Journal of Hydrologic Engineering

Accurate and reliable prediction of runoff generation is necessary for flood control scheduling, water supply planning, and hydropower generation. Support vector machine (SVM), which is at the forefront of current research of regression and classification, was used in this paper to conduct monthly mean streamflow prediction. A novel heuristic optimization named bat algorithm (BA) was introduced to determine the parameters of SVM [penalty parameter ($C$) and kernel parameter ($\eta$)], in which the initial fitness was supposed to be equal to the initial loudness for all bats. In order to evaluate the effectiveness of the proposed approach, monthly mean streamflow from 1952 to 2011 of Yichang station in the middle reaches of the Yangtze River were trained and tested. In the meantime, the given data set was also modeled using artificial neural networks (ANN) and cross validation–based SVM. The comparison results indicate that the proposed model (bat algorithm–based SVM) is more accurate compared with both ANN and cross validation–based SVM. However, two main shortages exist, i.e., time-consuming and relatively low accuracy in the break points of continued dry (wet) years. To relieve these shortages, local optimization algorithms [e.g., differential evolution (DE) algorithm, immune algorithm (IA), and genetic algorithm (GA)] were suggested to be combined with the bat algorithm to produce the initial population. Modifications of the stochastic term of the local search were also useful.

The ability to design and employ groundwater distribution models plays an important role in the development and application of regional water management policies and resource exploration. This paper presents a probabilistic reasoning approach for estimating groundwater levels over a geological map based on a limited number of available observations of hydraulic head and conductivity levels. The approach adapts, expands, and combines non-Euclidean distance kriging, probabilistic graphical modeling, and expectation maximization to provide a viable alternative to the currently existing, simulation-based methods of spatial interpolation. Upon outlining a conceptual framework for the proposed approach, this paper investigates the feasibility of using its key component, the Markov random field, with a flexible (learned) structure that recovers hydraulic conductivity maps from the knowledge of hydraulic head on those maps. The model is trained on a medium-sized data set of simulated hydraulic maps, and returns promising results. The paper also motivates future work in the area, pointing out several research directions.

Estimates of the USDA-NRCS runoff curve number (CN) are generally based on a soil map and observed land cover. Because the CN method is increasingly applied to disturbed and urbanized land, the objective of this work was to collect effective saturated hydraulic conductivities, $K_{\mathrm{sat}}$, and sorptivities, $S_o$, from a range of land use types, use the results to estimate a CN, and compare these CNs with CN estimates made from soil survey information and corresponding land cover. A total of 331 double ring infiltration tests were conducted over the 15 sites. Based on land use and site history, the test sites were classified into categories of engineered, urban altered, rural altered, rural unaltered, and prairie. Measured $K_{\mathrm{sat}}$ values were skewed so the medians of these data were a better predictor of central tendency. The prairie and rural unaltered median $K_{\mathrm{sat}}$ values were closer to soil map estimates than the other categories (between 0.0 and 91% different from soil survey). Two empirical methods developed in Hawaii using sprinkle infiltration tests were used to estimate CN values from infiltration data; method 1 used only $K_{\mathrm{sat}}$ as the predictor and method 2 used $K_{\mathrm{sat}}$ and $S_o$. Results from these two methods were not statistically different at 12 of the 15 sites ($\alpha =0.05$). When comparing these methods to CN values developed from soils data and land cover (method 3), better overall agreement existed between method 1 and method 3. The median CN value from method 1 was the best predictor for the mean CN based on measured runoff data for an urban altered land use site (0.6% different), whereas method 3 was the best predictor for the mean CN based on measured runoff data for a prairie land use (0.0% different).

