Geology Reports⌕ Search

SEARCH · Geology Reports

Results for “Algorithms”

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 1,297 records · Page 72Linked to original sources
USGS2010

Integrated satellite observations of the 2006 eruption of Augustine Volcano: Chapter 20 in The 2006 eruption of Augustine Volcano, Alaska

Satellite observations played an important role in monitoring the 2006 eruption of Augustine Volcano. It represented the first opportunity for observers to use, in an operational setting, new Web-based tools and techniques developed by the Alaska Volcano Observatory remote sensing group. The 'Okmok Algorithm' was used to analyze thermal infrared satellite data and highlight changes in the style and phases of activity. Temperature measurements were used to estimate ash cloud heights, which compared favorably to radar and ground-based observations, although larger discrepancies were seen when compared to pilot reports. Brightness temperature difference techniques were used to locate and track the 14 ash clouds produced during the explosive phase of the eruption. Stacking of these analyses allowed the creation of composite maps showing the distribution of airborne ash. The data from these maps were further combined with information from local reports and samples of ashfall to create a prototype of a concentration map that could be used to assess the potential hazard an eruption represents to aircraft, infrastructure and human health.

Alaska↗
USGS2023

Magnetotelluric investigation of northern Harrat Rahat, Kingdom of Saudi Arabia

Volcanism within the harrats (Arabic for “volcanic field”) of the Kingdom of Saudi Arabia includes at least one historical eruption occurring close to the holy city of Al Madīnah in 1256 C.E. As part of a volcanic- and seismic-hazard assessment of northern Harrat Rahat, magnetotelluric (MT) data were collected to investigate the structural setting of the area, the presence or absence of melt within the crust, and the mantle-derived magmatic source. Collected MT data were modeled in both two dimensions, where anisotropy can be estimated, and three dimensions. Interpretation of the preferred resistivity model includes a shallow sediment-filled graben beneath northern Harrat Rahat lavas, a melt-free upper crust, and a region of decompression melting in the asthenosphere below 60–70 kilometers depth. Models in two dimensions image the lower crust as anisotropic, demonstrating that a series of elongate conductivity anomalies with a strike of N. 10° E. within the lower crust of the three-dimensional model are artifacts of inverting anisotropic data with an isotropic modeling algorithm. Careful examination of the resistivity models, in combination with regional geological and geophysical data, suggests an anisotropic lower crust that is free of large zones of melt. Azimuthal anisotropy in the lower crust extends well beyond the limits of Harrat Rahat volcanic rocks, with a conductive direction oriented N. 10° E. and an anisotropy factor of 2–5 between the most and least conductive directions. Enhanced conductivity is likely caused by interconnected grain-boundary graphite, where the direction of anisotropy reflects either frozen-in fabric from the Neoproterozoic stabilization of the Arabian Shield or ductile deformation driven by channelized asthenospheric flow coupled with a thin rigid mantle lid. Asthenospheric melt is interpreted to transect the crust largely through diking, with limited melt storage and short residence times within the crustal column.

northern Harrat Rahat↗
USGS2023

Thickness of the Saudi Arabian crust

As part of a joint Saudi Geological Survey (SGS) and U.S. Geological Survey (USGS) project, we analyzed P-wave receiver functions from seismic stations covering most of the Kingdom of Saudi Arabia to map the thickness of the crust across the Arabia Plate. We present an update of crustal-thickness estimates and fill in gaps for the western Arabian Shield and the rifted margin at the Red Sea (the coastal plain), as well as the eastern Arabian Platform. We applied a conventional H-k stacking algorithm and included careful attention to stacking weights, two forms of sedimentary corrections for stations located on the Arabian Platform, and additional processing for noisy stations. We obtained useful results at 154 stations from 898 teleseismic events over a 2-year period from 1995–1997 (for non-SGS stations) and a 6-year period from 2008–2014 (for SGS stations). Average crustal thickness (that is, depth to the Mohorovičić discontinuity [Moho] below the surface) beneath the Red Sea coastal plain (the rift margin) is 29 kilometers (km), beneath the volcanic fields (known in Arabic as harra [plural] or harrat [singular]) is 35 km, beneath the Arabian Shield (excluding harrats) is 37 km, and beneath the Arabian Platform is 38 km. Crustal thinning appears not to extend east of the rift escarpment, suggesting uniform extension that is no broader at depth than at the surface. In contrast to some previous claims that the Arabian Platform crust is thicker than that of the Arabian Shield, we find no statistically significant difference between their whole crustal thicknesses. However, the average subsedimentary crustal thickness (that is, the crystalline crust) for stations on the Arabian Platform is 34 km, 3 km thinner than the crust of the Arabian Shield. Individual station P-wave (pressure) velocity and S-wave (shear) velocity ratios ( V P / V S ) are highly variable for the Arabia Plate, ranging from 1.60 to 1.97 and averaging 1.75, with a standard deviation of 0.07. There are no statistically significant differences between V P / V S ratios of the different geologic regions of Saudi Arabia. Similar V P / V S ratios, coupled with similar crustal thicknesses for harrats and the Arabian Shield, indicate that Cenozoic magmatism has contributed negligibly to crustal growth.

