USGS Science⌕ Search

SEARCH · USGS Science

Results for “Water Research”

Search indexed USGS publications on groundwater, aquifers, geologic maps, mineral resources and earthquakes. Explore source records by subject and place.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 973 records · Page 54Linked to original sources
USGS2011

Evaluating the potential for remote bathymetric mapping of a turbid, sand-bed river: 1. Field spectroscopy and radiative transfer modeling

Remote sensing offers an efficient means of mapping bathymetry in river systems, but this approach has been applied primarily to clear-flowing, gravel bed streams. This study used field spectroscopy and radiative transfer modeling to assess the feasibility of spectrally based depth retrieval in a sand-bed river with a higher suspended sediment concentration (SSC) and greater water turbidity. Attenuation of light within the water column was characterized by measuring the amount of downwelling radiant energy at different depths and calculating a diffuse attenuation coefficient, K d . Attenuation was strongest in blue and near-infrared bands due to scattering by suspended sediment and absorption by water, respectively. Even for red wavelengths with the lowest values of K d , only a small fraction of the incident light propagated to the bed, restricting the range of depths amenable to remote sensing. Spectra recorded above the water surface were used to establish a strong, linear relationship ( R 2 = 0.949) between flow depth and a simple band ratio; even under moderately turbid conditions, depth remained the primary control on reflectance. Constraints on depth retrieval were examined via numerical modeling of radiative transfer within the atmosphere and water column. SSC and sensor radiometric resolution limited both the maximum detectable depth and the precision of image-derived depth estimates. Thus, although field spectra indicated that the bathymetry of turbid channels could be remotely mapped, model results implied that depth retrieval in sediment-laden rivers would be limited to shallow depths (on the order of 0.5 m) and subject to a significant degree of uncertainty.

Water Resources Research↗
USGS2011

Evaluating the potential for remote bathymetric mapping of a turbid, sand-bed river: 2. application to hyperspectral image data from the Platte River

This study examined the possibility of mapping depth from optical image data in turbid, sediment-laden channels. Analysis of hyperspectral images from the Platte River indicated that depth retrieval in these environments is feasible, but might not be highly accurate. Four methods of calibrating image-derived depth estimates were evaluated. The first involved extracting image spectra at survey point locations throughout the reach. These paired observations of depth and reflectance were subjected to optimal band ratio analysis (OBRA) to relate ( R 2 = 0.596) a spectrally based quantity to flow depth. Two other methods were based on OBRA of data from individual cross sections. A fourth strategy used ground-based reflectance measurements to derive an OBRA relation ( R 2 = 0.944) that was then applied to the image. Depth retrieval accuracy was assessed by visually inspecting cross sections and calculating various error metrics. Calibration via field spectroscopy resulted in a shallow bias but provided relative accuracies similar to image-based methods. Reach-aggregated OBRA was marginally superior to calibrations based on individual cross sections, and depth retrieval accuracy varied considerably along each reach. Errors were lower and observed versus predicted regression R 2 values higher for a relatively simple, deeper site than a shallower, braided reach; errors were 1/3 and 1/2 the mean depth for the two reaches. Bathymetric maps were coherent and hydraulically reasonable, however, and might be more reliable than implied by numerical metrics. As an example application, linear discriminant analysis was used to produce a series of depth threshold maps for characterizing shallow-water habitat for roosting cranes.

Platte River↗
USGS2011

Independent effects of temperature and precipitation on modeled runoff in the conterminous United States

A water‐balance model is used to simulate time series of water‐year runoff for 4 km × 4 km grid cells for the conterminous United States during the 1900–2008 period. Model outputs are used to examine the separate effects of precipitation and temperature on runoff variability. Overall, water‐year runoff has increased in the conterminous United States and precipitation has accounted for almost all of the variability in water‐year runoff during the past century. In contrast, temperature effects on runoff have been small for most locations in the United States even during periods when temperatures for most of the United States increased significantly.