Weather data are usually required at several locations over a large watershed, especially when using distributed models for hydrological simulations. In many applications, spatially correlated weather data can be provided by a multisite stochastic weather generator, which considers the spatial correlation of weather variables. Prior to using a multisite weather generator for hydrological modeling, its ability to adequately represent the proper hydrological response needs to be assessed. This study assesses the effectiveness of a new multisite weather generator (MulGETS) for hydrological modeling over a Canadian watershed in the Province of Québec. Prior to hydrological modeling, MulGETS is first evaluated with respect to reproducing the spatial correlation and statistical characteristics of precipitation and temperature for the studied watershed. Hydrological simulations obtained from MulGETS-generated precipitation and temperature are then compared with those obtained from a single-site weather generator (WeaGETS) and a WeaGETS-based lumped approach (WeaGETS-lumped) that averages the climate series over all stations in a watershed before running the single-site weather generator. The hydrology is simulated using two hydrological models: the conceptually lumped model HSAMI and the physically based distributed model CEQUEAU. When using the conceptually lumped model, the weather time series is first averaged over all stations in the watershed. The results show that the monthly mean discharge is accurately represented by both MulGETS-generated and WeaGETS-lumped-generated precipitation and temperature, whereas it is considerably underestimated by WeaGETS data for the snowmelt period. The MulGETS and WeaGETS-lumped data also show significant advantages in representing the monthly streamflow variability, which is underestimated by the WeaGETS outputs. Additionally, MulGETS and WeaGETS-lumped consistently perform better than WeaGETS for simulating extreme flows (snowmelt high flows and summer-autumn high and low flows). However, no obvious difference in performance was found between MulGETS and WeaGETS-lumped data for hydrological modeling. Moreover, the use of a physically based distributed model with MulGETS did not result in any significant performance gain compared with the much simpler combination of WeaGETS-lumped with a lumped hydrological model for the studied watershed. Overall, this study indicates that a single-site weather generator combined with a lumped hydrological model is sufficient for accurate hydrological simulations, even in the case of a large watershed.

Identification of changes in hydrological extremes plays an important role in water resources management. This study investigates various properties in hydrological extremes including annual maximum daily flow (high flow) and annual minimum 7-day flow (low flow) in 252 unimpaired streamflow gauges in the continental United States. Three statistical methodologies are performed: the nonparametric Mann-Kendall test is used to evaluate temporal trends; the non-parametric change point Pettitt test, to identify abrupt shifts; and the quantile perturbation method, to assess temporal variability. In the results of trend analysis, low flow exhibited some spatial patterns compared to high flow. In addition, low flow has increased over 64 years (1950–2013) in many stations. The Pettitt test indicates that low flows have experienced more significant abrupt changes compared to high flow. The results of the quantile perturbation method confirm that many significant increasing trends obtained during the 64 years of low flow are attributable to infrequent low flow events in the period before 1970. These findings highlight that various statistical approaches complement each other and must be simultaneously applied to hydrological time series.

Rainfall data provide an essential input for most hydrologic analyses and designs for effective management of water resource systems. However, in practice, missing values often occur in rainfall data that can ultimately influence the results of hydrologic analysis and design. Conventionally, stochastic interpolation methods such as kriging are the most frequently used approach to estimate the missing rainfall values where the variogram model that represents spatial correlations among data points plays a vital role and significantly impacts the performance of the methods. In the past, the standard variogram models in ordinary kriging were replaced with the universal function approximator-based variogram models, such as artificial neural networks (ANN). In the current study, applicability of genetic programming (GP) to derive the variogram model and use of this GP-derived variogram model within ordinary kriging for spatial interpolation was investigated. Developed genetic programming-based ordinary kriging (GPOK) was then applied for estimating the missing rainfall data at a rain gauge station using the historical rainfall data from 19 rain gauge stations in the Middle Yarra River catchment of Victoria, Australia. The results indicated that the proposed GPOK method outperformed the traditional ordinary kriging as well as the ANN-based ordinary kriging method for spatial interpolation of rainfall. Moreover, the GP-derived variogram model is shown to have advantages over the standard and ANN-derived variogram models. Therefore, the GP-derived variogram model seems to be a potential alternative to variogram models applied in the past and the proposed GPOK method is recommended as a viable option for spatial interpolation.

