Geology topics

Ryan R. McShane

Publications and source records attributed to Ryan R. McShane.

17 recordsLinked to original sources

USGS2026

Spatial and temporal drivers of variability in abundance of two stream fishes

An understanding of how and why animal populations vary in space and time informs effective conservation and management. We used spatial and temporal flow metrics to explore flow-ecology relationships for two fish species, Mountain Sucker Pantosteus platyrhynchus (Cope, 1874) and Mottled Sculpin Cottus bairdii Girard, 1850, across 64 sites and 8 years in Wyoming, U.S.A. We found variation in Mottled Sculpin and Mountain Sucker abundance was driven to a greater extent by spatial flow metrics (e.g., drainage area, median summer flow) with differing direction of response between species. The opposing responses to spatial variation in median summer flow could reflect the ability of Mountain Sucker to persist in non-perennial habitats, likely due to their relatively high mobility and recolonization capacity. Responses to temporal flow metrics (e.g., previous year spring flow, previous year autumn flow) were weaker but consistent across species likely due to similar reproductive strategies and timing. The lack of stronger flow effects on population variability in both species likely corresponds to high variability in local habitat characteristics that are mediating population response. These findings underscore how considering spatial context informs conservation planning and suggest that flow-ecology relationships may be mediated by species-specific traits and local habitat conditions.

Wyoming↗

USGS2026

Machine learning generated streamflow drought forecasts for the conterminous United States (CONUS): developing and evaluating an operational tool to enhance sub-seasonal to seasonal streamflow drought early warning for gaged locations

Forecasts of streamflow drought, when streamflow declines below typical levels, are notably less available than for floods or meteorological drought, despite widespread impacts. We apply machine learning (ML) models to forecast streamflow drought 1–13 weeks ahead at 3,219 streamgages across the conterminous United States. We applied two ML methods (Long short-term memory neural networks; Light Gradient-Boosting Machine) and two benchmark models (persistence; Autoregressive Integrated Moving Average) to predict weekly streamflow percentiles with independent models for each forecast horizon. ML models outperformed benchmarks in predicting continuous streamflow percentiles below 30%. ML models generally performed worse than persistence models for discrete classification (moderate, severe, extreme) but exceeded the benchmark models for drought onset/termination. Performance was better for less intense droughts and shorter horizons, with predictive power for 1–4 weeks for severe droughts (10% threshold). This work highlights challenges and opportunities to advance hydrological drought forecasting and supports a new experimental forecasting tool.

conterminous United States↗

USGS2025

Predicted fish vulnerability to stream drying in the western U.S.A.

The frequency, magnitude and extent of stream drying is increasing due to climate change and human water demand. Fish vulnerability to increased stream drying is a combination of sensitivity (innate tolerance to low streamflow) and exposure to stream drying. To understand fish tolerance to low flow and susceptibility to decline under changing streamflow conditions, we estimated species-specific measures of sensitivity to low streamflow, determined relationships to species traits and evaluated vulnerability to future reductions in streamflow for 60 species. We found that sensitivity varied across species, and some variation was explained by life history strategy, spawning strategy and body size. Periodic life history strategy, pelagic spawning and larger size corresponded to an increased sensitivity to stream drying. Under future projections of August streamflow, 90% of sites were predicted to decrease in flow magnitude. Vulnerability to changes in streamflow, the combination of sensitivity and exposure, varied slightly across the study species, with the percent of inhospitable sites under future climate scenarios increasing for 87% of the species. Despite being relatively insensitive to low streamflow, vulnerability was high for multiple species dominant in mountainous areas, driven by high levels of exposure to stream drying. Our results illustrate the potential for species traits to predict sensitivity to low streamflow and demonstrate that exposure may play a large role when defining species vulnerability to stream drying. The ability to predict species tolerances and susceptibility to decline will become increasingly important in prioritising conservation efforts under changing environmental conditions.

