Information on water depth in river channels is important for a number of applications in water resource management but can be difficult to obtain via conventional field methods, particularly over large spatial extents and with the kind of frequency and regularity required to support monitoring programs. Remote sensing methods could provide a viable alternative means of mapping river bathymetry (i.e., water depth). The purpose of this study was to develop and test new, spectrally based techniques for estimating water depth from satellite image data. More specifically, a neural network-based temporal ensembling approach was evaluated in comparison to several other neural network depth retrieval (NNDR) algorithms. These methods are described in a manuscript titled "Neural Network-Based Temporal Ensembling of Water Depth Estimates Derived from SuperDove Images" and the purpose of this data release is to make available the depth maps produced using these techniques. The images used as input were acquired by the SuperDove cubesats comprising the PlanetScope constellation, but the original images cannot be redistributed due to licensing restrictions; the end products derived from these images are provided instead. The large number of cubesats in the PlanetScope constellation allows for frequent temporal coverage and the neural network-based approach takes advantage of this high density time series of information by estimating depth via one of four NNDR methods described in the manuscript: 1. Mean-spec: the images are averaged over time and the resulting mean image is used as input to the NNDR. 2. Mean-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is averaged to obtain the final depth map. 3. NN-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is then used as input to a second, ensembling neural network that essentially weights the depth estimates from the individual images so as to optimize the agreement between the image-derived depth estimates and field measurements of water depth used for training; the output from the ensembling neural network serves as the final depth map. 4. Optimal single image: a separate NNDR is applied independently to each image in the time series and only the image that yields the strongest agreement between the image-derived depth estimates and the field measurements of water depth used for training is used as the final depth map. MATLAB (Version 24.1, including the Deep Learning Toolbox) source code for performing this analysis is provided in the function NN_depth_ensembling.m and the figure included on this landing page provides a flow chart illustrating the four different neural network-based depth retrieval methods. As examples of the resulting models, MATLAB *.mat data files containing the best-performing neural network model for each site are provided below, along with a file that lists the PlanetScope image identifiers for the images that were used for each site. To develop and test this new NNDR approach, the method was applied to satellite images from three rivers across the U.S.: the American, Colorado, and Potomac. For each site, field measurements of water depth available through other data releases were used for training and validation. The depth maps produced via each of the four methods described above are provided as GeoTIFF files, with file name suffixes that indicate the method employed: X_mean-spec.tif, X_mean-depth.tif, X_NN-depth.tif, and X-single-image.tif, where X denotes the site name. The spatial resolution of the depth maps is 3 meters and the pixel values within each map are water depth estimates in units of meters.

Information on water depth in river channels is important for a number of applications in water resource management but can be difficult to obtain via conventional field methods, particularly over large spatial extents and with the kind of frequency and regularity required to support monitoring programs. Remote sensing methods could provide a viable alternative means of mapping river bathymetry (i.e., water depth). The purpose of this study was to develop and test new, spectrally based techniques for estimating water depth from satellite image data. More specifically, a neural network-based temporal ensembling approach was evaluated in comparison to several other neural network depth retrieval (NNDR) algorithms. These methods are described in a manuscript titled "Neural Network-Based Temporal Ensembling of Water Depth Estimates Derived from SuperDove Images" and the purpose of this data release is to make available the depth maps produced using these techniques. The images used as input were acquired by the SuperDove cubesats comprising the PlanetScope constellation, but the original images cannot be redistributed due to licensing restrictions; the end products derived from these images are provided instead. The large number of cubesats in the PlanetScope constellation allows for frequent temporal coverage and the neural network-based approach takes advantage of this high density time series of information by estimating depth via one of four NNDR methods described in the manuscript: 1. Mean-spec: the images are averaged over time and the resulting mean image is used as input to the NNDR. 2. Mean-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is averaged to obtain the final depth map. 3. NN-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is then used as input to a second, ensembling neural network that essentially weights the depth estimates from the individual images so as to optimize the agreement between the image-derived depth estimates and field measurements of water depth used for training; the output from the ensembling neural network serves as the final depth map. 4. Optimal single image: a separate NNDR is applied independently to each image in the time series and only the image that yields the strongest agreement between the image-derived depth estimates and the field measurements of water depth used for training is used as the final depth map. MATLAB (Version 24.1, including the Deep Learning Toolbox) source code for performing this analysis is provided in the function NN_depth_ensembling.m available on the main landing page for the data release of which this is a child item, along with a flow chart illustrating the four different neural network-based depth retrieval methods. To develop and test this new NNDR approach, the method was applied to satellite images from the Colorado River near Lees Ferry, AZ, acquired in March and April of 2021. Field measurements of water depth available through another data release (Legleiter, C.J., Debenedetto, G.P., and Forbes, B.T., 2022, Field measurements of water depth from the Colorado River near Lees Ferry, AZ, March 16-18, 2021: U.S. Geological Survey data release, https://doi.org/10.5066/P9HZL7BZ) were used for training and validation. The depth maps produced via each of the four methods described above are provided as GeoTIFF files, with file name suffixes that indicate the method employed: Colorado_mean-spec.tif, Colorado_mean-depth.tif, Colorado_NN-depth.tif, and Colorado-single-image.tif. In addition, to assess the robustness of the Mean-spec and NN-depth methods to the introduction of a large pulse of sediment by a flood event that occurred partway through the image time series, depth maps from before and after the flood are provided in the files Colorado_Mean-spec_after_flood.tif,

