Journal of Hydrologic Engineering

The first step in probabilistic multiobjective groundwater optimization is the conceptualization of the uncertain conductivity ($K$) field. The $K$ distribution can be considered as a lognormal random variable. Two geostatistical approaches for conditional field generation, sequential Gaussian simulation (SGSIM) and sequential indicator simulation (SISIM), are used to generate sets of multiple conditional realizations of the $K$ field. The SGSIM- and SISIM-generated $K$ fields are found to have similar statistical properties, but SGSIM results in a smoother $K$ field. The solutions to a probabilistic multiobjective optimization of groundwater remediation design case study are addressed by combining the probabilistic improved niched Pareto genetic algorithm (PINPGA) with the stochastic groundwater flow and transport model. The multiobjective optimization of groundwater remediation design for the removal of trichloroethylene (TCE) plume at Massachusetts Military Reservation (MMR) in Cape Cod is used as a test case. The impact of the sets of $\ln K$ realizations are evaluated by comparing the first and second moments of the resulting three-dimensional (3D) contaminant plume and by comparing the resulting Pareto frontiers. The resulting contaminant plumes of the two techniques have similar centroids, but the standard deviation and maximum concentrations are higher with the SISIM conductivity fields. Although PINPGA is able to identify valuable Pareto-optimal strategies about uncertain remediation options given either approach to generating $K$-fields, the choice of conductivity field model impacted the Pareto frontiers. The optimization results demonstrate that with the smoothed SGSIM-based $\ln K$ realizations, the PINPGA methodology obtains a smaller range between the upper and lower confidence intervals of the Pareto frontiers.

In this study, a numerical modeling framework for simulating extreme storm events was established using the Weather Research and Forecasting (WRF) model. Such a framework is necessary for the derivation of engineering parameters such as probable maximum precipitation that are the cornerstone of large water-management infrastructure design. Here, this framework was built based on a heavy storm that occurred in Nashville, Tennessee (U.S.), in 2010, and verified using two other extreme storms. To achieve the optimal setup, several combinations of model resolutions, initial/boundary conditions (IC/BC), cloud microphysics, and cumulus parameterization schemes were evaluated using multiple metrics of precipitation characteristics. The evaluation suggests that WRF is most sensitive to the IC/BC option. Simulation generally benefits from finer resolutions up to 5 km. At the 15 km level, NCEP2 IC/BC produces better results, whereas NAM IC/BC performs best at the 5 km level. The recommended model configuration from this study is: NAM or NCEP2 IC/BC (depending on data availability), 15 km or 15–5 km nested grids, Morrison microphysics, and Kain-Fritsch cumulus schemes. Validation of the optimal framework suggests that these options are good starting choices for modeling extreme events similar to the test cases. This optimal framework is proposed in response to emerging engineering demands of extreme storm event forecasting and analyses for design, operations, and risk assessment of large water infrastructures.

The study of moisture ingress and accumulation in porous building materials constitutes an essential facet of durability analysis and design of structural elements. Because the phenomenon is closely driven by the conditions of the service environment, a rational assessment of exposure conditions is needed. The pertinence is especially critical for those conditions which bear the potential to strongly influence the state of saturation of the material—exposure to rains, for instance. This paper presents an exploratory statistical study to characterize the manifestation of tropical rainfall events and quantify the concomitant cyclic wetting-drying effect. To account for the diversity of tropical conditions the study encompasses instances of five Indian meteorological stations, each pertaining to a distinct climatic zone of the subcontinent. The analysis uses the last fifteen years’ worth of available rainfall data obtained from the archives of the Indian Meteorological Department. The derived characteristics are used to design typical rainfall scenarios for the studied stations. The delineated scheme can be implemented in general to model rainfall scenarios for any given station of interest which subsequently could be adopted to facilitate the study of moisture intrusion in porous structural materials exposed to intermittent rains.