Water Resources Research↗
USGS2012

Quantifying riverine surface currents from time sequences of thermal infrared imagery

River surface currents are quantified from thermal and visible band imagery using two methods. One method utilizes time stacks of pixel intensity to estimate the streamwise velocity at multiple locations. The other method uses particle image velocimetry to solve for optimal two-dimensional pixel displacements between successive frames. Field validation was carried out on the Wolf River, a small coastal plain river near Landon, Mississippi, United States, on 26-27 May 2010 by collecting imagery in association with in situ velocities sampled using electromagnetic current meters deployed 0.1 m below the river surface. Comparisons are made between mean in situ velocities and image-derived velocities from 23 thermal and 6 visible-band image sequences (5 min length) during daylight and darkness conditions. The thermal signal was a small apparent temperature contrast induced by turbulent mixing of a thin layer of cooler water near the river surface with underlying warmer water. The visible-band signal was foam on the water surface. For thermal imagery, streamwise velocities derived from the pixel time stack and particle image velocimetry technique were generally highly correlated to mean streamwise current meter velocities during darkness (r 2 typically greater than 0.9) and early morning daylight (r 2 typically greater than 0.83). Streamwise velocities from the pixel time stack technique had high correlation for visible-band imagery during early morning daylight hours with respect to mean current meter velocities (r 2 > 0.86). Streamwise velocities for the particle image velocimetry technique for visible-band imagery had weaker correlations with only three out of six correlations performed having an r 2 exceeding 0.6. Copyright 2012 by the American Geophysical Union.

Water Resources Research↗
USGS2012

A method for physically based model analysis of conjunctive use in response to potential climate changes

Potential climate change effects on aspects of conjunctive management of water resources can be evaluated by linking climate models with fully integrated groundwater-surface water models. The objective of this study is to develop a modeling system that links global climate models with regional hydrologic models, using the California Central Valley as a case study. The new method is a supply and demand modeling framework that can be used to simulate and analyze potential climate change and conjunctive use. Supply-constrained and demand-driven linkages in the water system in the Central Valley are represented with the linked climate models, precipitation-runoff models, agricultural and native vegetation water use, and hydrologic flow models to demonstrate the feasibility of this method. Simulated precipitation and temperature were used from the GFDL-A2 climate change scenario through the 21st century to drive a regional water balance mountain hydrologic watershed model (MHWM) for the surrounding watersheds in combination with a regional integrated hydrologic model of the Central Valley (CVHM). Application of this method demonstrates the potential transition from predominantly surface water to groundwater supply for agriculture with secondary effects that may limit this transition of conjunctive use. The particular scenario considered includes intermittent climatic droughts in the first half of the 21st century followed by severe persistent droughts in the second half of the 21st century. These climatic droughts do not yield a valley-wide operational drought but do cause reduced surface water deliveries and increased groundwater abstractions that may cause additional land subsidence, reduced water for riparian habitat, or changes in flows at the Sacramento-San Joaquin River Delta. The method developed here can be used to explore conjunctive use adaptation options and hydrologic risk assessments in regional hydrologic systems throughout the world.

Water Resources Research↗
USGS2012

Interannual variability of snowmelt in the Sierra Nevada and Rocky Mountains, United States: examples from two alpine watersheds

The distribution of snow and the energy flux components of snowmelt are intrinsic characteristics of the alpine water cycle controlling the location of source waters and the effect of climate on streamflow. Interannual variability of these characteristics is relevant to the effect of climate change on alpine hydrology. Our objective is to characterize the interannual variability in the spatial distribution of snow and energy fluxes of snowmelt in watersheds of a maritime setting, Tokopah Basin (TOK) in California's southern Sierra Nevada, and a continental setting, Green Lake 4 Valley (GLV4) in Colorado's Front Range, using a 12 year database (1996–2007) of hydrometeorological observations and satellite-derived snow cover. Snowpacks observed in GLV4 exhibit substantially greater spatial variability than in TOK (0.75 versus 0.28 spatial coefficient of variation). In addition, modeling results indicate that the net turbulent energy flux contribution to snowmelt in GLV4 is, on average, 3 times greater in magnitude (mean 29% versus 10%) and interannual variability (standard deviation 17% versus 6%) than in TOK. These energy flux values exhibit strong seasonality, increasing as the melt season progresses to times later in the year (R 2 = 0.54–0.77). This seasonality of energy flux appears to be associated with snowmelt rates that generally increase with onset date of melt (0.02 cm d -2 ). This seasonality in snowmelt rate, coupled to differences in hydrogeology, may account for the observed differences in correspondence between the timing of snowmelt and timing of streamflow in these watersheds.