Idaho, Montana, Nebraska, North Dakota, South Dako↗

USGS2025

Machine learning generated streamflow drought forecasts for the Conterminous United States (CONUS): Developing and evaluating an operational tool to enhance sub-seasonal to seasonal streamflow drought early warning for gaged locations

Forecasts of streamflow drought, when streamflow declines below typical levels, are notably less available than for floods or meteorological drought, despite widespread impacts. To address this gap, we apply machine learning (ML) models to forecast streamflow drought 1-13 weeks into the future at > 3,000 streamgage locations across the conterminous United States (CONUS). We applied two ML methods (Long short-term memory (LSTM) neural networks; Light Gradient-Boosting Machine - LightGBM) and two benchmark model approaches (persistence; Autoregressive Integrated Moving Average - ARIMA) to predict weekly streamflow percentiles with independent models for each forecast horizon. To explore whether a training focus on dry weeks improved performance, both ML models were trained using all percentiles (LSTM-all, LightGBM-all) and only percentiles below 30% (LSTM<30, LightGBM<30). We evaluated model performance regionally and nationally for drought occurrence (the classification performance for a future date) and for drought onset/termination (performance identifying drought starts and ends). ML models generally performed worse than the persistence model for discrete classification (moderate, severe, extreme drought) of drought occurrence but exceeded the benchmark models for onset/termination. ML models outperformed benchmarks in predicting continuous streamflow percentiles below 30%. Occurrence performance was better for less intense droughts and shorter forecast horizons, with the ML models having predictive power at 1-4 week horizons for severe droughts (10th percentile threshold). All models struggled to forecast onset, though the best ML model was the LSTM<30 (sensitivity of 22%). Termination performance was greater, with the drought termination performance greatest for the LightGBM-all model. When estimating model uncertainty, the LSTM<30 model had the narrowest 90% percentile interval with closest to optimal capture. This work highlights the challenges and opportunities to further advance hydrological drought forecasting and supports an experimental operational streamflow drought assessment and forecast tool.

Continental United States↗

USGS2023

Regional streamflow drought forecasting in the Colorado River Basin using Deep Neural Network models

Process-based, large-scale (e.g., conterminous United States [CONUS]) hydrologic models have struggled to achieve reliable streamflow drought performance in arid regions and for low-flow periods. Deep learning has recently seen broad implementation in streamflow prediction and forecasting research projects throughout the world with performance often equaling or exceeding that of process-based models. Deep learning models are a possible approach to increase the accuracy of streamflow drought predictions and to expand the spatial coverage of river locations with available streamflow drought forecasts. As part of a multi-component Data-Driven Drought Prediction project, the U.S. Geological Survey is developing and testing deep learning models for streamflow drought forecasting. In this work, we present preliminary results of a deep learning model capable of predicting streamflow drought occurrence at ungaged locations for the Colorado River Basin (CRB). A long short-term memory (LSTM) neural network model was trained using 40 years (1980-2020) of daily streamflow data from 425 streamgages within and surrounding the CRB using static watershed attributes as well as meteorological and remotely sensed dynamic forcing inputs. Model tests were performed to evaluate model accuracy for now-casting streamflow drought conditions at ungaged locations and for forecasting drought conditions at lead times ranging from 0 to 14 days. Nearly all model configurations showed behavioral performance for predicting daily streamflow percentiles. Comparisons of LSTM model performance for predicting drought using fixed drought thresholds (calculated over all days and years) and variable drought thresholds (unique threshold calculated for each day of the year) identify differences in model skill between locations with implications for model design.

Colorado River basin↗

USGS2022

Predictions and drivers of sub-reach-scale annual streamflow permanence for the upper Missouri River basin: 1989-2018

The presence of year-round surface water in streams (i.e., streamflow permanence) is an important factor for identifying aquatic habitat availability, determining the regulatory status of streams, managing land use change, allocating water resources, and designing scientific studies. However, accurate, high resolution, and dynamic prediction of streamflow permanence that accounts for year-to-year variability at a regional extent is a major gap in modeling capability. Herein, we expand and adapt the U.S. Geological Survey (USGS) PRObability of Streamflow PERmanence (PROSPER) model from its original implementation in the Pacific Northwest (PROSPER PNW ) to the upper Missouri River basin (PROSPER UM ), a geographical region that includes mountain and prairie ecosystems of the northern United States. PROSPER UM is an empirical model used to estimate the probability that a stream channel has year-round flow in response to climatic conditions (monthly and annual) and static physiographic predictor variables of the upstream basin. The structure and approach of PROSPER UM are generally consistent with the PROSPER PNW model but include improved spatial resolution (10 m) and a longer modeling period. Average model accuracy was 81 %. Drainage area, upstream proportion as wetlands, and upstream proportion as developed land cover were the most important predictor variables. The PROSPER UM model identifies decreases in streamflow permanence during climatically drier years, although there is variability in the magnitude across basins highlighting geographically varying sensitivity to drought. Variability in the response of perennial streams to drought conditions among basins in the study area was also observed.