Professional Paper↗
USGS2023

Ambient seismic noise tomography of the Kingdom of Saudi Arabia

Harrat Rahat is a Cenozoic volcanic field in the west-central part of the Kingdom of Saudi Arabia, 150 kilometers east of the Red Sea, and is the site of the most recent eruption in the country (1256 C.E.; 654 in the year of the Hijra). The city of Al Madīnah lies at the north end of Harrat Rahat, and its volcanic and seismic risks are frequently reassessed. In 2009 C.E. an earthquake swarm at Harrat Lunayyir, 200 km west-northwest of Al Madīnah, produced significant shaking and some building damage in nearby communities, which prompted a revision of seismic hazard models for the region. A network of seismic stations on this volcanically active western side of the Arabia Plate was installed, and stations were also added in the tectonically active northern part of the country. Although regional earthquakes may be used to determine the crustal structure of the western Arabia Plate, such crustal models are hindered by insufficient numbers of earthquakes in the stable plate interior. Tomography studies can be used to infer material properties of the subsurface, such as presence of partial melt, and are beneficial for volcanic hazard assessments. We use ambient seismic noise to compute Rayleigh and Love surface-wave dispersion maps between 5 and 12 second periods for a subset of seismic stations within and near northern Harrat Rahat. The surface-wave maps are inverted to produce shear-wave velocities using a neighborhood algorithm and interpolated into a pseudo three-dimensional model. The distributions of surface-wave and shear-wave velocities are heterogenous, varying from ±3–8 percent. However, low velocities are not restricted to Harrat Rahat. We observed a difference between Rayleigh- and Love-wave velocities that extends north of the site of the 1256 C.E. eruption and coincides with a low gravity anomaly. We obtain a shear-wave velocity increase of 10–15 percent between 15- and 25-kilometers depth, which is consistent with the presence of a transition between the felsic upper crust and the mafic lower crust of the Arabian Shield. The average shear-wave velocities of the upper and lower crust are estimated to be 3.64 and 3.95 kilometers per second using Rayleigh waves and 3.53 and 4.16 kilometers per second using Love waves, which are in good agreement with the results of other geophysical surveys in this area. The modest magnitude of the low-velocity anomalies within the crust and their locations extending well beyond the limits of Harrat Rahat indicate that they are not caused by a crustal magma chamber. If magma chambers exist, they are smaller than can be imaged with our seismological method (resolution corresponds to a 15-kilometer wavelength), deeper than 30 kilometers, or shallower than 5 kilometers with a small velocity contrast. The last possibility is discounted by weak-to-absent surface geothermal activity and by a lack of shallow seismicity that characterizes other areas with known shallow magmas, such as Hawaiʻi. We then expanded our work to the entire Arabian Shield with principally the same methodology to look at large-scale regional patterns. We found modest shear-wave velocity deviations on the order of only ±3 percent, which are within expected ranges for lithological variation.

Professional Paper↗
USGS2012

Maps of estimated nitrate and arsenic concentrations in basin-fill aquifers of the southwestern United States

Human-health concerns and economic considerations associated with meeting drinking-water standards motivated a study of the vulnerability of basin-fill aquifers to nitrate contamination and arsenic enrichment in the southwestern United States. Statistical models were developed by using the random forest classifier algorithm to predict concentrations of nitrate and arsenic across a model grid representing about 190,600 square miles of basin-fill aquifers in parts of Arizona, California, Colorado, Nevada, New Mexico, and Utah. The statistical models, referred to as classifiers, reflect natural and human-related factors that affect aquifer vulnerability to contamination and relate nitrate and arsenic concentrations to explanatory variables representing local- and basin-scale measures of source and aquifer susceptibility conditions. Geochemical variables were not used in concentration predictions because they were not available for the entire study area. The models were calibrated to assess model accuracy on the basis of measured values. Only 2 percent of the area underlain by basin-fill aquifers in the study area was predicted to equal or exceed the U.S. Environmental Protection Agency drinking-water standard for nitrate as N (10 milligrams per liter), whereas 43 percent of the area was predicted to equal or exceed the standard for arsenic (10 micrograms per liter). Areas predicted to equal or exceed the drinking-water standard for nitrate include basins in central Arizona near Phoenix; the San Joaquin Valley, the Santa Ana Inland, and San Jacinto Basins of California; and the San Luis Valley of Colorado. Much of the area predicted to equal or exceed the drinking-water standard for arsenic is within a belt of basins along the western portion of the Basin and Range Physiographic Province that includes almost all of Nevada and parts of California and Arizona. Predicted nitrate and arsenic concentrations are substantially lower than the drinking-water standards in much of the study area-about 93 percent of the area underlain by basin-fill aquifers was less than one-half the standard for nitrate as N (5.0 milligrams per liter), and 50 percent was less than one-half the standard for arsenic (5.0 micrograms per liter). The predicted concentrations and the improved understanding of the susceptibility and vulnerability of southwestern basin-fill aquifers to nitrate contamination and arsenic enrichment can be used by water managers as a qualitative tool to assess and protect the quality of groundwater resources in the Southwest.

