Geology topics

Hayley R. Corson-Dosch

Publications and source records attributed to Hayley R. Corson-Dosch.

15 recordsLinked to original sources

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

Water supply in the conterminous United States, Alaska, Hawaii, and Puerto Rico, water years 2010–20

We present an assessment of water supply across the conterminous United States (CONUS), Alaska, Hawaii, and Puerto Rico covering water years 2010–20. Our analysis drew on two national hydrologic models, the National Hydrologic Model Precipitation-Runoff Modeling System and the Weather Research and Forecasting model hydrologic modeling system. Both models produced estimates of streamflow, evapotranspiration, soil moisture, snow water equivalent, and other hydrologic states and fluxes. The models were driven by the bias-adjusted 4-kilometer-resolution, long-term regional hydroclimate simulation over the conterminous United States dataset (CONUS404). We assessed spatial and temporal error distributions by comparing monthly simulations at the 12-digit hydrologic unit code and regional scale from both models against external benchmarking datasets. Results showed that average annual rainfall across the CONUS was 857 millimeters per year for the period of analysis, with water year 2012 the driest year (729 millimeters) and water year 2019 the wettest year (995 millimeters). Key interannual variability results included the following: (1) the California–Nevada hydrologic region had the highest variability in precipitation and snow accumulation, and (2) the Texas hydrologic region was among hydrologic regions with the highest variability in precipitation. We related interannual variability in precipitation to storage volumes in soil moisture, snow water equivalent, and lakes and reservoirs to highlight areas with little storage and large year-to-year variability in precipitation. These areas included the Southern High Plains, Central High Plains, Texas, Souris–Red–Rainy, Mississippi Embayment, and Midwest regions. Our analysis of groundwater-level data showed that several of these areas overlap aquifers where groundwater levels were considerably lower than historical averages, including the Colorado Plateaus aquifers, the Rio Grande aquifer system, and the Central and Southern regions of the High Plains aquifer. Many of these lowered groundwater levels are continuations of decades-long declines from overpumping that started well before the assessment period. The resulting water budgets and their analyses provide a high-resolution foundational assessment of the mean state and variability of the terrestrial hydrologic cycle across the CONUS and Alaska, Hawaii, and Puerto Rico to support a wide range of water resource management applications.

Professional Paper↗

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↗

USGS2024

Where is the water? Coast [poster]

Description Water is an integral part of how we interact with the environment and live our everyday lives. This educational poster illustrates where water is on the coast, how water moves, and different ways that water is used both naturally and through human interaction. This poster is intended for eighth-grade audiences and younger.

General Information Product↗

USGS2024

Where is the water? Urban [poster]

Description Water is an integral part of how we interact with the environment and live our everyday lives. This educational poster illustrates where water is in urban environments, how it moves, and different ways water is used both naturally and through human interaction. This poster is intended for eighth-grade audiences and younger.

General Information Product↗

USGS2024

Where is the water? Agriculture [poster]

Description Water is an integral part of how we interact with the environment and live our everyday lives. This educational poster illustrates agricultural water use, how water moves, and different ways that water is used both naturally and through human interaction. This poster is intended for eighth-grade audiences and younger.

General Information Product↗

USGS2024

Water cycle processes [poster]

Description The water cycle describes how water moves from Earth’s surface into the atmosphere, then back to the surface again or to below Earth’s surface. This educational poster depicts five key water-cycle processes that transport or transform water between states: evaporation, transpiration, condensation, precipitation, and infiltration. It illustrates examples of natural and human interactions with these processes. This poster is intended for eighth-grade audiences.

General Information Product↗

USGS2024

Where is the water? Suburban [poster]

Description Water is an integral part of how we interact with the environment and live our everyday lives. This educational poster illustrates where water is in the suburbs, how it moves, and different ways water is used both naturally and through human interaction. This poster is intended for eighth-grade audiences and younger.

General Information Product↗

USGS2024

Where is the water? Forest [poster]

Description Water is an integral part of how we interact with the environment and live our everyday lives. This educational poster illustrates where water is in forests, how it moves, and different ways water is used both naturally and through human interaction. This poster is intended for eighth-grade audiences and younger.

General Information Product↗

USGS2024

Where is the water? Desert [poster]

Description Water is an integral part of how we interact with the environment and live our everyday lives. This educational poster illustrates where water is in a desert environment, how it moves, and different ways water is used both naturally and through human interaction. This poster is intended for eighth-grade audiences and younger.

