using AEP and ARI to reliably convey the expected incidence of flood magnitudes to decision makers and the public. An alternative expression—exceedance events per 1,000 years ($\mathrm{EE}/\mathrm{1,000}\mathrm{years}$)—is proposed, which offers a less ambiguous definition of design discharge frequency. A case is argued for the adoption of $\mathrm{EE}/\mathrm{1,000}\mathrm{years}$ based on several perceived advantages to AEP and ARI, including (1) by definition, it is a measure of number of flood exceedances over a timespan but can also communicate shifts in flood frequency over time, (2) the results of analysis using peak-over-threshold and annual maximum series can be both presented on the same $\mathrm{EE}/\mathrm{1,000}\mathrm{year}$ chart, and (3) $\mathrm{EE}/\mathrm{1,000}\mathrm{years}$ as a probability-based expression can be easily converted to alternatives used in floodplain risk analysis.

Most stormwater models oversimplify evapotranspiration (ET) processes, which restricts their applications in water budget analyses, ecosystem services assessments, and coupled modeling studies such as with urban climate models to investigate benefits of green infrastructure (GI) implementation. To address this need, the Penman–Monteith scheme was incorporated into the United States Environmental Protection Agency’s storm water management model (SWMM) to improve the latter’s ET routine. The modified SWMM allows heterogeneous and subdaily potential ET (PET) inputs, and includes the effects of water stress on actual ET (AET). The updated ET routine was validated for GI, including bioretention ($R^2=0.57$) and green roof ($R^2=0.78$), by comparing with direct measurements. A case study showed the upgraded SWMM model generates improved daily ET estimates and more accurate temporal patterns than previous SWMM versions. The results also suggest that water cycle fluxes could be miscalculated if a homogeneous potential ET is used in SWMM and other stormwater models. For example, the annual percolation and AET amounts could be overestimated and underestimated by 406% and 9% respectively for landscapes, and the annual AET and runoff amounts could be overestimated and underestimated by 12–19% and 14–19% respectively for GI, if the mismatched PET time series is used.