Arizona, California, Colorado, Idaho, Nevada, New ↗
USGS2018

Isostatic gravity map of Mountain Pass and vicinity, California and Nevada

Gravity investigations of Mountain Pass and vicinity were begun as part of an effort to study regional crustal structures as an aid to understanding the geologic framework and mineral resources of the eastern Mojave Desert. The study area, which straddles the state boundary between southeastern California and southern Nevada, encompasses Mountain Pass, which is host to one of the world’s largest rare earth element carbonatite deposits. The deposit is found along a north-northwest-trending, fault-bounded block that extends along the eastern parts of the Clark Mountain Range, Mescal Range, and Ivanpah Mountains. This Paleoproterozoic block is composed of a 1.7-Ga metamorphic complex of gneiss and schist that underwent widespread metamorphism and associated plutonism during the Ivanpah orogeny. The Paleoproterozoic rocks were intruded by a Mesoproterozoic (1.4 Ga) ultrapotassic alkaline intrusive suite and carbonatite body. The intrusive rocks include, from oldest to youngest, shonkinite, mesosyenite, syenite, quartz syenite, potassic granite, carbonatite, carbonatite dikes, and late shonkinite dikes. Generally speaking, gravity anomalies can be used to infer subsurface geologic structure, revealing variations in lithology and delineating features such as faults, plutons, volcanic centers, calderas, and deep sedimentary basins. As part of this study, gravity data from more than 2,400 stations were collected and processed to identify lateral changes in subsurface density. Gravity stations were distributed across parts of Shadow Valley, Clark Mountain Range, Mescal Range, Ivanpah Mountains, and Ivanpah Valley. The new gravity data were combined with preexisting gravity data from the surrounding areas in California and Nevada. All gravity data were gridded using a minimum curvature algorithm at an interval of 200 m, and the result is displayed as a color-contour isostatic gravity map.

California, Nevada↗
USGS2005

Analysis and mapping of post-fire hydrologic hazards for the 2002 Hayman, Coal Seam, and Missionary Ridge wildfires, Colorado

Wildfires caused extreme changes in the hydrologic, hydraulic, and geomorphologic characteristics of many Colorado drainage basins in the summer of 2002. Detailed assessments were made of the short-term effects of three wildfires on burned and adjacent unburned parts of drainage basins. These were the Hayman, Coal Seam, and Missionary Ridge wildfires. Longer term runoff characteristics that reflect post-fire drainage basin recovery expected to develop over a period of several years also were analyzed for two affected stream reaches: the South Platte River between Deckers and Trumbull, and Mitchell Creek in Glenwood Springs. The 10-, 50-, 100-, and 500-year flood-plain boundaries and water-surface profiles were computed in a detailed hydraulic study of the Deckers-to-Trumbull reach. The Hayman wildfire burned approximately 138,000 acres (216 square miles) in granitic terrain near Denver, and the predominant potential hazard in this area is flooding by sediment-laden water along the large tributaries to and the main stem of the South Platte River. The Coal Seam wildfire burned approximately 12,200 acres (19.1 square miles) near Glenwood Springs, and the Missionary Ridge wildfire burned approximately 70,500 acres (110 square miles) near Durango, both in areas underlain by marine shales where the predominant potential hazard is debris-flow inundation of low-lying areas. Hydrographs and peak discharges for pre-burn and post-burn scenarios were computed for each drainage basin and tributary subbasin by using rainfall-runoff models because streamflow data for most tributary subbasins were not available. An objective rainfall-runoff model calibration method based on nonlinear regression and referred to as the ?objective calibration method? was developed and applied to rainfall-runoff models for three burned areas. The HEC-1 rainfall-runoff model was used to simulate the pre-burn rainfall-runoff processes in response to the 100-year storm, and HEC-HMS was used for runoff hydrograph generation. Post-burn rainfall-runoff parameters were determined by adjusting the runoff-curve numbers on the basis of a weighting procedure derived from the U.S. Soil Conservation Service (now the National Resources Conservation Service) equation for precipitation excess and the effect of burn severity. This weighting procedure was determined to be more appropriate than simple area weighting because of the potentially marked effect of even small burned areas on the runoff hydrograph in individual drainage basins. Computed water-peak discharges from HEC-HMS models were increased volumetrically to account for increased sediment concentrations that are expected as a result of accelerated erosion after burning. Peak discharge estimates for potential floods in the South Platte River were increased by a factor that assumed a volumetric sediment concentration (Cv) of 20 percent. Flood hydrographs for the South Platte River and Mitchell Creek were routed down main-stem channels using watershed-routing algorithms included in the HEC-HMS rainfall-runoff model. In areas subject to debris flows in the Coal Seam and Missionary Ridge burned areas, debris-flow discharges were simulated by 100-year rainfall events, and the inflow hydrographs at tributary mouths were simulated by using the objective calibration method. Sediment concentrations (Cv) used in debris-flow simulations were varied through the event, and were initial Cv 20 percent, mean Cv approximately 31 percent, maximum Cv 48 percent, Cv 43 percent at the time of the water hydrograph peak, and Cv 20 percent for the duration of the event. The FLO-2D flood- and debris-flow routing model was used to delineate the area of unconfined debris-flow inundation on selected alluvial fan and valley floor areas. A method was developed to objectively determine the post-fire recovery period for the Hayman and Coal Seam burned areas using runoff-curve numbers (RCN) for all drainage basins for a 50-year period. A