Sierra Nevada;Rocky Mountains↗
USGS2012

Factors controlling nitrate fluxes in groundwater in agricultural areas

The impact of agricultural chemicals on groundwater quality depends on the interactions of biogeochemical and hydrologic factors. To identify key processes affecting distribution of agricultural nitrate in groundwater, a parsimonious transport model was applied at 14 sites across the U.S. Simulated vertical profiles of NO 3 - , N 2 from denitrification, O 2 , Cl - , and environmental tracers of groundwater age were matched to observations by adjusting the parameters for recharge rate, unsaturated zone travel time, fractions of N and Cl - inputs leached to groundwater, O 2 reduction rate, O 2 threshold for denitrification, and denitrification rate. Model results revealed important interactions among biogeochemical and physical factors. Chloride fluxes decreased between the land surface and water table possibly because of Cl - exports in harvested crops (averaging 22% of land-surface Cl - inputs). Modeled zero-order rates of O 2 reduction and denitrification were correlated. Denitrification rates at depth commonly exceeded overlying O 2 reduction rates, likely because shallow geologic sources of reactive electron donors had been depleted. Projections indicated continued downward migration of NO 3 - fronts at sites with denitrification rates <0.25 mg-N L -1 yr -1 . The steady state depth of NO 3 - depended to a similar degree on application rate, leaching fraction, recharge, and NO 3 - and O 2 reaction rates. Steady state total mass in each aquifer depended primarily on the N application rate. In addition to managing application rates at land surface, efficient water use may reduce the depth and mass of N in groundwater because lower recharge was associated with lower N fraction leached. Management actions to reduce N leaching could be targeted over aquifers with high-recharge and low-denitrification rates.

Water Resources Research↗
USGS2012

Target loads of atmospheric sulfur and nitrogen deposition for protection of acid sensitive aquatic resources in the Adirondack Mountains, New York

The dynamic watershed acid-base chemistry model of acidification of groundwater in catchments (MAGIC) was used to calculate target loads (TLs) of atmospheric sulfur and nitrogen deposition expected to be protective of aquatic health in lakes in the Adirondack ecoregion of New York. The TLs were calculated for two future dates (2050 and 2100) and three levels of protection against lake acidification (acid neutralizing capacity (ANC) of 0, 20, and 50 eq L -1). Regional sulfur and nitrogen deposition estimates were combined with TLs to calculate exceedances. Target load results, and associated exceedances, were extrapolated to the regional population of Adirondack lakes. About 30% of Adirondack lakes had simulated TL of sulfur deposition less than 50 meq m -2 yr to protect lake ANC to 50 eq L -1. About 600 Adirondack lakes receive ambient sulfur deposition that is above this TL, in some cases by more than a factor of 2. Some critical criteria threshold values were simulated to be unobtainable in some lakes even if sulfur deposition was to be decreased to zero and held at zero until the specified endpoint year. We also summarize important lessons for the use of target loads in the management of acid-impacted aquatic ecosystems, such as those in North America, Europe, and Asia. Copyright 2012 by the American Geophysical Union.

Water Resources Research↗
USGS2012

The use of multiobjective calibration and regional sensitivity analysis in simulating hyporheic exchange

We describe an approach for calibrating a two-dimensional (2-D) flow model of hyporheic exchange using observations of temperature and pressure to estimate hydraulic and thermal properties. A longitudinal 2-D heat and flow model was constructed for a riffle-pool sequence to simulate flow paths and flux rates for variable discharge conditions. A uniform random sampling approach was used to examine the solution space and identify optimal values at local and regional scales. We used a regional sensitivity analysis to examine the effects of parameter correlation and nonuniqueness commonly encountered in multidimensional modeling. The results from this study demonstrate the ability to estimate hydraulic and thermal parameters using measurements of temperature and pressure to simulate exchange and flow paths. Examination of the local parameter space provides the potential for refinement of zones that are used to represent sediment heterogeneity within the model. The results indicate vertical hydraulic conductivity was not identifiable solely using pressure observations; however, a distinct minimum was identified using temperature observations. The measured temperature and pressure and estimated vertical hydraulic conductivity values indicate the presence of a discontinuous low-permeability deposit that limits the vertical penetration of seepage beneath the riffle, whereas there is a much greater exchange where the low-permeability deposit is absent. Using both temperature and pressure to constrain the parameter estimation process provides the lowest overall root-mean-square error as compared to using solely temperature or pressure observations. This study demonstrates the benefits of combining continuous temperature and pressure for simulating hyporheic exchange and flow in a riffle-pool sequence. Copyright 2012 by the American Geophysical Union.