The impact of prospective climate change on the hydrological and ecological status of the mountain stretch of the Serio river (ca. $300{\mathrm{km}}^2$) in the Northern Italian Alps was investigated. A hydrological model was used to mimic theflow regime, and experimental suitability curves were used to assess weighted usable area (WUA) for brown trout (Salmo trutta) in different stages (adult, young, and spawning), and four macroinvertebrates families (Leuctridae, Heptageniidae, Limnephilidae, and Limoniidae). Discharge-WUA curves were obtained using instream flow incremental methodology/physical habitat simulation system (IFIM-PHABSIM), and a seasonal WUA assessment was carried out. The future (until 2100) hydrological cycle was projected using outputs from two general circulation models from Assessment Report 5 of the Intergovernmental Panel on Climate Change (IPCC). The results display a potential for a large flow decrease yearly, until $-60\%$ at 2050 and $-56\%$ at 2090. Spring melt peaks will be largely dampened, with the largest projected decrease of $-60\%$ in 2050 and $-66\%$ in 2090. Also fall flows will be largely smoothed, down to $-70\%$ for 2050 and $-56\%$ for 2090. To quantify habitat quality, a specific quantile ${\mathrm{WUA}}_{20\%}$, was used and exceeded presently for 292 days in a year, and the percentage of critical days $n_{c\%}$ when $\mathrm{WUA}<{\mathrm{WUA}}_{20\%}$. Adult trouts will have decreasing $n_{c\%}$ yearly (from 20% now to down to 13% in 2090), increasing in winter and spring and decreasing in fall. Young trouts will have decreasing $n_{c\%}$ yearly (from 20% now to down to 2% in 2040 and 2090), but increasing in winter. Spawning, occurring in winter, will have lower $n_{c\%}$ (from 16% now, to down to 8% in 2090). Macroinvertebrate families will have constant $n_{c\%}$ yearly, unless for Leuctridae (from 20% now to up to 38% in 2050), and noticeable decrease of $n_{c\%}$ in winter for Leuctridae until 2090, increase in spring and summer, and decrease in fall for all families until 2090. Correlation analysis of the WUA against weather and hydrological variables displayed that temperature is the most influencing variable, decreasing habitat suitability for all species in fall, and increasing suitability in spring, unless for young trouts.

Substantial energy losses occur when water flows over vegetated ground. Vegetation enhances surface resistance and decreases flow velocity, which results in energy loss. Since the ground is considered to be permeable, both surface and subsurface water flow are simultaneously resolved. The objectives of this study are to analyze the dynamic effects of vegetated flows including the distributions of velocity, shear stress, and energy by an analytical approach for better understanding the physical mechanism. An inflection point in the velocity distribution is found as that in conventional research for denser vegetation cases with a constant water depth. A new dimensionless parameter, ${\text{δ}}_2$, related to permeability of porous medium and clearly affecting the shear stress distribution is proposed. The value of ${\text{δ}}_2$ approximately equals 9.5 as the ratio of the water layer depth to the vegetation layer height is unity, then the shear stresses at the interface of water/vegetation layers and at the ground surface appear equivalent and reach a maximum. The results also reveal that most of the energy loss occurs near the interface of water/vegetation layers.

Derivation of spatial distribution of travel time is required in some conceptual rainfall-runoff transformation models. Application of a kinematic wave (KW) solution to delineate isochrones of travel time may be straightforward, particularly in an overland flow regime. In this article, a new analytical approach based on the diffusive wave (DW) solution is developed for calculation of travel time. The discharge hydrograph over a rectangular plane will be calculated based on the developed analytical DW method through the application of the time-area method. Differences between calculated maximum travel time (i.e., equilibrium time), isochrone locations, and outflow discharge between the KW and DW solutions are compared with those of the numerical dynamic wave (DYW) solution. It is shown that analytical DW results are closer to those of the DYW in comparison with those of the KW. The effect of the $K_0{F}_0^2$ number on the solution accuracy is investigated, where $K_0$ is the kinematic wave number and $F_0$ is the Froude number. It is perceived that as $K_0{F}_0^2$ decreases, the differences between KW and analytical DW solutions grow. For instance, at equilibrium time, the new DW solution improved the results by 5 to 12% over the KW solution. In addition, the DW isochrones lie upstream of the corresponding KW isochrones. Finally, the $1.5<K_0{F}_0^2<20$ interval is proposed for using an analytical DW solution in lieu of the KW and DYW solution.

The hydrologic behavior of an underdrained porous pavement system was characterized using an effective curve number (ECN), which is synonymous to the common runoff curve number (RCN). A model previously developed for porous pavements without an underdrain, which generates the rainfall-runoff data for the system, was modified to include an underdrain. However, the addition of an underdrain adds three new parameters (drain size, drain placement height, and drain spacing) that both influence the ECN and fundamentally change the design procedure. A sensitivity analysis was completed to quantify the impacts of different rainfall depths and distributions on this modified model. The RCN equations were then fit to the rainfall-runoff data to find the ECN value. ECN values for a wide range of pavement and underdrain configurations were summarized in figures that can be used as preliminary design aids for designers and regulators. These figures can be used to estimate the ECN value for a pavement design using only the basic pavement system, soil, and drain properties. A new design procedure was also proposed that uses these figures as a tool to estimate the required size and location of the underdrains for a pavement system to have a specified design ECN.