Montana, North Dakota, South Dakota, Wyoming↗

USGS2021

Estimates of water use associated with continuous oil and gas development in the Permian Basin, Texas and New Mexico, 2010–19, with comparisons to the Williston Basin, North Dakota and Montana

The Permian Basin, in west Texas and southeastern New Mexico is one of the largest conventional oil and gas reservoirs in the United States and is becoming one of the world’s largest continuous oil and gas (COG) reservoirs. Advances in technology have enabled oil and gas to be extracted from reservoirs that historically were developed using conventional, or vertical, well drilling techniques. Conventional oil and gas reservoirs have discrete deposits that are well defined and are typically trapped by an overlying geologic formation or caprock, whereas COG reservoirs contain deposits that are distributed evenly throughout the geologic formation, typically have much lower permeability (the capacity of a porous rock to transmit a fluid) than the conventional deposits, and require specialized horizontal extraction techniques. The methods to extract the oil from the two different reservoirs require differing amounts of water, and the horizontal extraction methods typically require substantially more water. In 2015, the U.S. Geological Survey started a topical study to quantify water used during COG development. The Permian Basin, which contains both types of reservoirs (continuous and conventional), was the second basin in the United States in the U.S. Geological Survey’s topical study to quantify water used during COG development.

Montana, New Mexico, North Dakota, Texas↗

USGS2021

A machine learning approach to modeling streamflow with sparse data in ungaged watersheds on the Wyoming Range, Wyoming, 2012–17

Scant availability of streamflow data can impede the utility of streamflow as a variable in ecological models of aquatic and terrestrial species, especially when studying small streams in watersheds that lack streamgages. Streamflow data at fine resolution and broad extent were needed by collaborators for ecological research on small streams in several ungaged watersheds of southwestern Wyoming, where streamflow data are sparse. To improve the utility of sparse streamflow data to ecological research in ungaged watersheds, we developed a machine learning approach in R for modeling spatially and temporally continuous monthly streamflow from 2012 through 2017 in three semiarid montane-steppe watersheds (with drainage areas of 26–55 square miles and mean elevations of 8,031–8,455 feet) on the Wyoming Range in the upper Green River Basin. A machine learning streamflow (MLFLOW) model was calibrated and validated with 971 discrete streamflow observations and 24 static and dynamic predictor variables derived from geospatial and time series data on climatic, physiographic, and anthropogenic characteristics affecting streamflow. The predictor variables were temporally and spatially conditioned to amplify the relation of predictor variables to monthly streamflow. The MLFLOW model had satisfactory agreement between observed and predicted streamflow (coefficient of determination [ R 2 ]=0.80, Nash-Sutcliffe efficiency [NSE]=0.79, NSE with log-transformed data [logNSE]=0.82, and percent bias [PBIAS]=0.7 percent). NSE and logNSE indicated the MLFLOW model performed equally well for high and low flows, and PBIAS indicated the MLFLOW model did not overpredict or underpredict monthly streamflow. Streamflow predictions seemed to well represent the annual hydrograph within the study area during the study period. The most important variables (statistically important in the MLFLOW model) for explaining monthly streamflow were temporally and spatially conditioned dynamic climatic variables, mostly precipitation and snow water equivalent. Importance of the static and dynamic variables did not differ substantially among the three watersheds but differed considerably among the 6 years. Monthly streamflow increased with increasing precipitation, snow water equivalent, and drainage area but decreased with increasing forest cover, elevation, evapotranspiration, and temperature. The MLFLOW model was most sensitive to selection of dynamic climatic variables. Unconditioned dynamic climatic variables alone explained 54 percent of the variance ( R 2 =0.54) in monthly streamflow, whereas adding static physiographic and anthropogenic variables only explained 12 percent more of the variance ( R 2 =0.66). Also, spatial conditioning of all variables together with temporal conditioning of dynamic variables increased the variance explained in the MLFLOW model by another 14 percent ( R 2 =0.80). The MLFLOW model also had greater sensitivity to temporal than to spatial differences in the data. For the MLFLOW model trained with observations from all watersheds and years or for models trained with observations from all except one watershed or 1 year left out sequentially, performance was better in testing on observations from each watershed than from each year separately. Also, performance was better for models fitted to fewer sites than to fewer months of observations. The greatest utility of the modeling approach is the ease of use and the speed of processing input data, running the model, and interpreting the model output, whereas the greatest limitation is the need for spatially and temporally representative streamflow observations to drive the model. Although familiarity with R is necessary, only a working knowledge of hydrology (for selecting appropriate predictor variables and evaluating the quality of streamflow observations) and a rudimentary understanding of machine learning models are needed. Therefore, this modeling approach is practicable for other scientists who work with water but who are not hydrologists.