Nevada↗
USGS2012

Role of turbulence fluctuations on uncertainties of acoutic Doppler current profiler discharge measurements

This work presents a systematic analysis quantifying the role of the presence of turbulence fluctuations on uncertainties (random errors) of acoustic Doppler current profiler (ADCP) discharge measurements from moving platforms. Data sets of three-dimensional flow velocities with high temporal and spatial resolution were generated from direct numerical simulation (DNS) of turbulent open channel flow. Dimensionless functions relating parameters quantifying the uncertainty in discharge measurements due to flow turbulence (relative variance and relative maximum random error) to sampling configuration were developed from the DNS simulations and then validated with field-scale discharge measurements. The validated functions were used to evaluate the role of the presence of flow turbulence fluctuations on uncertainties in ADCP discharge measurements. The results of this work indicate that random errors due to the flow turbulence are significant when: (a) a low number of transects is used for a discharge measurement, and (b) measurements are made in shallow rivers using high boat velocity (short time for the boat to cross a flow turbulence structure).

Water Resources Research↗
USGS2012

Stability of infinite slopes under transient partially saturated seepage conditions

Prediction of the location and timing of rainfall‐induced shallow landslides is desired by organizations responsible for hazard management and warnings. However, hydrologic and mechanical processes in the vadose zone complicate such predictions. Infiltrating rainfall must typically pass through an unsaturated layer before reaching the irregular and usually discontinuous shallow water table. This process is dynamic and a function of precipitation intensity and duration, the initial moisture conditions and hydrologic properties of the hillside materials, and the geometry, stratigraphy, and vegetation of the hillslope. As a result, pore water pressures, volumetric water content, effective stress, and thus the propensity for landsliding vary over seasonal and shorter time scales. We apply a general framework for assessing the stability of infinite slopes under transient variably saturated conditions. The framework includes profiles of pressure head and volumetric water content combined with a general effective stress for slope stability analysis. The general effective stress, or suction stress, provides a means for rigorous quantification of stress changes due to rainfall and infiltration and thus the analysis of slope stability over the range of volumetric water contents and pressure heads relevant to shallow landslide initiation. We present results using an analytical solution for transient infiltration for a range of soil texture and hydrological properties typical of landslide‐prone hillslopes and show the effect of these properties on the timing and depth of slope failure. We follow by analyzing field‐monitoring data acquired prior to shallow landslide failure of a hillside near Seattle, Washington, and show that the timing of the slide was predictable using measured pressure head and volumetric water content and show how the approach can be used in a forward manner using a numerical model for transient infiltration.

Water Resources Research↗
USGS2012

A low-cost method to measure the timing of post-fire flash floods and debris flows relative to rainfall

Data on the specific timing of post-fire flash floods and debris flows are very limited. We describe a method to measure the response times of small burned watersheds to rainfall using a low-cost pressure transducer, which can be installed quickly after a fire. Although the pressure transducer is not designed for sustained sampling at the fast rates ({less than or equal to}2 sec) used at more advanced debris-flow monitoring sites, comparisons with high-data rate stage data show that measured spikes in pressure sampled at 1-min intervals are sufficient to detect the passage of most debris flows and floods. Post-event site visits are used to measure the peak stage and identify flow type based on deposit characteristics. The basin response timescale (tb) to generate flow at each site was determined from an analysis of the cross correlation between time series of flow pressure and 5-min rainfall intensity. This timescale was found to be less than 30 minutes for 40 post-fire floods and 11 post-fire debris flows recorded in 15 southern California watersheds ({less than or equal to} 1.4 km 2 ). Including data from 24 other debris flows recorded at 5 more instrumentally advanced monitoring stations, we find there is not a substantial difference in the median tb for floods and debris flows (11 and 9 minutes, respectively); however, there are slight, statistically significant differences in the trends of flood and debris-flow tb with basin area, which are presumably related to differences in flow speed between floods and debris flows.