Mantiqueira Range is the most important headwater region in southeastern Brazil, being responsible for streamflows that feed a significant part of Brazilian hydroelectric energy production. This region is extremely fragile with endemic species like those that compose the upper mountain cloud forest (UMCF), and the dominant soils are typical Inceptisols. These forest environments are rapidly disappearing in Brazil. The changing land use combined with shallow Inceptisols lead to important effects on hydrology. Therefore, studies addressing streamflow behavior in these landscapes are urgently needed for understanding the groundwater recharge process. The objectives of this study were to investigate the role of shallow Inceptisols linked to the land use changes in the streamflow behavior of the Mantiqueira Range region and to understand the mechanisms of water flow into the saturated zone. To validate the study, a comparative analysis was developed for a representative watershed of Mantiqueira Range, whose land use is divided into pasture and Atlantic Forest, known as Lavrinha Creek Watershed (LCW) and a typical UMCF located inside of LCW. The results showed that in general, UMCF has greater potential for water percolation beginning from the surface layer towards the saturated zone. Micromorphological soil images provided useful information about the soil pore system associated with the physical properties of Inceptisols, and this background helped to understand the water recharge process and streamflow behavior. In addition, the water balance for two years (2009–2011) in the UMCF showed that the flow stored in the saturated zone occurs rapidly, increasing its potential for storage, characterized by the process of refilling water in the Inceptisols soil profile under forests in this region.

This research presents a spatial clustering of mean air temperature in Iran based on monthly data of 44 synoptic and climatologic stations from 1951 to 2010 via multivariate and L-moment methods. After standardizing the data, two multivariate methods of factor analysis (FA) and cluster analysis (CA) were used to derive primary homogeneous regions. Then, two L-moment based indices were applied to examine the homogeneity of the derived clusters along with the discordancy of each cluster’s member(s). Next, considering geographical similarities of different parts of Iran as well as local weather regimes, discordant stations were moved to other clusters. The two latter steps were repeated (a trial-and-error process) to derive final homogeneous clusters. The optimum status, in terms of air temperature, resulted in a grouping of eight homogeneous regions in Iran for which there were reasonable relationships between clusters and the natural geography of the country. Considering the results, in most of the regions, four months of June through September showed the highest mean air temperature in Iran; while December, January, and February indicated the lowest ones.

A novel multisite cascading calibration (MSCC) approach using the shuffled complex evolution–University of Arizona (SCE-UA) optimization method, developed at the University of Arizona, was employed to calibrate the variable infiltration capacity (VIC) model in the Red River Basin. Model simulations were conducted at 35 nested gauging stations. Compared with simulated results using a priori parameters, single-site calibration can improve VIC model performance at specific calibration sites; however, improvement is still limited in upstream locations. The newly developed MSCC approach overcomes this limitation. Simulations using MSCC not only utilize all of the available streamflow observations but also better represent spatial heterogeneities in the model parameters. Results indicate that MSCC largely improves model performance by decreasing the number of stations with negative Nash-Sutcliffe coefficient of efficiency (NSCE) values from 69% (66%) for a priori parameters to 37% (34%) for single-site calibration to 3% (3%) for MSCC, and by increasing the number of stations with NSCE values larger than 0.5 from 9% (9%), to 23% (23%) to 34% (29%) during calibration (and validation) periods across all sites.

The Al Batinah coastal plain in the Sultanate of Oman is one of the area’s most treasured regions for the existence of fresh groundwater and agriculture. Growth of irrigation, urbanization, and industrial endeavors have resulted in a severe decline in groundwater levels in the Al Batinah coastal aquifer, leading to significant seawater intrusion in recent decades. This study examines the groundwater and seawater intrusion pattern in the eastern Al Batinah coastal plain using a groundwater flow model along with a solute transport model. Both steady and transient calibration was carried out using data from the Ministry of Regional Municipalities and Water Resources. The calibrated model was then utilized to simulate two scenarios to mitigate seawater intrusion: (1) reduction in pumping rate; and (2) artificial recharge with constant pumping. Model predictions suggest that the effect of optimizing the pumping rate and artificial recharge reducing seawater intrusion would prove to be effective in the eastern Al Batinah coastal plain.