Scientific Investigations Report↗
USGS2006

MODOPTIM: A general optimization program for ground-water flow model calibration and ground-water management with MODFLOW

MODOPTIM is a non-linear ground-water model calibration and management tool that simulates flow with MODFLOW-96 as a subroutine. A weighted sum-of-squares objective function defines optimal solutions for calibration and management problems. Water levels, discharges, water quality, subsidence, and pumping-lift costs are the five direct observation types that can be compared in MODOPTIM. Differences between direct observations of the same type can be compared to fit temporal changes and spatial gradients. Water levels in pumping wells, wellbore storage in the observation wells, and rotational translation of observation wells also can be compared. Negative and positive residuals can be weighted unequally so inequality constraints such as maximum chloride concentrations or minimum water levels can be incorporated in the objective function. Optimization parameters are defined with zones and parameter-weight matrices. Parameter change is estimated iteratively with a quasi-Newton algorithm and is constrained to a user-defined maximum parameter change per iteration. Parameters that are less sensitive than a user-defined threshold are not estimated. MODOPTIM facilitates testing more conceptual models by expediting calibration of each conceptual model. Examples of applying MODOPTIM to aquifer-test analysis, ground-water management, and parameter estimation problems are presented.

Scientific Investigations Report↗
USGS2008

Mobility statistics and automated hazard mapping for debris flows and rock avalanches

Power-law equations that are physically motivated and statistically tested and calibrated provide a basis for forecasting areas likely to be inundated by debris flows, rock avalanches, and lahars with diverse volumes. The equations A=α 1 V 2/3 and B=α 2 V 2/3 are based on the postulate that the maximum valley cross-sectional area (A) and total valley planimetric area (B) likely to be inundated by a flow depend only on its volume (V) and the topography of the flow path. Testing of these equations involves determining whether or not they fit data for documented flows satisfactorily, and calibration entails determining best-fit values of the coefficients α 1 and α 2 for debris flows, rock avalanches, and lahars. This report describes statistical testing and calibration of the equations by using field data compiled from many sources, and it describes application of the equations to delineation of debris-flow hazard zones. Statistical results show that for each type of flow (debris flows, rock avalanches, and lahars), the dependence of A and B on V is described well by power laws with exponents equal to 2/3. This value of the exponent produces fits that are effectively indistinguishable from the best fits obtained by using adjustable power-law exponents. Statistically calibrated values of the coefficients α 1 and α 2 provide scale-invariant indices of the relative mobilities of rock avalanches (α 1 = 0.2, α 2 = 20), nonvolcanic debris flows (α 1 = 0.1, α 2 = 20), and lahars (α 1 = 0.05, α 2 = 200). These values show, for example, that a lahar of specified volume can be expected to inundate a planimetric area ten times larger than that inundated by a rock avalanche or nonvolcanic debris flow of the same volume. The utility of the calibrated debris-flow inundation equations A=0.1V 2/3 and B=20V 2/3 is demonstrated by using them within the GIS program LAHARZ to delineate nested hazard zones for future debris flows in an area bordering the Umpqua River in the south-central Oregon Coast Range. This application requires use of high-resolution topographic data derived form LIDAR surveys, knowledge of local geology to specify a suitable range of prospective debris-flow volumes, and development and use of a new algorithm for identification of prospective debris-flow source areas in finely dissected terrain.