Wyoming↗

USGS2021

Estimates of water use associated with continuous oil and gas development in the Permian Basin, Texas and New Mexico, 2010–19

In 2015, the U.S. Geological Survey started a topical study to quantify water use in areas of continuous oil and gas (COG) development. The first phase of the study was completed in 2019 and analyzed the Williston Basin. The second phase of the study analyzed the Permian Basin using the same techniques and approaches used for the Williston Basin analysis. The Permian Basin was selected for the second phase of water-use analysis for the following reasons: (1) the basin has the largest undiscovered technically recoverable oil and gas resource in the United States, (2) the basin has a continuous resource in tight shale that primarily produces oil, and (3) the basin is within the contiguous United States. This study used data from 60 counties in Texas and New Mexico with spatial coverage based on the Permian Basin extent defined by the U.S. Energy Information Administration, a representation of the geologically defined Permian Basin. Data from several sources were used in the analysis of direct, indirect, and ancillary water use associated with COG development in the Permian Basin and are available in an associated data release. Hydraulic fracturing water-use data were used to determine the start of the recent (before 2019) COG development boom in oil production in the Permian Basin in the same way that the data were used for the Williston Basin study. Water-use data were aggregated by county and year, which were the sampling units used in the analysis. The water-use analysis of the Permian Basin contained three elements: (1) estimates of water use, in million gallons, by county and year; (2) coefficients of water use from regression models, in million gallons per developed oil and gas well; and (3) performance (based on goodness-of-fit metrics) of the regression models in estimating the observed water use. Coefficients from the linear and quantile regression models of direct, indirect, and ancillary water use in the Permian Basin were produced as aggregate values for the counties and years. The mean estimate of direct water use had a 95-percent confidence interval of 4.13–5.45 million gallons (Mgal) per developed oil and gas well. The coefficient from the linear regression model of indirect water use was 0.111 Mgal per well, with a 95-percent confidence interval of 0.104–0.117 Mgal per well. The mean estimate of ancillary water use in the Permian Basin was 1.09 Mgal per well, with a 95-percent confidence interval of 1.05–1.13 Mgal per well. Model performance was evaluated with goodness-of-fit metrics including coefficient of determination ( R 2 ), root mean square error, and the ratio of root mean square error to standard deviation of observations computed from leave-one-out cross validation of the linear and quantile regression models of direct, indirect, and ancillary water use. Model performance for direct water use was acceptable, with an R 2 value of 0.91. The model performance of indirect water use was acceptable, with an R 2 value of 0.89. Values of R 2 for the ancillary water-use categories were at least 0.89. Annual mean estimates for hydraulic fracturing, cementing, drilling, indirect, and ancillary water use per well for the years 2010–17 were comparable between the Permian and Williston Basins. Hydraulic fracturing water use increased similarly from 2010 to 2015 in the Permian Basin and the Williston Basin, increasing from 0.6 Mgal per well in 2010 to 5.4 Mgal per well in 2015 in the Permian Basin and from 1.4 Mgal per well in 2010 to 4.7 Mgal per well in 2015 in the Williston Basin. By design, the Permian water-use assessment is a simplification of a complex and continually developing system and therefore has uncertainty and limitations in the interpretation of results. Despite the modeling limitations, the results summarized in the report, when compared to other studies, compare well with water-use estimations. The favorable comparison highlights the transferability of the water-use methodology to other areas of COG development.