Despite the documented evidence of geomorphic characteristics on flooding, an understanding of the relative effects of the geomorphic characteristics on flooding remains elusive. The objective of this study is to investigate the relationship between flooding and geomorphic characteristics in Indiana, United States, by using data at 94 streamflow gauging stations. The flood magnitude as determined through the flash flood index is categorized into three groups, including moderate, extreme, and severe. The flood in each group is then related to geomorphic characteristics, including topography, morphometry, slope, land use, soil, channel network, and aspect through stepwise regression. Results show that extreme flooding is most affected by watershed morphometry, particularly watershed length, whereas severe flooding is most affected by watershed slope and land use type. The methodology used in this study also highlights that the stronger the severity of flooding, the more it is explained by geomorphic characteristics. Additional investigation of the effect of land cover type and watershed size using the Kruskal-Wallis rank test shows that land cover type affects flooding occurrence despite its relatively role is relatively smaller compared with other variables. Finally, flooding is more affected by watershed length than watershed area for the watersheds included in this study.

Unanticipated peak inflows that can exceed the inflow design flood (IDF) for spillways and result in possible storage loss in reservoirs from increased sedimentation rates lead to a greater risk for downstream floods. Probable maximum precipitation (PMP) and probable maximum flood (PMF) are mostly used to determine IDF. Any possible change of PMP and PMF resulting from future land use and land cover (LULC) change therefore requires a methodical investigation. However, the consequential sediment yield resulting from altered precipitation and flow patterns into the reservoir has not been addressed in literature. Thus, this study aims to determine the combined impact of a modified PMP on PMF and sediment yield for an artificial reservoir. The Owyhee Dam of the Owyhee River watershed (ORW) in Oregon is selected as a case study area for understanding the impact of LULC change on PMF and sedimentation rates. Variable infiltration capacity (VIC) is used for simulating streamflow (PMF) and the revised universal soil loss equation (RUSLE) to estimate sediment yield over ORW as a result of change in precipitation intensity and LULC. Scenarios that represent pre-Owyhee Dam (pre-dam) and post-Owyhee Dam (post-dam; nonirrigation, control) are used to simulate PMF’s and consequential sediment yield. Peak PMF result for pre-dam scenarios increased by 26 (1%) and $81{\mathrm{m}}^3{\mathrm{s}}^{-1}$ (3%) from the nonirrigation and control scenario, respectively. Considering only LULC change, sediment yield decreased over ORW owing to the transformation of LULC from grassland to shrubland (from the pre-dam period to the post-dam years). However, increase in precipitation intensity caused a significant (0.1% storage loss over a 21-day storm period) increase in sediment yield primarily resulting from reservoir sedimentation. This study underscores the need to consider the future impact of LULC change on IDF calculation and sedimentation rates for more robust reservoir operations and planning.

Hydrologic design and planning requires dealing with extremes in precipitation and temperature events caused due to climate change. Long-term trends and variability in 20 extreme climate indices (ECIs) and the influence of El Niño and La Niña over the Pranhita catchment, India are studied for the period of 1951–2013. An overall increase in daily precipitation, more intense rain, and an increase in continuous dry days are observed; however, total rain shows a negative trend. A warming trend is observed with an increase in dry days, warm days, and hot nights, and a decrease in the frequency of cool days and nights. Observed $+\mathrm{ve}$ trends for maximum and minimum temperatures are 0.7 and 0.8°C, respectively. El Niño years resulted in $-\mathrm{ve}$ anomaly and $+\mathrm{ve}$ anomaly was observed during La Niña years for the majority of ECIs. The association between ENSO and yearly temperature indices is found to be weak in comparison with precipitation indices. The study confirmed the multimodal nature of precipitation during El Niño, high precipitation during La Niña, and excess postmonsoon rains during La Niña. A comparison between the Oceanic Nino Index (ONI) and the precipitation anomaly showed that the anomaly changes with a phase change in ENSO with the relation being very prominent during El Niño. The El Niño years have resulted in a precipitation deficit, and the La Niña years have resulted in excess precipitation.