Information on water depth in river channels is important for a number of applications in water resource management but can be difficult to obtain via conventional field methods, particularly over large spatial extents and with the kind of frequency and regularity required to support monitoring programs. Remote sensing methods could provide a viable alternative means of mapping river bathymetry (i.e., water depth). The purpose of this study was to develop and test new, spectrally based techniques for estimating water depth from satellite image data. More specifically, a neural network-based temporal ensembling approach was evaluated in comparison to several other neural network depth retrieval (NNDR) algorithms. These methods are described in a manuscript titled "Neural Network-Based Temporal Ensembling of Water Depth Estimates Derived from SuperDove Images" and the purpose of this data release is to make available the depth maps produced using these techniques. The images used as input were acquired by the SuperDove cubesats comprising the PlanetScope constellation, but the original images cannot be redistributed due to licensing restrictions; the end products derived from these images are provided instead. The large number of cubesats in the PlanetScope constellation allows for frequent temporal coverage and the neural network-based approach takes advantage of this high density time series of information by estimating depth via one of four NNDR methods described in the manuscript: 1. Mean-spec: the images are averaged over time and the resulting mean image is used as input to the NNDR. 2. Mean-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is averaged to obtain the final depth map. 3. NN-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is then used as input to a second, ensembling neural network that essentially weights the depth estimates from the individual images so as to optimize the agreement between the image-derived depth estimates and field measurements of water depth used for training; the output from the ensembling neural network serves as the final depth map. 4. Optimal single image: a separate NNDR is applied independently to each image in the time series and only the image that yields the strongest agreement between the image-derived depth estimates and the field measurements of water depth used for training is used as the final depth map. MATLAB (Version 24.1, including the Deep Learning Toolbox) source code for performing this analysis is provided in the function NN_depth_ensembling.m available on the main landing page for the data release of which this is a child item, along with a flow chart illustrating the four different neural network-based depth retrieval methods. To develop and test this new NNDR approach, the method was applied to satellite images from the Colorado River near Lees Ferry, AZ, acquired in March and April of 2021. Field measurements of water depth available through another data release (Legleiter, C.J., Debenedetto, G.P., and Forbes, B.T., 2022, Field measurements of water depth from the Colorado River near Lees Ferry, AZ, March 16-18, 2021: U.S. Geological Survey data release, https://doi.org/10.5066/P9HZL7BZ) were used for training and validation. The depth maps produced via each of the four methods described above are provided as GeoTIFF files, with file name suffixes that indicate the method employed: Colorado_mean-spec.tif, Colorado_mean-depth.tif, Colorado_NN-depth.tif, and Colorado-single-image.tif. In addition, to assess the robustness of the Mean-spec and NN-depth methods to the introduction of a large pulse of sediment by a flood event that occurred partway through the image time series, depth maps from before and after the flood are provided in the files Colorado_Mean-spec_after_flood.tif,

Information on water depth in river channels is important for a number of applications in water resource management but can be difficult to obtain via conventional field methods, particularly over large spatial extents and with the kind of frequency and regularity required to support monitoring programs. Remote sensing methods could provide a viable alternative means of mapping river bathymetry (i.e., water depth). The purpose of this study was to develop and test new, spectrally based techniques for estimating water depth from satellite image data. More specifically, a neural network-based temporal ensembling approach was evaluated in comparison to several other neural network depth retrieval (NNDR) algorithms. These methods are described in a manuscript titled "Neural Network-Based Temporal Ensembling of Water Depth Estimates Derived from SuperDove Images" and the purpose of this data release is to make available the depth maps produced using these techniques. The images used as input were acquired by the SuperDove cubesats comprising the PlanetScope constellation, but the original images cannot be redistributed due to licensing restrictions; the end products derived from these images are provided instead. The large number of cubesats in the PlanetScope constellation allows for frequent temporal coverage and the neural network-based approach takes advantage of this high density time series of information by estimating depth via one of four NNDR methods described in the manuscript: 1. Mean-spec: the images are averaged over time and the resulting mean image is used as input to the NNDR. 2. Mean-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is averaged to obtain the final depth map. 3. NN-depth: a separate NNDR is applied independently to each image in the time series and the resulting time series of depth estimates is then used as input to a second, ensembling neural network that essentially weights the depth estimates from the individual images so as to optimize the agreement between the image-derived depth estimates and field measurements of water depth used for training; the output from the ensembling neural network serves as the final depth map. 4. Optimal single image: a separate NNDR is applied independently to each image in the time series and only the image that yields the strongest agreement between the image-derived depth estimates and the field measurements of water depth used for training is used as the final depth map. MATLAB (Version 24.1, including the Deep Learning Toolbox) for performing this analysis is provided in the function NN_depth_ensembling.m available on the main landing page for the data release of which this is a child item, along with a flow chart illustrating the four different neural network-based depth retrieval methods. To develop and test this new NNDR approach, the method was applied to satellite images from the American River near Fair Oaks, CA, acquired in October 2020. Field measurements of water depth available through another data release (Legleiter, C.J., and Harrison, L.R., 2022, Field measurements of water depth from the American River near Fair Oaks, CA, October 19-21, 2020: U.S. Geological Survey data release, https://doi.org/10.5066/P92PNWE5) were used for training and validation. The depth maps produced via each of the four methods described above are provided as GeoTIFF files, with file name suffixes that indicate the method employed: American_mean-spec.tif, American_mean-depth.tif, American_NN-depth.tif, and American-single-image.tif. The spatial resolution of the depth maps is 3 meters and the pixel values within each map are water depth estimates in units of meters.