New Mexico, Texas↗

USGS2021

U.S. Geological Survey science for the Wyoming Landscape Conservation Initiative—2018 annual report

The Wyoming Landscape Conservation Initiative (WLCI) was established in 2007 as a collaborative interagency partnership to develop and implement science-based conservation actions. During the past 11 years, partners from U.S. Geological Survey (USGS), State and Federal land management agencies, universities, and the public have collaborated to implement a long-term (more than 10 years) science-based program that assesses and enhances the quality and quantity of wildlife habitats in the southwest Wyoming region while facilitating responsible development. The USGS WLCI Science Team completes scientific research and develops tools that inform and support WLCI partner planning, decision making, and on-the-ground management actions. In fiscal year 2018, the USGS initiated 3 new projects and continued efforts on 21 ongoing science and web-development projects. The first new project was initiated to support Secretarial Order 3362 which calls on the USGS to assist Western States in mapping big-game migration corridors and developing new mapping tools. During 2018, the USGS hosted a workshop in Laramie, Wyoming, which included more than 70 State and Federal wildlife experts from Colorado, New Mexico, Texas, and Wyoming. Most of the mapping and migration tool curricula used in the workshop were derived from prior WLCI studies and mapping efforts of big-game migration movement in habitats undergoing large-scale energy development. The second new project was in response for WLCI partners to better understand sedimentation and hydrogeomorphic processes in a cold-desert headwater and the third new project was designed to improve our approach for people to access, manage, and analyze WLCI data and WLCI resource information. The USGS published 18 products (including peer-reviewed journal articles, USGS series publications, and data releases) and provided more than a dozen professional oral and poster presentations at scientific meetings and numerous informal presentations to WLCI partners at meetings and workshops. This report summarizes the objectives and status of each project and highlights the USGS 2018 accomplishments and products.

Wyoming↗

USGS2021

Beyond streamflow: Call for a national data repository of streamflow presence for streams and rivers in the United States

Observations of the presence or absence of surface water in streams are useful for characterizing streamflow permanence, which includes the frequency, duration, and spatial extent of surface flow in streams and rivers. Such data are particularly valuable for headwater streams, which comprise the vast majority of channel length in stream networks, are often non-perennial, and are frequently the most data deficient. Datasets of surface water presence exist across multiple data collection groups in the United States but are not well aligned for easy integration. Given the value of these data, a unified approach for organizing information on surface water presence and absence collected by diverse surveys would facilitate more effective and broad application of these data and address the gap in streamflow data in headwaters. In this paper, we highlight the numerous existing datasets on surface water presence in headwater streams, including recently developed crowdsourcing approaches. We identify the challenges of integrating multiple surface water presence/absence datasets that include differences in the definitions and categories of streamflow status, data collection method, spatial and temporal resolution, and accuracy of geographic location. Finally, we provide a list of critical and useful components that could be used to integrate different streamflow permanence datasets.

Water↗

USGS2020

Estimates of water use associated with continuous oil and gas development in the Williston Basin, North Dakota and Montana, 2007–17