With recent acknowledgment that New Zealand requires updated guidelines in predicting rainfall and runoff, this case study compares three approaches in forming flood regions for the North Island of New Zealand. With comparison to regions in the literature, new regions are delineated on the basis of administrative boundaries, similar catchment climate and dominant topographies, and comparable linear moment ratios of the coefficient of variation (L-CV) of the annual maximum flow series. Across 204 individual flow gauging sites, the regions on the basis of L-CV were observed to provide the best predictors of flood discharges when compared with those derived using at-site analyses, with relative errors of 6.9 and 21% for 10- and 100-year peak discharges, respectively. Although these regions are delineated on the basis of spatial proximity and similar hydrologic flood flow regimes, concerns are raised regarding the homogeneity of these regions as a whole. Although further breakdowns of regions could be beneficial, these regions are still recommended for use in ungauged catchments over those previously published.

Climate change will have a significant impact on the hydrological cycle. This paper presents the results of a pilot project for the Yamaska River in Québec. The objective of this project is to evaluate the river’s vulnerability to low flows attributable to climate change and to analyze a no-regrets adaptation strategy at locations identified as vulnerable. The vulnerability was evaluated using statistical indicators (low flow indices) based on long-term observations at four locations in the basin. A distributed physically-based hydrological model in use in Québec was calibrated and validated against observed streamflow data to properly represent low flows. Hydrological simulations used seven climate projections provided by the north american regional climate change assessment program (NARCCAP) s project. Also, five members of the canadian regional climate model (CRCM), nested with the coupled global climate model (CGCM) under the special report on emission scenarios (SRES) A2 emission scenario, were run for a reference (1971–2000) period and a future (2041–2070) period. Streamflow simulations indicate a degradation of future low flow conditions, particularly in June and August, when compared to the reference period. In addition, the 7-day low flow value with a 2-year return period (7Q2) and the 7-day low flow value with a 10-year return period (7Q10) decrease by 16–64% and 18–45% respectively. A no-regrets adaptation strategy allowing stakeholders to reduce withdrawal according to alert levels was implemented. Simulations of the application of the no-regrets adaptation strategy reduced the number of days where streamflows are below the Crisis level in the future period by at least 20%.

HYDRUS (2D/3D) model was used to simulate spatial and temporal distributions of nitrate-nitrogen (${\mathrm{NO}}_3\text{-}\mathrm{N}$) within and below the onion root zone under conventional furrow fertigation with the urea-ammonium-nitrate (UAN) liquid fertilizer. The simulated water contents in the furrow irrigated onion field agreed well with the measurements. Simulations produced similar patterns of the measured ${\mathrm{NO}}_3\text{-}\mathrm{N}$ concentration profiles throughout the growing season. ${\mathrm{NO}}_3\text{-}\mathrm{N}$ concentrations remained higher and accumulation of ${\mathrm{NO}}_3\text{-}\mathrm{N}$ was observed within the root zone. Higher ${\mathrm{NO}}_3\text{-}\mathrm{N}$ within the root zone was dependent on the rate of the UAN fertilizer application, quantity of ${\mathrm{NO}}_3\text{-}\mathrm{N}$ removed by root uptake, and ${\mathrm{NO}}_3\text{-}\mathrm{N}$ drainage fluxes below the root zone. Simulations also suggested that ${\mathrm{NO}}_3\text{-}\mathrm{N}$ below the root zone during different growth stages remained much higher than a recommended (for drinking water) standard concentration level ($10\mathrm{mg}{\mathrm{L}}^{-1}$). This resulted in higher ${\mathrm{NO}}_3\text{-}\mathrm{N}$ drainage fluxes, particularly during the fertigation events between the establishment and vegetative growth stages. This indicates the need to apply most fertigation events at an early stage of bulb formation to provide the maximum ${\mathrm{NO}}_3\text{-}\mathrm{N}$ demands by onions and to reduce potential ${\mathrm{NO}}_3\text{-}\mathrm{N}$ leaching.