Oregon↗
USGS2008

Simulation of the Quantity, Variability, and Timing of Streamflow in the Dennys River Basin, Maine, by Use of a Precipitation-Runoff Watershed Model

The U.S. Geological Survey (USGS), in cooperation with the Maine Department of Marine Resources Bureau of Sea Run Fisheries and Habitat, began a study in 2004 to characterize the quantity, variability, and timing of streamflow in the Dennys River. The study included a synoptic summary of historical streamflow data at a long-term streamflow gage, collecting data from an additional four short-term streamflow gages, and the development and evaluation of a distributed-parameter watershed model for the Dennys River Basin. The watershed model used in this investigation was the USGS Precipitation-Runoff Modeling System (PRMS). The Geographic Information System (GIS) Weasel was used to delineate the Dennys River Basin and subbasins and derive parameters for their physical geographic features. Calibration of the models used in this investigation involved a four-step procedure in which model output was evaluated against four calibration data sets using computed objective functions for solar radiation, potential evapotranspiration, annual and seasonal water budgets, and daily streamflows. The calibration procedure involved thousands of model runs and was carried out using the USGS software application Luca (Let us calibrate). Luca uses the Shuffled Complex Evolution (SCE) global search algorithm to calibrate the model parameters. The SCE method reliably produces satisfactory solutions for large, complex optimization problems. The primary calibration effort went into the Dennys main stem watershed model. Calibrated parameter values obtained for the Dennys main stem model were transferred to the Cathance Stream model, and a similar four-step SCE calibration procedure was performed; this effort was undertaken to determine the potential to transfer modeling information to a nearby basin in the same region. The calibrated Dennys main stem watershed model performed with Nash-Sutcliffe efficiency (NSE) statistic values for the calibration period and evaluation period of 0.79 and 0.76, respectively. The Cathance Stream model had an NSE value of 0.68. The Dennys River Basin models make use of limited streamflow-gaging station data and provide information to characterize subbasin hydrology. The calibrated PRMS watershed models of the Dennys River Basin provide simulated daily streamflow time series from October 1, 1985, through September 30, 2006, for nearly any location within the basin. These models enable natural-resources managers to characterize the timing and quantity of water moving through the basin to support many endeavors including geochemical calculations, water-use assessment, Atlantic salmon population dynamics and migration modeling, habitat modeling and assessment, and other resource-management scenario evaluations. Characterizing streamflow contributions from subbasins in the basin and the relative amounts of surface- and ground-water contributions to streamflow throughout the basin will lead to a better understanding of water quantity and quality in the basin. Improved water-resources information will support Atlantic salmon protection efforts.

Scientific Investigations Report↗
USGS2009

Magnitude and frequency of rural floods in the southeastern United States, 2006: Volume 1, Georgia

A multistate approach was used to update methods for estimating the magnitude and frequency of floods in rural, ungaged basins in Georgia, South Carolina, and North Carolina that are not substantially affected by regulation, tidal fluctuations, or urban development. Annual peak-flow data through September 2006 were analyzed for 943 streamgaging stations having 10 or more years of data on rural streams in Georgia, South Carolina, North Carolina, and adjacent parts of Alabama, Florida, Tennessee, and Virginia. Flood-frequency estimates were computed for the 943 stations by fitting the logarithms of annual peak flows for each station to a Pearson Type III distribution. As part of the computation of flood-frequency estimates for these streamgaging stations, a new value for the generalized-skew coefficient was developed by using a Bayesian generalized least-squares regression model. Additionally, basin characteristics for the streamgaging stations were computed by using a geographical information system and automated computer algorithms. Regional regression analysis, using generalized least-squares regression, was used to develop a set of predictive equations for estimating the 50-, 20-, 10-, 4-, 2-, 1-, 0.5-, and 0.2-percent chance exceedance flows for rural ungaged basins in Georgia, South Carolina, and North Carolina. Flood-frequency estimates and basin characteristics for 828 stream-gaging stations were combined to form the final database used in the regional regression analysis. Five hydrologic regions were developed for Georgia, South Carolina, and North Carolina. The final predictive equations are all functions of drainage area and percentage of the drainage basin within each hydrologic region. Average standard errors of prediction for these regression equations range from 34.5 to 47.7 percent.

Georgia↗
USGS2009

Development, Testing, and Application of a Coupled Hydrodynamic Surface-Water/Groundwater Model (FTLOADDS) with Heat and Salinity Transport in the Ten Thousand Islands/Picayune Strand Restoration Project Area, Florida