In this study, a Copula Monte Carlo (CMC) method was proposed to improve flood risk for confluence flooding control downstream of a reservoir. The effects of reservoir operation and joint relationships between two rivers’ streamflow were both considered in the method. The Copula function modeled the dependence between the main stream and tributary; whereas the Monte Carlo (MC) method estimated the flood risk after the reservoir routing of main-stream flood and the simulated tributary flood from the Copula function. This proposed method was applied to a case study in the confluence flooding control downstream (Jinsha and Min Rivers) of Xiluodu-Xiangjiaba reservoirs, southwest China. After the simulation, the results were compared with the current MC method. Our results indicated that with the CMC method, the downstream flood risk caused by the Min River flood was 4.32, 2.35, and 1.08% for the 0.5, 1, and 2% design floods of the Jinsha River, respectively; whereas the MC method underestimated the flood risk. The CMC method is more robust than the MC method because it can consider both the flood spatial correlations and the inside flood domain stochastic characteristics. Further, this method can provide decision support for joint operation of Xiluodu and Xiangjiaba cascaded reservoirs and flood risk planning and management.

This study assesses the relative influence of four independent variables on green roof hydrological performance under rainstorm conditions. Twenty-four extensive green roofs representing all combinations of the following four design factors were used: native meadow species versus Sedum; mineral-based versus biologically derived planting medium; 10-cm versus 15-cm depth; and irrigation provided daily, sensor controlled, or not at all. From events covering the period May–October in 2013 and 2014, mean values were determined for the seasonal volumetric runoff coefficient ($C_{\text{vol}}=0.4$), peak runoff coefficient ($C_{\text{peak}}=0.12$), and U.S. Natural Resources Conservation Service (NRCS) curve number ($\mathrm{CN}=94$). Irrigation had the largest overall impact: daily irrigation increased $C_{\text{vol}}$ to 0.5 compared to 0.3 for systems with sensor-controlled or no irrigation. The biologically derived planting medium, composed of a high proportion of aged wood compost, made a significant improvement, maintaining $C_{\text{vol}}$ of 0.3 compared to 0.4 for the mineral-based product in the modules without irrigation. A similar pattern was found in the NRCS curve numbers.

Flow-measurement weirs in natural channels experience sedimentation. Removing accumulated sediment is often impractical, prohibitively expensive, or both. This paper describes the results of hundreds of laboratory experiments conducted over a 6-year period to quantify the effect of sedimentation on the discharge coefficients for short-crested 120° and 140° triangular weirs over a range of flows, channel slopes, and different measurement positions along and across the channel. Two weir crest geometries were tested. These included the U.S. Department of Agriculture, Agricultural Research Service (ARS) design, and a modified ARS-type weir crest intended to reduce the frequency of plugging by floating debris called the Panama crest design. Two different weir crest thicknesses were tested in a hydraulics laboratory in two different flumes, with and without horizontal planar sedimentation, to identify the effects of sedimentation on discharge coefficients. Sediment transport was not considered. Results showed that in all cases sedimentation reduced the discharge coefficient by up to 10%. Channel slope had a relatively minor effect provided that the channel slope was less than or equal to 2% with a horizontal planar bed above the weir. Depth measurement position was of minor importance provided that the measurement was made out of the region of drawdown near the weir. The findings reported in this paper extend the usefulness of short-crested triangular weirs to conditions where the upstream pool is filled with sediment to the weir invert.

Conditions of shortage in a water supply system (WSS) occur when the available water resources are unable to satisfy the related demand (failure). The definition of risk of shortage conventionally relies on three indexes that capture the characteristics of possible failures in terms of probability of occurrence (reliability), duration (resiliency), and intensity (vulnerability). Although the conceptual bases for these definitions are largely acknowledged, the operative way to define them can largely affect the final value of the risk of shortage, making it difficult to compare among different WSSs when different formulations are applied. In this paper, a robust method to quantify the risk of shortage for WSSs that rely on surface water is proposed. The major novelties are to consider the extreme events in the risk analysis and to overcome the issue of the representativeness of the observed time series with respect to the characteristic return periods of drought events. To this aim, a stochastic approach based on a zero mean autoregressive (AR) model of standardized precipitation indexes (SPIs) is combined with a multilinear regressive model learning from observed SPI and associated inflow anomalies. This approach has been applied to the case study of the Ridracoli Reservoir in Central Italy, taking into account several climate scenarios, as well as several configurations of the WSS. Results show the ability of the proposed procedure to obtain convergence in the risk indexes and to distinguish among different levels of shortage risk, giving additional information for drought episodes with longer return periods. In particular, the standardized indexes focusing also on extreme events led to a quantification of the risk able to capture, at least in the proposed case study, the benefit of management options aiming to reduce the vulnerability (i.e., not fully meeting the demand in anticipation of a dry period), while those considering only the average features do not.