Water Resources Research↗
USGS2012

Hydrologic conditions controlling runoff generation immediately after wildfire

We investigated the control of postwildfire runoff by physical and hydraulic properties of soil, hydrologic states, and an ash layer immediately following wildfire. The field site is within the area burned by the 2010 Fourmile Canyon Fire in Colorado, USA. Physical and hydraulic property characterization included ash thickness, particle size distribution, hydraulic conductivity, and soil water retention curves. Soil water content and matric potential were measured indirectly at several depths below the soil surface to document hydrologic states underneath the ash layer in the unsaturated zone, whereas precipitation and surface runoff were measured directly. Measurements of soil water content showed that almost no water infiltrated below the ash layer into the near-surface soil in the burned site at the storm time scale (i.e., minutes to hours). Runoff generation processes were controlled by and highly sensitive to ash thickness and ash hydraulic properties. The ash layer stored from 97% to 99% of rainfall, which was critical for reducing runoff amounts. The hydrologic response to two rain storms with different rainfall amounts, rainfall intensity, and durations, only ten days apart, indicated that runoff generation was predominantly by the saturation-excess mechanism perched at the ash-soil interface during the first storm and predominantly by the infiltration-excess mechanism at the ash surface during the second storm. Contributing area was not static for the two storms and was 4% (saturation excess) to 68% (infiltration excess) of the catchment area. Our results showed the importance of including hydrologic conditions and hydraulic properties of the ash layer in postwildfire runoff generation models.

Colorado↗
USGS2012

Synthesis of benthic flux components in the Patos Lagooncoastal zone, Rio Grande do Sul, Brazil

The primary objective of this work is to synthesize components of benthic flux in the Patos Lagoon coastal zone, Rio Grande do Sul, Brazil. Specifically, the component of benthic discharge flux forced by the terrestrial hydraulic gradient is 0.8 m 3 d -1 ; components of benthic discharge and recharge flux associated with the groundwater tidal prism are both 2.1 m 3 d -1 ; components of benthic discharge and recharge flux forced by surface-gravity wave setup are both 6.3 m 3 d -1 ; the component of benthic discharge flux that transports radium-228 is 350 m 3 d -1 ; and components of benthic discharge and recharge flux forced by surface-gravity waves propagating over a porous medium are both 1400 m 3 d -1 . (All models are normalized per meter shoreline.) Benthic flux is a function of components forced by individual mechanisms and nonlinear interactions that exist between components. Constructive and destructive interference may enhance or diminish the contribution of benthic flux components. It may not be possible to model benthic flux by summing component magnitudes. Geochemical tracer techniques may not accurately model benthic discharge flux or submarine groundwater discharge (SGD). A conceptual model provides a framework on which to quantitatively characterize benthic discharge flux and SGD with a multifaceted approach.

Patos Lagoon↗
USGS2012

Thresholds of flow-induced bed disturbances and their effects on stream metabolism in an agricultural river

Storm-driven flow pulses in rivers destroy and restructure sediment habitats that affect stream metabolism. This study examined thresholds of bed disturbances that affected patch- and reach-scale sediment conditions and metabolism rates. A 4 year record of discharge and diel changes in dissolved oxygen concentrations (ΔDO) was analyzed for disturbances and recovery periods of the ΔDO signal. Disturbances to the ΔDO signal were associated with flow pulses, and the recovery times for the ΔDO signal were found to be in two categories: less than 5 days (30% of the disturbances) or greater than 15 days (70% of the disturbances). A field study was performed during the fall of 2007, which included a storm event that increased discharge from 3.1 to 6.9 m 3 /s over a 7 h period. During stable flow conditions before the storm, variability in patch-scale stream metabolism values were associated with sediment texture classes with values ranging from −16.4 to 2.3 g O 2 2 /d (negative sign indicates net respiration) that bounded the reach-averaged rate of −5.6 g O 2 2 /d. Hydraulic modeling of bed shear stresses demonstrated a storm-induced flow pulse mobilized approximately 25% of the bed and reach-scale metabolism rates shifted from −5 to −40 g O 2 2 /d. These results suggest that storm-induced bed disturbances led to threshold behavior with respect to stream metabolism. Small flow pulses resulted in partial-bed mobilization that disrupted stream metabolism by increased turbidity with short recovery times. Large flow pulses resulted in full-bed mobilization that disrupted stream metabolism by destroying periphyton habitats with long recovery times.