This study of water use associated with development of continuous oil and gas resources in the Williston Basin is intended to provide a preliminary model-based analysis of water use in major regions of production of continuous oil and gas resources in the United States. Direct, indirect, and ancillary water use associated with development of continuous oil and gas resources in the Williston Basin was estimated in North Dakota and Montana from 2007 to 2017. Water-use data were aggregated by county and year, which were the sampling units used in this analysis. Linear and quantile regression models of water use in relation to the number of oil and gas wells developed were fit for the direct, indirect, and ancillary water-use categories for each State. A 95-percent confidence interval for each parameter estimate from the linear regression models was computed as a measure of uncertainty. Additional information on uncertainty can be gained from modeling other distribution parameters, so quantile regression models of the 5th, 50th, and 95th percentiles also were fit. To assess uncertainty in the estimates from the regression models of direct, indirect, and ancillary water use, leave-one-out cross-validation was used. Model performance was evaluated with three goodness-of-fit metrics used to compare the estimates and observations of water use. Mean annual direct and indirect water use for development of continuous oil and gas resources in North Dakota was estimated at 4,512 million gallons (Mgal) per year (Mgal/yr), with a 95-percent confidence interval of 4,021–5,152 Mgal/yr, and in Montana was estimated at 196 Mgal/yr, with a 95-percent confidence interval of 189–203 Mgal/yr. Ancillary water use (for domestic and public supply) had an estimated annual mean of 2,753 Mgal/yr in North Dakota and 396 Mgal/yr in Montana. The coefficient from the linear regression model of direct water use was 3.86 Mgal per well and hydraulic fracturing water use was 3.70 Mgal per well for North Dakota. The mean estimate of direct water use had a 95-percent confidence interval of 3.48–4.23 Mgal per well. For North Dakota, the coefficient from the linear regression model of indirect water use was 0.453 Mgal per well, with a 95-percent confidence interval of 0.415–0.492 Mgal per well. Direct and indirect water use had a mean estimate of about 4.31 Mgal per well in North Dakota. The mean estimate of ancillary water use (for domestic and public supply) in North Dakota was 2.03 Mgal per well, with a 95-percent confidence interval of 1.76–2.31 Mgal per well. For Montana, the linear regression model of hydraulic fracturing water use had a mean estimate of 2.04 Mgal per well. The 95-percent confidence interval for the mean estimate was 1.80–2.28 Mgal per well. Direct and indirect water use in Montana had a mean estimate of 2.49 Mgal per well. The mean estimate of ancillary water use (for domestic and public supply) in Montana was 2.43 Mgal per well, with a 95-percent confidence interval of 1.76–3.11 Mgal per well.

Montana, North Dakota, South Dakota↗

USGS2019

Analytical framework to estimate water use associated with continuous oil and gas development

An analytical framework was designed to estimate water use associated with continuous oil and gas (COG) development in support of the U.S. Geological Survey Water Availability and Use Science Program. This framework was developed to better understand the relation between the production of COG resources for energy and the amount of water needed to sustain this type of energy development in the United States. The total mean undiscovered, technically recoverable volume of COG has increased, highlighting the continued need to develop approaches to better characterize water use associated with COG development. The analytical framework can be used to estimate water use associated with COG development for three water-use components—direct, indirect, and ancillary water use—that are related to the life cycle of COG development. Direct water use is defined as water used in a wellbore to complete a well, including the water used for drilling, cementing, stimulating, and maintaining the well during production. Indirect water use is the water used at or near the well site, including water used for dust abatement, for cleaning equipment, and for crew and staff use. Ancillary water use is all other water used during the life cycle of COG development that is not categorized as direct or indirect, such as additional local or regional water use resulting from a change (for example, population) related to COG development. The analytical framework includes the data inputs, the processes involved in estimating the water-use coefficients and analyzing their uncertainties, and the outputs. The analytical framework was developed as an R script, which contains the statistical models used to estimate water-use components. The availability of data across COG reservoirs in the United States is variable and presents challenges for estimating water use for extracting COG from their reservoirs; thus, the R script can be modified for the types of data available within a COG reservoir, the extent and resolution of data available for each water-use component, and the desired output of the water-use assessment. The script was written so that the units of the data in the script were standardized. Water-use estimates were simulated for the mean and 10th, 50th, and 90th percentiles of the data distributions. Uncertainties were quantified with confidence intervals for the estimated coefficients. Uncertainty for estimated or simulated data can be calculated with the R script by providing a range of representative values that are within the appropriate confidence intervals of the mean of the data.