Soil and water assessment tool (SWAT) and advanced circulation (ADCIRC) models were integrated to generate a hydrologic (SWAT)–hydrodynamic (ADCIRC) model applicable for flood prediction in coastal areas. The model is applied to the lower St. Johns River Basin for a holistic postevent analysis of Tropical Storm Fay (2008). Validation of the model is presented, followed by physical-forcing and temporal assessments of inundation within the river-adjacent watershed basins. The model validation and inundation assessment demonstrates the need to apply watershed runoff as an additional boundary condition in order to more fully capture the peak surge and recession, which added $\sim 0.5\mathrm{m}$ to storm tide elevation in the lower St. Johns River, extended the surge recession by nearly 5 days, and increased the inundated watershed area by almost 50%. Watershed inundation ranges between $\sim 33{\mathrm{km}}^2$ for normal tide conditions and $\sim 150{\mathrm{km}}^2$ for an approximate 1-in-100-year event (Tropical Storm Fay). Implementation of the approach requires careful definition of the SWAT-ADCIRC interfaces at tributaries and river offshoots, which for the case of the lower St. Johns River Basin was accomplished with 8 inflow boundary condition locations. Other details of the approach are discussed, and general guidance for application to other estuaries is provided.

Hydrologic models are not able to compute unsteady flood flows, and the hydraulic models are sometimes found less effective in uncertainty estimation in the modeling outcomes. This study presents a coupled modeling framework to analyze the unsteady flows in the downstream of the Teesta and Lachung Rivers using hydrodynamic and hydrological models. The present approach leads to unsteady flow simulations along stream channel reach. Soil and Water Assessment Tool (SWAT) model is used for the projection of streamflow, water depth, and precipitation at various gauged and ungauged locations over the selected catchment using Coupled Model Intercomparison Phase 5 (CMIP5) CM3 model data sets with their representative concentration pathway (RCP) experiments and observed hydrometeorological data. SWAT-based projected scenarios are incorporated in the hydrodynamic model for the projection of water level, flood discharge, and flooded area at different sections in the downstream of the Teesta and Lachung Rivers. Using projected flood flows and water-level data, the rating curve equations have been developed at the outlets such as Chungthang and Lachung. The study outcomes resulting from both models show a very good agreement between the simulated streamflow and observed streamflow at the outlet locations. The uncertainties can also be associated in the modeling outcomes. Thus, a statistical-cum-stochastic downscaling method has been applied to downscale the CMIP5 CM3 model-based meteorological variables (e.g., temperature and precipitation). The modeling errors prior to project streamflow have been adjusted using a sequential parameter fitting approach (SUFI2). The hydrodynamic model-based projected unsteady flood flows are found highly uncertain in the downstream portion of both rivers, and thus several extreme flood peaks have been observed in the final outcomes. The observations show that the water velocities have increased in the projected scenarios (recorded as $8.5\mathrm{m}/\mathrm{s}$), as compared to historical scenarios ($6.5\mathrm{m}/\mathrm{s}$). Using projected time series discharge and water level (stage), the rating curve equations, which have been developed as per the low to extreme emission scenarios, can be used in the future planning and management of the water resources systems.