A numerical model application was developed for the coastal area inland of the Ten Thousand Islands (TTI) in southwestern Florida using the Flow and Transport in a Linked Overland/Aquifer Density-Dependent System (FTLOADDS) model. This model couples a two-dimensional dynamic surface-water model with a three-dimensional groundwater model, and has been applied to several locations in southern Florida. The model application solves equations for salt transport in groundwater and surface water, and also simulates surface-water temperature using a newly enhanced heat transport algorithm. One of the purposes of the TTI application is to simulate hydrologic factors that relate to habitat suitability for the West Indian Manatee. Both salinity and temperature have been shown to be important factors for manatee survival. The inland area of the TTI domain is the location of the Picayune Strand Restoration Project, which is designed to restore predevelopment hydrology through the filling and plugging of canals, construction of spreader channels, and the construction of levees and pump stations. The effects of these changes are simulated to determine their effects on manatee habitat. The TTI application utilizes a large amount of input data for both surface-water and groundwater flow simulations. These data include topography, frictional resistance, atmospheric data including rainfall and air temperature, aquifer properties, and boundary conditions for tidal levels, inflows, groundwater heads, and salinities. Calibration was achieved by adjusting the parameters having the largest uncertainty: surface-water inflows, the surface-water transport dispersion coefficient, and evapotranspiration. A sensitivity analysis did not indicate that further parameter changes would yield an overall improvement in simulation results. The agreement between field data from GPS-tracked manatees and TTI application results demonstrates that the model can predict the salinity and temperature fluctuations which affect manatee behavior. Comparison of the existing conditions simulation with the simulation incorporating restoration changes indicated that the restoration would increase the period of inundation for most of the coastal wetlands. Generally, surface-water salinity was lowered by restoration changes in most of the wetlands areas, especially during the early dry season. However, the opposite pattern was observed in the primary canal habitat for manatees, namely, the Port of the Islands. Salinities at this location tended to be moderately elevated during the dry season, and unchanged during the wet season. Water temperatures were in close agreement between the existing conditions and restoration simulations, although minimum temperatures at the Port of the Islands were slightly higher in the restoration simulation as a result of the additional surface-water ponding and warming that occurs in adjacent wetlands. The TTI application output was used to generate salinity and temperature time series for comparison to manatee field tracking data and an individually-based manatee-behavior model. Overlaying field data with salinity and temperature results from the TTI application reflects the effect of warm water availability and the periodic need for low-salinity drinking water on manatee movements. The manatee-behavior model uses the TTI application data at specific model nodes along the main manatee travel corridors to determine manatee migration patterns. The differences between the existing conditions and restoration scenarios can then be compared for manatee refugia. The TTI application can be used to test a variety of hydrologic conditions and their effect on important criteria.

Scientific Investigations Report↗
USGS2009

Magnitude and Frequency of Rural Floods in the Southeastern United States, 2006: Volume 3, South Carolina

A multistate approach was used to update methods for estimating the magnitude and frequency of floods in rural, ungaged basins in South Carolina, Georgia, and North Carolina that are not substantially affected by regulation, tidal fluctuations, or urban development. Annual peak-flow data through September 2006 were analyzed for 943 streamgaging stations having 10 or more years of data on rural streams in South Carolina, Georgia, North Carolina, and adjacent parts of Alabama, Florida, Tennessee, and Virginia. Flood-frequency estimates were computed for the 943 stations by fitting the logarithms of annual peak flows for each station to a Pearson Type III distribution. As part of the computation of flood-frequency estimates for the stations, a new value for the generalized skew coefficient was developed using a Bayesian generalized least-squares regression model. Additionally, basin characteristics for these stations were computed by using a geographical information system and automated computer algorithms. Exploratory regression analyses using ordinary least-squares regression completed on the initial database of 943 gaged stations resulted in defining five hydrologic regions for South Carolina, Georgia, and North Carolina. Stations with drainage areas less than 1 square mile were removed from the database, and a procedure to examine for basin redundancy (based on drainage area and periods of record) also resulted in the removal of some stations from the regression database. Regional regression analysis, using generalized least-squares regression, was used to develop a set of predictive equations for estimating the 50-, 20-, 10-, 4-, 2-, 1-, 0.5-, and 0.2-percent chance exceedance flows for rural ungaged basins in Georgia, South Carolina, and North Carolina. Flood-frequency estimates and basin characteristics for 828 streamgaging stations were combined to form the final database used in the regional regression analysis. The final predictive equations are all functions of drainage area and percentage of the drainage basin within each hydrologic region. Average errors of prediction for these regression equations range from 34.0 to 47.7 percent. Peak-flow records at 25 regulated stations were assessed to determine if a flood-frequency analysis was appropriate. Based on those assessments, flood-frequency estimates are provided for three regulated stations. Annual peak-flow data are provided for the regulated stations in an appendix.