Lower 48 states↗

USGS2019

U.S. Geological Survey science for the Wyoming Landscape Conservation Initiative—2017 annual report

The Wyoming Landscape Conservation Initiative (WLCI) was established in 2008 to address the scientific and conservation questions associated with land use changes because of energy development and other factors in southwest Wyoming. Over the past decade, partners from U.S. Geological Survey (USGS), State and Federal land management agencies, universities, and the public have collaborated to implement a long-term (defined here as more than 10 years), science-based program that assesses and enhances the quality and quantity of wildlife habitats in this region while facilitating responsible development. The USGS Science Team completes scientific research and develops tools that inform and support WLCI partner planning, decision making, and on-the-ground management actions. In fiscal year 2017, USGS published 18 products (including peer-reviewed journal articles, USGS series publications, and data releases), prepared an additional 7 products for publication, and presented 14 talks or posters at professional scientific meetings in addition to numerous informal presentations to WLCI partners at meetings and workshops. In this report, we summarize the science themes that describe USGS science for the WLCI and highlight work completed in fiscal year 2017 for each science theme. We also provide information on how USGS science is being used by land managers to better achieve habitat conservation objectives.

Wyoming↗

USGS2019

Probability of streamflow permanence model (PROSPER): A spatially continuous model of annual streamflow permanence throughout the Pacific Northwest

The U.S. Geological Survey (USGS) has developed the PRObability of Streamflow PERmanence (PROSPER) model, a GIS raster-based empirical model that provides streamflow permanence probabilities (probabilistic predictions) of a stream channel having year-round flow for any unregulated and minimally-impaired stream channel in the Pacific Northwest region, U.S. The model provides annual predictions for 2004-2016 at a 30-m spatial resolution based on monthly or annually updated values of climatic conditions and static physiographic variables associated with the upstream basin. Predictions correspond to any pixel on the channel network consistent with the medium resolution National Hydrography Dataset channel network stream grid. Total annual precipitation and percent forest cover were consistently the most important predictor variables among global and most subregional models, which had error rates between 17 and 22%. Probabilities were converted to wet and dry streamflow permanence classes with an associated confidence. Wet and dry classifications were used to derive descriptors that characterize the statistical and spatial distribution of streamflow permanence in three focal basins. Predicted dry channel segments account for 52 to 92% of the stream network across the three focal basins; streamflow permanence decreased during climatically drier years. Predictions are publicly available through the USGS StreamStats platform. Results demonstrate the utility of the PROSPER model as a tool for identifying areas that may be resilient or sensitive to drought conditions, allowing for management efforts that target protecting critical reaches. Importantly, PROSPER’s successful predictive performance can be improved with new datasets of streamflow permanence underscoring the importance of field observations.

Idaho, Montana, Nevada, Oregon, Utah, Washington↗

USGS2018

Conceptual model to assess water use associated with the life cycle of unconventional oil and gas development