General Information Product↗

USGS2023

Evaluating deep learning architecture and data assimilation for improving water temperature forecasts at unmonitored locations

Deep learning (DL) models are increasingly used to forecast water quality variables for use in decision making. Ingesting recent observations of the forecasted variable has been shown to greatly increase model performance at monitored locations; however, observations are not collected at all locations, and methods are not yet well developed for DL models for optimally ingesting recent observations from other sites to inform focal sites. In this paper, we evaluate two different DL model structures, a long short-term memory neural network (LSTM) and a recurrent graph convolutional neural network (RGCN), both with and without data assimilation for forecasting daily maximum stream temperature 7 days into the future at monitored and unmonitored locations in a 70-segment stream network. All our DL models performed well when forecasting stream temperature as the root mean squared error (RMSE) across all models ranged from 2.03 to 2.11°C for 1-day lead times in the validation period, with substantially better performance at gaged locations (RMSE = 1.45–1.52°C) compared to ungaged locations (RMSE = 3.18–3.27°C). Forecast uncertainty characterization was near-perfect for gaged locations but all DL models were overconfident (i.e., uncertainty bounds too narrow) for ungaged locations. Our results show that the RGCN with data assimilation performed best for ungaged locations and especially at higher temperatures (>18°C) which is important for management decisions in our study location. This indicates that the networked model structure and data assimilation techniques may help borrow information from nearby monitored sites to improve forecasts at unmonitored locations. Results from this study can help guide DL modeling decisions when forecasting other important environmental variables.

Frontiers in Water↗

USGS2023

The water cycle

An illustrated diagram of the water cycle. This is a modern, updated version of the widely used diagram featured on the USGS Water Science School. Notably, this new water cycle diagram depicts humans and major categories of human water use as key components of the water cycle, in addition to the key pools and fluxes of the hydrologic cycle. This product targets an 8th grade audience and is designed to be printed as a poster.

General Information Product↗

USGS2023

Connecting habitat to species abundance: The role of light and temperature on the abundance of walleye in lakes

Walleye ( Sander vitreus) are an ecologically important species managed for recreational, tribal, and commercial harvest. Walleye prefer cool water and low light conditions, and therefore changing water temperature and clarity potentially impacts walleye habitat and populations across the landscape. Using survey data collected from 1993 to 2018 from 312 lakes in Minnesota, we evaluated the relationship between thermal-optical habitat and the relative abundance of small (0–300 mm), medium (300–450 mm), and large (450 + mm) walleye. Thermal-optical habitat was positively correlated with the relative abundance of small and medium walleye but not large walleye. Walleye were more abundant in larger, naturally reproducing lakes opposed to smaller, stocked lakes. Thermal-optical habitat changed in 59% of lakes since 1980 (26% increasing and 33% decreasing) and appears to be driven primarily by changes in water clarity and thus optical habitat area. Our study provides important insights into local and regional drivers that influence walleye populations that can be used to assist fisheries managers in setting population goals and managing harvest.

Minnesota↗

USGS2022

Near-term forecasts of stream temperature using deep learning and data assimilation in support of management decisions

Deep learning (DL) models are increasingly used to make accurate hindcasts of management-relevant variables, but they are less commonly used in forecasting applications. Data assimilation (DA) can be used for forecasts to leverage real-time observations, where the difference between model predictions and observations today is used to adjust the model to make better predictions tomorrow. In this use case, we developed a process-guided DL and DA approach to make 7-day probabilistic forecasts of daily maximum water temperature in the Delaware River Basin in support of water management decisions. Our modeling system produced forecasts of daily maximum water temperature with an average root mean squared error (RMSE) from 1.1 to 1.4°C for 1-day-ahead and 1.4 to 1.9°C for 7-day-ahead forecasts across all sites. The DA algorithm marginally improved forecast performance when compared with forecasts produced using the process-guided DL model alone (0%–14% lower RMSE with the DA algorithm). Across all sites and lead times, 65%–82% of observations were within 90% forecast confidence intervals, which allowed managers to anticipate probability of exceedances of ecologically relevant thresholds and aid in decisions about releasing reservoir water downstream. The flexibility of DL models shows promise for forecasting other important environmental variables and aid in decision-making.

Journal of the American Water Resources Associatio↗

USGS2021

Heterogeneous stream-reservoir graph networks with data assimilation

Accurate prediction of water temperature in streams is critical for monitoring and understanding biogeochemical and ecological processes in streams. Stream temperature is affected by weather patterns (such as solar radiation) and water flowing through the stream network. Additionally, stream temperature can be substantially affected by water releases from man-made reservoirs to downstream segments. In this paper, we propose a heterogeneous recurrent graph model to represent these interacting processes that underlie stream-reservoir networks and improve the prediction of water temperature in all river segments within a network. Because reservoir release data may be unavailable for certain reservoirs, we further develop a data assimilation mechanism to adjust the deep learning model states to correct for the prediction bias caused by reservoir releases. Our evaluation for the Delaware River Basin has demonstrated the superiority of our proposed method over multiple existing methods. We have extensively studied the effect of the data assimilation mechanism under different scenarios.

IEEE International Conference on Data Mining (ICDM↗