Conceptual rainfall-runoff models (CRRMs) are widely used for flood forecasting and hydrologic simulations. However, parameter calibration poses a major challenge for using CRRMs, especially given that climate changes and effects of human activities often necessitate recalibration of CRRMs. Genetic algorithms (GAs) are one of the most widely used optimization techniques for hydrological model calibration, and have been widely and successfully used in model calibration; however, the complexity and high dimensionality of parameter calibration make them time-consuming and prone to local optima. Moreover, repetitive computation of fitness values in the GAs greatly reduces the efficiency. Therefore, new methods must be explored to improve the computational efficiency. Multicore parallel technology, which enables resource sharing and has low parallel costs and computing burdens, offers considerable benefits for parameter calibration. Thus, a multicore parallel genetic algorithm (MCPGA) based on the fork–join parallel framework with a tabu strategy is proposed in this paper for CRRM calibration. The method uses a multicore to divide the original task into several small subtasks and complete them concurrently, and adopts tabu strategy to avoid redundant computation in the GA to enhance the computing efficiency. The current methodology is applied to parameter calibration for the Xinanjiang model (which is a typical CRRM and has extensive applications in humid and semi-humid regions in China) of the Shuangpai Reservoir. Result comparisons between the MCPGA and serial GA indicate that the proposed method ensures a high degree of accuracy and significantly improves the computational efficiency.

Hydrologic models are widely used to represent and study water flows in various land uses and environments. This paper presents work done to model stormwater runoff from a highway and forested watershed in Alabama using two models: Gridded Surface/Subsurface Hydrologic Analysis (GSSHA) and Storm Water Management Model (SWMM). Of particular importance in this work is exploring similarities and differences in surface and groundwater modeling results between the two models. SWMM and GSSHA calculated similar precipitation infiltration via the Green-Ampt equation and similar volumes of surface runoff for all events shown here. Comparisons of modeled data to measured streamflow and groundwater table elevation data are made for major storm events in 2013–2014. GSSHA reproduced event streamflow volumes with total sum of absolute residuals (TSAR) error ranges from $-4.87$ to $-0.61{\mathrm{m}}^3/\mathrm{s}$ compared with SWMM flow volume TSAR error ranges from $-7.66$ to $9.29{\mathrm{m}}^3/\mathrm{s}$. Both programs performed well in calculating the groundwater table elevation change due to stream bank storage, with sum-square error ranges of 0.0074–0.0206 for GSSHA and 0.0066–0.0249 for SWMM. GSSHA represented spatially varying groundwater table elevations, whereas SWMM calculated the average groundwater table elevation in response to rain events and stream bank storage.

Application of a deterministic geometric approach for the simulation of mildly intermittent hydrologic data, exhibiting a few peaks and displaying relatively slowly rising and falling limbs and yielding slowly decaying autocorrelation functions that reach a zero value at a lag that is at least 5% of the length of the records, is presented. Specifically, adaptations of the original fractal-multifractal (FM) method and an extension, yielding more general attractors instead of fractal functions (and relying on five and eight parameters, respectively), are advanced in order to simulate (1) continuous rainfall events gathered every few seconds or minutes and lasting a few hours, and (2) continuous streamflow records measured at the daily scale and encompassing a year. It is shown, using as case studies one rainfall event in Boston, three storms gathered in Iowa City, and 4 years of streamflow records at the Sacramento River in California, all having distinct geometries, that the (computationally efficient) FM approach is capable of closely preserving either the complete record’s autocorrelation function or the data’s whole histogram (including moments), and even both, resulting in suitable rainfall and streamflow simulations, whose features and textures are similar to those of the observed data sets. The study hence establishes, for the first time, the possibility of parsimoniously simulating hydrologic sets in time in a deterministic manner, as a novel way to supplement stochastic frameworks.