As the demand for energy increases in the United States, so does the demand for water used to produce many forms of that energy. Technological advances, limited access to conventional oil and gas accumulations, and the rise of oil and gas prices resulted in increased development of unconventional oil and gas (UOG) accumulations. Unconventional oil and gas is developed using a method that combines directional drilling and hydraulic fracturing techniques, allowing for greater oil and gas production from previously unrecoverable reservoirs. Quantification of the water resources required for UOG development and production is difficult because of disparate data sources, variable reporting requirements across boundaries (local, State, and national), and incomplete or proprietary datasets. A topical study was started in 2015 under the U.S. Geological Survey’s Water Availability and Use Science Program, as part of the directive in the Secure Water Act for the U.S. Geological Survey to conduct a National Water Census, to better understand the relation between production of UOG resources for energy and the amount of water needed to produce and sustain this type of energy development in the United States. The Water Availability and Use Science Program goal for this topical study is to develop and apply a statistical model to better estimate the water use associated with UOG development, regardless of the location and target geologic formation. As a first step, a conceptual model has been developed to characterize the life cycle of water use in areas of UOG development. Categories of water use and the way water-use data are collected might change over time; therefore, a generic approach was used in developing the conceptual model to allow for greater flexibility in adapting to future changes or newly available data. UOG development can be summarized into four stages: predrilling construction, drilling, hydraulic fracturing, and ongoing production. The water used in UOG production can be categorized further as direct, indirect, or ancillary water use. Direct water use is defined as the water used for drilling and hydraulic fracturing a well and for maintaining the well during ongoing production. Indirect water use is defined as the water used at or near a well pad. The water used for dust abatement also is considered an indirect use but may be applied away from the well pad. Ancillary water use is defined as the additional local or regional water use resulting from a change (for example, population) directly related to UOG development throughout the life cycle that is not used directly in the well or indirectly for any other purpose at the well pad. The conceptual model presented in this report consists of five elements: (1) input data, (2) processes, (3) decisions, (4) output data, and (5) outcomes. The input data requirements for estimating water use associated with UOG development are somewhat onerous, and obtaining suitable datasets can be challenging because local, State, and Federal agencies do not collect data similarly. The quality of a water-use assessment that uses the conceptual model presented in this report is dependent on the quality and quantity of data that are available for a UOG play. The conceptual model can be used for an assessment with sparse data; however, having sparse data likely will result in greater uncertainty in the water-use estimates. The conceptual model presented in this report is designed to be robust to characterize and simulate the data processing to estimate water use associated with UOG development. Although the results of an analysis that includes missing data have greater uncertainty, the analysis still can be insightful because it can establish a baseline estimate of UOG water use that may be refined further as more data become available. Analysis of models that include missing data also could aid in identifying the data most needed for future water-use estimates. Characterizing individual model limitations is important because the conceptual model can be used in future water-use studies to facilitate data compiling, data processing, estimating, and assessing UOG activities regardless of location.

Scientific Investigations Report↗

USGS2017

A review of surface energy balance models for estimating actual evapotranspiration with remote sensing at high spatiotemporal resolution over large extents

Many approaches have been developed for measuring or estimating actual evapotranspiration ( ET a ), and research over many years has led to the development of remote sensing methods that are reliably reproducible and effective in estimating ET a . Several remote sensing methods can be used to estimate ET a at the high spatial resolution of agricultural fields and the large extent of river basins. More complex remote sensing methods apply an analytical approach to ET a estimation using physically based models of varied complexity that require a combination of ground-based and remote sensing data, and are grounded in the theory behind the surface energy balance model. This report, funded through cooperation with the International Joint Commission, provides an overview of selected remote sensing methods used for estimating water consumed through ET a and focuses on Mapping Evapotranspiration at High Resolution with Internalized Calibration (METRIC) and Operational Simplified Surface Energy Balance (SSEBop), two energy balance models for estimating ET a that are currently applied successfully in the United States. The METRIC model can produce maps of ET a at high spatial resolution (30 meters using Landsat data) for specific areas smaller than several hundred square kilometers in extent, an improvement in practice over methods used more generally at larger scales. Many studies validating METRIC estimates of ET a against measurements from lysimeters have shown model accuracies on daily to seasonal time scales ranging from 85 to 95 percent. The METRIC model is accurate, but the greater complexity of METRIC results in greater data requirements, and the internalized calibration of METRIC leads to greater skill required for implementation. In contrast, SSEBop is a simpler model, having reduced data requirements and greater ease of implementation without a substantial loss of accuracy in estimating ET a . The SSEBop model has been used to produce maps of ET a over very large extents (the conterminous United States) using lower spatial resolution (1 kilometer) Moderate Resolution Imaging Spectroradiometer (MODIS) data. Model accuracies ranging from 80 to 95 percent on daily to annual time scales have been shown in numerous studies that validated ET a estimates from SSEBop against eddy covariance measurements. The METRIC and SSEBop models can incorporate low and high spatial resolution data from MODIS and Landsat, but the high spatiotemporal resolution of ET a estimates using Landsat data over large extents takes immense computing power. Cloud computing is providing an opportunity for processing an increasing amount of geospatial “big data” in a decreasing period of time. For example, Google Earth Engine TM has been used to implement METRIC with automated calibration for regional-scale estimates of ET a using Landsat data. The U.S. Geological Survey also is using Google Earth Engine TM to implement SSEBop for estimating ET a in the United States at a continental scale using Landsat data.

Scientific Investigations Report↗