South Carolina↗
USGS2009

Capacitively Coupled Resistivity Survey of Selected Irrigation Canals Within the North Platte River Valley, Western Nebraska and Eastern Wyoming, 2004 and 2007-2009

Due to water resources of portions of the North Platte River basin being designated as over-appropriated by the State of Nebraska Department of Natural Resources (DNR), the North Platte Natural Resources District (NPNRD), in cooperation with the DNR, is developing an Integrated Management Plan (IMP) for groundwater and surface water in the NPNRD. As part of the IMP, a three-dimensional numerical finite difference groundwater-flow model is being developed to evaluate the effectiveness of using leakage of water from selected irrigation canal systems to manage groundwater recharge. To determine the relative leakage potential of the upper 8 m of the selected irrigation canals within the North Platte River valley in western Nebraska and eastern Wyoming, the U.S. Geological Survey performed a land-based capacitively coupled (CC) resistivity survey along nearly 630 km of 13 canals and 2 laterals in 2004 and from 2007 to 2009. These 13 canals were selected from the 27 irrigation canals in the North Platte valley due to their location, size, irrigated area, and relation to the active North Platte valley flood plain and related paleochannels and terrace deposits where most of the saturated thickness in the alluvium exists. The resistivity data were then compared to continuous cores at 62 test holes down to a maximum depth of 8 m. Borehole electrical conductivity (EC) measurements at 36 of those test holes were done to correlate resistivity values with grain sizes in order to determine potential vertical leakage along the canals as recharge to the underlying alluvial aquifer. The data acquired in 2004, as well as the 25 test hole cores from 2004, are presented elsewhere. These data were reprocessed using the same updated processing and inversion algorithms used on the 2007 through 2009 datasets, providing a consistent and complete dataset for all collection periods. Thirty-seven test hole cores and borehole electrical conductivity measurements were acquired based on the 2008 data. This report presents comparisons between the CC resistivity data and results from the 37 test holes and includes all binned and inverted CC resistivity datasets from all four years as well as the EC log data for the 37 test holes acquired in 2008 and 2009. The information gained from these data can help State and local water managers and scientists better understand the characteristics of the shallow subsurface underlying the irrigation canals so that the water resources can be managed more effectively.

Scientific Investigations Report↗
USGS2010

Implementation and Evaluation of the Streamflow Statistics (StreamStats) Web Application for Computing Basin Characteristics and Flood Peaks in Illinois

Illinois StreamStats (ILSS) is a Web-based application for computing selected basin characteristics and flood-peak quantiles based on the most recently (2010) published (Soong and others, 2004) regional flood-frequency equations at any rural stream location in Illinois. Limited streamflow statistics including general statistics, flow durations, and base flows also are available for U.S. Geological Survey (USGS) streamflow-gaging stations. ILSS can be accessed on the Web at http://streamstats.usgs.gov/ by selecting the State Applications hyperlink and choosing Illinois from the pull-down menu. ILSS was implemented for Illinois by obtaining and projecting ancillary geographic information system (GIS) coverages; populating the StreamStats database with streamflow-gaging station data; hydroprocessing the 30-meter digital elevation model (DEM) for Illinois to conform to streams represented in the National Hydrographic Dataset 1:100,000 stream coverage; and customizing the Web-based Extensible Markup Language (XML) programs for computing basin characteristics for Illinois. The basin characteristics computed by ILSS then were compared to the basin characteristics used in the published study, and adjustments were applied to the XML algorithms for slope and basin length. Testing of ILSS was accomplished by comparing flood quantiles computed by ILSS at a an approximately random sample of 170 streamflow-gaging stations computed by ILSS with the published flood quantile estimates. Differences between the log-transformed flood quantiles were not statistically significant at the 95-percent confidence level for the State as a whole, nor by the regions determined by each equation, except for region 1, in the northwest corner of the State. In region 1, the average difference in flood quantile estimates ranged from 3.76 percent for the 2-year flood quantile to 4.27 percent for the 500-year flood quantile. The total number of stations in region 1 was small (21) and the mean difference is not large (less than one-tenth of the average prediction error for the regression-equation estimates). The sensitivity of the flood-quantile estimates to differences in the computed basin characteristics are determined and presented in tables. A test of usage consistency was conducted by having at least 7 new users compute flood quantile estimates at 27 locations. The average maximum deviation of the estimate from the mode value at each site was 1.31 percent after four mislocated sites were removed. A comparison of manual 100-year flood-quantile computations with ILSS at 34 sites indicated no statistically significant difference. ILSS appears to be an accurate, reliable, and effective tool for flood-quantile estimates.

Scientific Investigations Report↗
USGS2009

Regression models to estimate real-time concentrations of selected constituents in two tributaries to Lake Houston near Houston, Texas, 2005-07