In the present work, the flood hazard exposure in an ungauged basin in Africa is assessed exploiting the basin morphological characteristics. Flood-prone areas are identified using linear binary classifiers based on several geomorphic descriptors extracted from digital elevation models (DEMs). The classifiers are calibrated individually and evaluated by comparing their outputs with a flood inundation map obtained by two-dimensional (2D) hydraulic simulations and using receiver operating characteristics (ROC) curves as performance measures. The best-performing descriptors for the subcatchment of the Bulbula River, near the city of Addis Ababa (Ethiopia), are the elevation difference, $H$ between the location under exam and the nearest drainage network, and the composite index $\ln [h_r/H]$, that compares an estimate of the water level in the nearest point of the river network to the difference in elevation between the point under exam and the river. These simple procedures allow extending the flood delineation derived with the hydraulic model over the entire river basin. The study highlights the potential for the detection of flood-prone areas over ungauged basins and large areas.

Calibration of distributed parameter groundwater models in the Bayesian framework using Markov-chain Monte Carlo (MCMC) random sampling is often hampered by the large number of simulations required to make reliable uncertainty estimates. In particular, naive application of the ubiquitous random walk metropolis Hastings (MH) algorithm can take an unsatisfactorily long time to draw samples from the posterior distribution and hence make the required uncertainty estimates. This note addresses the issue of obtaining feasible uncertainty estimates using accelerated MCMC. A pragmatic approach is investigated, based on the adaptive delayed acceptance MH algorithms of Cui et al. First, adoption of an appropriate prior model over the parameters indicates that the number of estimated parameters can be reduced from a over a thousand parameters to several tens without essential loss of information. Secondly, the algorithm is initialized by a least squares [maximum a posteriori (MAP)] estimate and the covariance of the parameters approximated by the Hessian of the objective function, which is then taken to be an initial proposal distribution for the MH algorithm. The method is evaluated with a numerical simulation, in which the calibration time is reduced five fold compared with previous results of the authors.

Estimation of seepage losses is important in the management of an irrigation scheme because such losses significantly affect the efficiency of water usage. An attempt was made to determine seepage losses as a function of soil texture. Losses were measured three times, with the inflow-outflow method, for each of the 30 channels made of sandy loam, clay, clay loam, silt clay, and silt loam (six channels for each soil texture), and the soil-specific losses were determined. The soil-specific values were validated in terms of computing outflows. The maximum seepage of $294\mathrm{mm}{\mathrm{day}}^{-1}$ was found for sandy loam, followed by 200, 185, 144, and $96\mathrm{mm}{\mathrm{day}}^{-1}$ for silt loam, silty clay, clay loam, and clay material channels, respectively. Two empirical equations requiring soil type constants were also evaluated—namely, the Moritz formula and the Molesworth and Yennidunia formula. Investigation using these two equations for seepage estimation showed either underestimation or overestimation for the soil textures investigated. Therefore, these equations should be used after calibration, in accordance with channel conditions.

Karst aquifers differ from other types of hydrogeological systems because of their complex behavior, which originates from strong heterogeneity. A karst spring carries an imprint of hydrologic information for the karst aquifer. The shape of the outflow hydrograph recorded at a spring is a unique reflection of the aquifer’s response. A karstic aquifer consisting of a fracture network and a conduit was proposed, and the spring recession curves generated from the designed karst aquifer were analyzed in this study. The purpose of the study was to discuss the influence of the main spring conduit diameter and the saturated thickness of the aquifer on spring recession curves. A combination of 7 groups of outlet pipes with different diameters and 7 initial water levels were used to simulate changes in diameter and saturated thickness, respectively. Thus, a total of 49 experimental tests were carried out in the laboratory karst aquifer. The results indicate that spring recession curves can be separated into two flow segments, and that the main spring conduit diameter has a remarkable influence on the initial flow of recession, recession duration, and recession coefficient. The results also show that the saturated thickness of the aquifer has a great effect on the initial flow of the recession and recession duration, but that its effect on the recession coefficient and the shape of the recession curve is slight.

A numerical model of variably saturated axis-symmetric radial flow is employed to simulate and parameterize the gravity-delayed drainage, viewing it as the vertical accretion over falling water table. Numerical experiments are conducted to simulate the head fields at advancing discrete times as pumping is sustained at the assigned well. The head fields are invoked to generate the corresponding time series of water table elevation and the accretion rate over it. Employing these time series, the optimal structure/parameters of gravity-delayed drainage equation are estimated by the least-squares approach under a variety of hydrogeologic/geometric conditions. It is revealed that generally, a two-parameter gravity-delayed drainage equation reproduces parsimoniously the accretion time series.