Sediment concentration in open channels is fundamental to modeling sediment and pollutant transport. This study employs Renyi entropy for deriving the vertical distribution of suspended sediment concentration in open-channel flow. The derivation maximizes entropy by invoking the principle of maximum entropy, which selects the least-biased probability distribution out of many probability distributions that satisfy a given set of constraints. By considering point source release of sediment particles along with the assumption that the movement of sediment particles follow a nonlinear differential equation, the concentration distribution of suspended sediment is also investigated using a random walk hypothesis. The distribution obtained here is found to be similar to that obtained using entropy. The distribution is evaluated with experimental and field observations and good agreement is observed between computed and measured data. An error analysis is carried out to support the results and the relative root-mean-square error varies from 0.125 to 0.872 for experimental and from 0.141 to 0.510 for field data. Comparison with another entropy-based distribution shows higher accuracy of the proposed distribution.

A brief overview of a statistical method to estimate the probable maximum precipitation (PMP) is presented. This study addresses some issues associated with Hershfield’s $K_m$-value method to estimate PMP in China, which can be solved by the revised Hershfield’s $K_m$-value method. This new derivation makes it clear that the frequency factor $K_m$ is dependent on only two variables, the standardized variable, ${\mathit{\varphi }}_m$, the maximum deviation from the mean scaled by its standard deviation, and the sample size, $n$. It is found that there is a consistent relationship between $K_m$ and ${\mathit{\varphi }}_m$. Therefore, $K_m$ can be used to make a preliminary estimate of PMP under some conditions when sufficient rainfall data are available. The advantages and disadvantages of this revised $K_m$-value method are also discussed here with a case study for the estimation of 24 h PMP in Hong Kong. The 24 h PMP estimate in Hong Kong based on the local rainfall data is approximately 1,753 mm.

Flow duration curves (FDCs) are widely used to assess water availability. With the aim of overcoming the scarcity of available data in many river basins, this study proposed a process to estimate FDCs at ungauged locations. Twenty-two river basins in Taiwan with major demands were analyzed through multivariate statistical techniques. First, 158 gauges were grouped into homogeneous regions using principal component analysis (PCA), cluster analysis (CA), and discriminant analysis (DA). Second, 99 gauges with data containing unregulated flow were selected and models were built for FDC estimations, applying three regression methods including the traditional multiple linear regression (MLR), principle component regression (PCR), and partial least-squares regression (PLSR). A comparison of the model performances obtained using a cross-validation procedure showed some unstable conditions for MLR in the regional regression models for estimating the FDC; this was due to the variation of selected variables among percentile flows. The PCR and PLSR approaches can address the difficulties of variable selection and achieve a more-robust model of FDC estimation.

Daily flow regulation of a hydropower station has a greater influence on waterway transport than long-term regulation such as seasonal or annual regulation. Although a physical model can reflect the influence of daily regulation on waterways exactly, this approach is not economical for smaller hydropower stations. In this study, a mathematical modeling approach was adopted to simulate the influence of daily regulation of the Xiaonanhai (XNH) hydropower station on downstream navigation. A one-dimensional unsteady-flow dynamic model was established by using Saint-Venant partial differential equations with the Preissmann implicit difference scheme and further calibrated with observed hydrological data to ensure simulation accuracy. As a case study, the mathematical model was successfully applied to simulating the hydrodynamics of XNH under daily regulation and estimating the downstream navigability. The results showed that daily flow regulation based on the predicted power demand reduces the flow conditions downstream of the dam during the dry season, which will inevitably impede navigation. This paper presents a practical method for evaluating the navigability downstream of small and midsize hydropower stations.

This paper applies HEC-HMS to the Karkheh River basin (KRB), Iran, and facilitates the calibration of a continuous hydrologic model (CHM) with soil moisture accounting (SMA) and snowmelt degree-day parameters. Manual calibration was performed to ensure the physical relevance of HEC-HMS parameter values. Because manual calibration entails changing each parameter value in a user-defined setting, it is often a time-consuming procedure complicated by multitude of interacting parameters. To address this setback, an event-based calibration technique (EBCT) was implemented in KRB and its interior sub-basins whereby the governing parameters of specific fall, spring, and winter events were initially estimated in a precalibration step and used as inputs to facilitate calibration of the CHM. Model performance analyzed based on goodness-of-fit criteria with respect to peak flows, low flows, and hydrograph shape reflects uncertainties associated with streamflow naturalization and use of average annual parameter values for the snowmelt component. Sensitivity analysis provided insights into the basin’s snowfall and melt characteristics, distinguishing antecedent temperature index (ATI) cold rate coefficient and baseflow recession coefficient as key parameters affecting hydrograph shape and magnitude of the peak flow, respectively. Results based on goodness of fit metrics suggest that event-based parameter estimation using seasonal characteristics improved the efficiency and accuracy of the continuous HEC-HMS model (CORRL and NSE 0.78–0.87 and 0.5–0.7, respectively) while facilitating application to a large, data-poor river basin with heterogeneous climatic conditions.