In December 2005, the U.S. Geological Survey in cooperation with the City of Houston, Texas, began collecting discrete water-quality samples for nutrients, total organic carbon, bacteria (total coliform and Escherichia coli), atrazine, and suspended sediment at two U.S. Geological Survey streamflow-gaging stations upstream from Lake Houston near Houston (08068500 Spring Creek near Spring, Texas, and 08070200 East Fork San Jacinto River near New Caney, Texas). The data from the discrete water-quality samples collected during 2005-07, in conjunction with monitored real-time data already being collected - physical properties (specific conductance, pH, water temperature, turbidity, and dissolved oxygen), streamflow, and rainfall - were used to develop regression models for predicting water-quality constituent concentrations for inflows to Lake Houston. Rainfall data were obtained from a rain gage monitored by Harris County Homeland Security and Emergency Management and colocated with the Spring Creek station. The leaps and bounds algorithm was used to find the best subsets of possible regression models (minimum residual sum of squares for a given number of variables). The potential explanatory or predictive variables included discharge (streamflow), specific conductance, pH, water temperature, turbidity, dissolved oxygen, rainfall, and time (to account for seasonal variations inherent in some water-quality data). The response variables at each site were nitrite plus nitrate nitrogen, total phosphorus, organic carbon, Escherichia coli, atrazine, and suspended sediment. The explanatory variables provide easily measured quantities as a means to estimate concentrations of the various constituents under investigation, with accompanying estimates of measurement uncertainty. Each regression equation can be used to estimate concentrations of a given constituent in real time. In conjunction with estimated concentrations, constituent loads were estimated by multiplying the estimated concentration by the corresponding streamflow and applying the appropriate conversion factor. By computing loads from estimated constituent concentrations, a continuous record of estimated loads can be available for comparison to total maximum daily loads. The regression equations presented in this report are site specific to the Spring Creek and East Fork San Jacinto River streamflow-gaging stations; however, the methods that were developed and documented could be applied to other tributaries to Lake Houston for estimating real-time water-quality data for streams entering Lake Houston.

Scientific Investigations Report↗
USGS2011

Simulation of streamflow in the Pleasant, Narraguagus, Sheepscot, and Royal Rivers, Maine, using watershed models

The U.S. Geological Survey (USGS) began a study in 2008 to investigate anticipated changes in summer streamflows and stream temperatures in four coastal Maine river basins and the potential effects of those changes on populations of endangered Atlantic salmon. To achieve this purpose, it was necessary to characterize the quantity and timing of streamflow in these rivers by developing and evaluating a distributed-parameter watershed model for a part of each river basin by using the USGS Precipitation-Runoff Modeling System (PRMS). The GIS (geographic information system) Weasel, a USGS software application, was used to delineate the four study basins and their many subbasins, and to derive parameters for their geographic features. The models were calibrated using a four-step optimization procedure in which model output was evaluated against four datasets for calibrating solar radiation, potential evapotranspiration, annual and seasonal water balances, and daily streamflows. The calibration procedure involved thousands of model runs that used the USGS software application Luca (Let us calibrate). Luca uses the Shuffled Complex Evolution (SCE) global search algorithm to calibrate the model parameters. The calibrated watershed models performed satisfactorily, in that Nash-Sutcliffe efficiency (NSE) statistic values for the calibration periods ranged from 0.59 to 0.75 (on a scale of negative infinity to 1) and NSE statistic values for the evaluation periods ranged from 0.55 to 0.73. The calibrated watershed models simulate daily streamflow at many locations in each study basin. These models enable natural resources managers to characterize the timing and amount of streamflow in order to support a variety of water-resources efforts including water-quality calculations, assessments of water use, modeling of population dynamics and migration of Atlantic salmon, modeling and assessment of habitat, and simulation of anticipated changes to streamflow and water temperature resulting from changes forecast for air temperature and precipitation.

Scientific Investigations Report↗
USGS2011

Magnitude and Frequency of Floods for Urban and Small Rural Streams in Georgia, 2008

A study was conducted that updated methods for estimating the magnitude and frequency of floods in ungaged urban basins in Georgia that are not substantially affected by regulation or tidal fluctuations. Annual peak-flow data for urban streams from September 2008 were analyzed for 50 streamgaging stations (streamgages) in Georgia and 6 streamgages on adjacent urban streams in Florida and South Carolina having 10 or more years of data. Flood-frequency estimates were computed for the 56 urban streamgages by fitting logarithms of annual peak flows for each streamgage to a Pearson Type III distribution. Additionally, basin characteristics for the streamgages were computed by using a geographical information system and computer algorithms. Regional regression analysis, using generalized least-squares regression, was used to develop a set of equations for estimating flows with 50-, 20-, 10-, 4-, 2-, 1-, 0.5-, and 0.2-percent annual exceedance probabilities for ungaged urban basins in Georgia. In addition to the 56 urban streamgages, 171 rural streamgages were included in the regression analysis to maintain continuity between flood estimates for urban and rural basins as the basin characteristics pertaining to urbanization approach zero. Because 21 of the rural streamgages have drainage areas less than 1 square mile, the set of equations developed for this study can also be used for estimating small ungaged rural streams in Georgia. Flood-frequency estimates and basin characteristics for 227 streamgages were combined to form the final database used in the regional regression analysis. Four hydrologic regions were developed for Georgia. The final equations are functions of drainage area and percentage of impervious area for three of the regions and drainage area, percentage of developed land, and mean basin slope for the fourth region. Average standard errors of prediction for these regression equations range from 20.0 to 74.5 percent.

Georgia↗