Water Resources Research↗
USGS2012

Physical controls and predictability of stream hyporheic flow evaluated with a multiscale model

Improved predictions of hyporheic exchange based on easily measured physical variables are needed to improve assessment of solute transport and reaction processes in watersheds. Here we compare physically based model predictions for an Indiana stream with stream tracer results interpreted using the Transient Storage Model (TSM). We parameterized the physically based, Multiscale Model (MSM) of stream-groundwater interactions with measured stream planform and discharge, stream velocity, streambed hydraulic conductivity and porosity, and topography of the streambed at distinct spatial scales (i.e., ripple, bar, and reach scales). We predicted hyporheic exchange fluxes and hyporheic residence times using the MSM. A Continuous Time Random Walk (CTRW) model was used to convert the MSM output into predictions of in stream solute transport, which we compared with field observations and TSM parameters obtained by fitting solute transport data. MSM simulations indicated that surface-subsurface exchange through smaller topographic features such as ripples was much faster than exchange through larger topographic features such as bars. However, hyporheic exchange varies nonlinearly with groundwater discharge owing to interactions between flows induced at different topographic scales. MSM simulations showed that groundwater discharge significantly decreased both the volume of water entering the subsurface and the time it spent in the subsurface. The MSM also characterized longer timescales of exchange than were observed by the tracer-injection approach. The tracer data, and corresponding TSM fits, were limited by tracer measurement sensitivity and uncertainty in estimates of background tracer concentrations. Our results indicate that rates and patterns of hyporheic exchange are strongly influenced by a continuum of surface-subsurface hydrologic interactions over a wide range of spatial and temporal scales rather than discrete processes.

Water Resources Research↗
USGS2012

Influence of permafrost distribution on groundwater flow in the context of climate-driven permafrost thaw: Example from Yukon Flats Basin, Alaska, United States

Understanding the role of permafrost in controlling groundwater flow paths and fluxes is central in studies aimed at assessing potential climate change impacts on vegetation, species habitat, biogeochemical cycling, and biodiversity. Recent field studies in interior Alaska show evidence of hydrologic changes hypothesized to result from permafrost degradation. This study assesses the hydrologic control exerted by permafrost, elucidates modes of regional groundwater flow for various spatial permafrost patterns, and evaluates potential hydrologic consequences of permafrost degradation. The Yukon Flats Basin (YFB), a large (118,340 km 2 ) subbasin within the Yukon River Basin, provides the basis for this investigation. Model simulations that represent an assumed permafrost thaw sequence reveal the following trends with decreasing permafrost coverage: (1) increased groundwater discharge to rivers, consistent with historical trends in base flow observations in the Yukon River Basin, (2) potential for increased overall groundwater flux, (3) increased spatial extent of groundwater discharge in lowlands, and (4) decreased proportion of suprapermafrost (shallow) groundwater contribution to total base flow. These trends directly affect the chemical composition and residence time of riverine exports, the state of groundwater-influenced lakes and wetlands, seasonal river-ice thickness, and stream temperatures. Presently, the YFB is coarsely mapped as spanning the continuous-discontinuous permafrost transition that model analysis shows to be a critical threshold; thus, the YFB may be on the verge of major hydrologic change should the current permafrost extent decrease. This possibility underscores the need for improved characterization of permafrost and other hydrogeologic information in the region via geophysical techniques, remote sensing, and ground-based observations.

Alaska↗
USGS2013

Evaluating model structure adequacy: The case of the Maggia Valley groundwater system, southern Switzerland

Model adequacy is evaluated with alternative models rated using model selection criteria (AICc, BIC, and KIC) and three other statistics. Model selection criteria are tested with cross-validation experiments and insights for using alternative models to evaluate model structural adequacy are provided. The study is conducted using the computer codes UCODE_2005 and MMA (MultiModel Analysis). One recharge alternative is simulated using the TOPKAPI hydrological model. The predictions evaluated include eight heads and three flows located where ecological consequences and model precision are of concern. Cross-validation is used to obtain measures of prediction accuracy. Sixty-four models were designed deterministically and differ in representation of river, recharge, bedrock topography, and hydraulic conductivity. Results include: (1) What may seem like inconsequential choices in model construction may be important to predictions. Analysis of predictions from alternative models is advised. (2) None of the model selection criteria consistently identified models with more accurate predictions. This is a disturbing result that suggests to reconsider the utility of model selection criteria, and/or the cross-validation measures used in this work to measure model accuracy. (3) KIC displayed poor performance for the present regression problems; theoretical considerations suggest that difficulties are associated with wide variations in the sensitivity term of KIC resulting from the models being nonlinear and the problems being ill-posed due to parameter correlations and insensitivity. The other criteria performed somewhat better, and similarly to each other. (4) Quantities with high leverage are more difficult to predict. The results are expected to be generally applicable to models of environmental systems.

Water Resources Research↗