The modified Universal Soil Loss Equation (USLE-M) was empirically deduced by a statistical analysis of the original data set of soil loss measurements used to derive the Universal Soil Loss Equation (USLE). The USLE-M, including the effect of runoff in the event rainfall-runoff erosivity factor, is characterized by a better capacity to predict event soil loss. At first, in this paper, using the soil erosion representative variables of USLE-M and the reference condition adopted in the USLE, the dimensional analysis and the self-similarity theory are applied to theoretically deduce a multiplicative equation similar to the USLE-M. Then using the database of the Sparacia experimental site, the ability to predict event soil loss by the USLE-M and modified USLE-M (USLE-MM) models are tested. The analysis demonstrates that the relationships used to predict the topographic factors of USLE-MM can be applied in areas different from the one of its original derivation. Finally, the analysis shows that, overall, USLE-MM [Nash-Sutcliffe index $(\mathrm{N}\mathrm{S}\mathrm{E}\mathrm{I})=0.80$; root mean square error $(\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E})=10.3$] performed slightly better than USLE-M ($\mathrm{N}\mathrm{S}\mathrm{E}\mathrm{I}=0.75$; $\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E}=11.4$) whereas the soil loss prediction accuracy especially improved for the relatively low and the highest values. In fact, for event soil loss per unit area $(A_e)<10\mathrm{Mg}{\mathrm{ha}}^{-1}$ the RMSE was equal to 7.1 for USLE-M and 4.4 for USLE-MM, and for $A_e>100\mathrm{Mg}{\mathrm{ha}}^{-1}$ the maximum factor of difference between predicted and measured soil losses was equal to 1.8 for USLE-M and 1.2 for USLE-MM.

The standardized precipitation index (SPI) is widely used in meteorological drought classification on a monthly or seasonal scale. A new monthly drought classification scheme using daily precipitation–based climate indices—consecutive two dry days (CDD2), maximum one-day rainfall (RX1), and maximum five-day rainfall (RX5)—is attempted in this study. A rule was created to derive the drought class from climate indices using northwest Indiana regional rainfall. The U.S. Drought Monitor, which uses the SPI for drought classification, was compared with the proposed new drought classification scheme. The SPIs derived using monthly data do not have a bearing on extreme precipitation indices because of lumping. This study hypothesizes that the climate indices can improve drought classification on a monthly/seasonal scale. The results are validated using the northwest Upper Mississippi and northeastern Indiana regions. The new classification and the SPI are very similar in most of the months, when the indices and monthly rainfall deviate similarly from normal values. The proposed reclassification incorporates stress to the watershed system when a major portion of the monthly rainfall is due to one or two events. It also captures deviations in the climate indices from normal values and uses them in the classification.

Droughts are extreme natural phenomena that cause severe damage to economy, environment, and life. Palmer’s drought model is widely used in drought characterization. However, its use in a mountainous snow-dominated region is often a concern because Palmer’s method does not take into account the contribution of snowmelt. Snowmelt is an important contributor to renewable water resources in many parts of the world, including Afghanistan. In this study, the standard Palmer method is modified to incorporate snow using a snowmelt module based on the temperature-index method. The Palmer model with and without snowmelt is used in this study to characterize drought, quantify drought severity, and determine the spatial extent of historical droughts in Afghanistan. Reanalysis data for precipitation and temperature over a period of 45 years (1957–2002) are used in the analysis. Results are analyzed using three drought indices—Palmer drought severity index (PDSI); Palmer hydrological drought index (PHDI); and Palmer moisture anomaly index (Z-index). Both the standard and modified Palmer models captured the documented major national droughts, including the severe events in 1971 and 2001. The model with snowmelt is found to represent the historical droughts as less severe compared with the standard Palmer method. In addition, the Panj and Kokcha subbasins, which are located at a higher elevation, showed less severe drought compared with low-lying regions such as Shoretepa and Balkaab during the severe event in 2000–2001. Moreover, larger differences between the standard and modified Palmer’s drought models are observed at higher elevations. The authors further observed that the occurrence of meteorological drought does not necessarily mean that there exist agricultural and hydrological droughts. The modified method has the capability to be used in monitoring droughts in snow-dominated regions and thus can assist decision makers in providing suitable mitigation measures.

This paper describes a new application of the Fourier series for a detailed simulation of runoff in a small catchment in dry periods, when the streamflow is significantly impacted by evapotranspiration, particularly during daytime hours. The catchment was considered as a dynamic system, in which evapotranspiration has an impact on the day–night fluctuation of discharges. Measurements of these discharges were accomplished by a high-resolution water-level sensor attached to a V-notch. In parallel, the free water evaporation and also the soil moisture content have been measured nearby. The paper describes three runoff recession episodes in dry periods. Using short time step measurements and calculations, it was not difficult to analyze the diurnal streamflow fluctuations as harmonic waves. An application of the finite Fourier series model (FSM) to a quasi-periodic hydrologic data series clearly shows how the actual evapotranspiration influences surface runoff from small catchments. The method was verified by direct numerical evaluation of the convolution integral. The Fourier transformation works better if the number of discharge points ($n$) is large. The method allows computing the missing discharges in order to bridge accidental data gaps. The automatic measurement of free water evaporation multiplied by the measured soil water content compared semiquantitatively with the Fourier transformation function derived backward from the discharge hydrograph. Hence, this measurement can, to some extent, substitute the actual evapotranspiration on the catchment scale. The observation has been carried out with a time delay of the stream discharges behind the water evaporation, caused by the unsaturated zone processes and the hydraulic resistance to water flow, both in the saturated zone and in the streambed.

This study investigates the effect of the merging of subbasins on the storage coefficient of the entire basin with the concept of parallel and serial systems of linear reservoirs having different storage coefficients. Based on the linear reservoir theory, both the parallel and serial connections of two subbasins are analyzed and the storage coefficient is derived from the impulse response function of the merged subbasins. The storage coefficient of the entire basin could also be derived by merging subbasins one by one from upstream to downstream. The applicability of the proposed methodology is tested considering the subbasin arrangement of the Chungju Dam Basin in Korea. The findings are as follows: (1) the storage coefficient of a basin with parallel subbasins is represented as the weighted average by the subbasin areas, and that of a basin with serial subbasins is represented simply by the sum of storage coefficients; and (2) the storage coefficient of the Chungju Dam Basin is estimated to be approximately 30.53 h by considering the structure of 44 subbasins, i.e., the combination of parallel and serial subbasins.

Although numerous studies conducted over the past four decades have shown the significant role that drainage network geometry plays in determining the streamflow response and hence the peak flood magnitude, a review of regional flood frequency equations established for various hydrologic regions in the United States shows that drainage network geometry is largely excluded from regional regression equations. This study demonstrates how river network geometry affects flood frequency using two catchments located in the same hydrologic region and have the same drainage area, but with different drainage network geometry. The results from the analysis of historical peak discharge data obtained from streamflow gauging stations located at the outlet of the watersheds and the systematic numerical simulation of the watersheds show that the apparent difference in the drainage network geometry of the two watersheds accounts for the observed difference in the flood frequency. By showing the significant effects of drainage network geometry on flood frequency, the results highlight the need for a revision of the current approach to regional flood frequency equations to improve the accuracy of flood quantile estimation in ungauged basins.