Field measurements of water depth were acquired from a reach of the Colorado River near Lees Ferry, Arizona, March16-18, 2021, to support research on remote sensing of water depth from satellite images. The depth measurements included in this data release were obtained along a series of cross-sections using a SonTek RiverSurveyor M9 acoustic Doppler current profiler (ADCP) deployed from a boat. The spatial location of each measurement was obtained using a differential GPS included as part of the RiverSurveyor M9 ADCP instrument package. The map projection and datum for these data are UTM Zone 12S and WGS84, respectively. The USGS Qrev software program was used to ingest and process the raw ADCP data. The Qrev data file was then imported into MATLAB, which was used to create a comma-delimited (*.csv) text file with three columns: Easting_m, Northing_m, and Depth_m; the units of the spatial coordinates and the depths are meters. This ground-based depth data set was used to calibrate (i.e., train) models for inferring water depths from passive optical image data and to assess the accuracy of image-derived depth estimates.

Field-based real-time kinematic (RTK) GNSS surveys of water surface elevation and channel bed topography were collected along the Colorado River, focusing on two cross-sections from which remotely sensed data were obtained. These data were used to assess the accuracy of river bed elevations inferred from the ASTRALiTe bathymetric lidar, acquired from an unmanned aircraft system (UAS). These data sets were collected to support research focused on developing innovative methods for non-contact measurement of river discharge based on various forms of remotely sensed data. The RTK GNSS surveys were performed using a local base station and Trimble R8 and R10 receivers while wading the channel at each cross-section. The survey data were post-processed by performing an OPUS correction of the static observations collected by the base and adjusting all of the survey points accordingly. The survey data were exported to a comma-separated text file and the resulting *.csv file contain for each point number a point ID, spatial coordinates, elevations, and a depth.

This dataset contains river depth, water velocity, and direction data at three focal areas of research included in the Sustainable Rivers Program (SRP) in the Allegheny River on three dates in 2021 and 2022, at two different river conditions (10,500/10,800 cubic feet per second (cfs) and 6,600 cfs as recorded at USGS gage Allegheny River at West Hickory - 03016000). Data was collected with a Sontek M9* Acoustic Doppler Current Profiler (ADCP) and processed within the USGS software QREV (Mueller, 2020). Depth averaged data are provided at intervals across the river along with data at various depths. Data was collected in the same general area during the three collection periods, but the cross-section locations do not overlap. Data Form Summary: Data is given in station number named .zip folders. Each .zip folder contains folders of data collected on each date (YYYYMMDD). For station numbers 03015650 and 03015950, data was collected at two spots and that data is separated as well. See "Transect Details" for location designation specifics. Each date_LOCATION DESIGNATION folder consists of Excel files that contain data processed within QREV. Output includes a point location, depth, water current velocity, and water current direction at that location. See each individual metadata file included in the zip file for details. This data can be processed in Geographic Information System (GIS) software for visualization purposes. * Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

U.S. Geological Survey (USGS) and Virginia Institute of Marine Science (VIMS) scientists conducted field data collection efforts during the week of April 8th - 14th, 2018, using a combination of remote sensing technologies to map riverbank and wetland topography and vegetation at four sites in the Chesapeake Bay Region of Virginia. The four sites are located along the James, Severn, and York Rivers. The work was initiated to evaluate the utility of different remote sensing technologies in mapping river bluff and wetland topography and vegetation for change detection and sediment transport modeling. The USGS team collected Global Navigation Satellite System (GNSS), total station, and ground based lidar (GBL) data while the VIMS team collected aerial imagery using an Unmanned Aerial System (UAS). This data release contains shapefiles of the processed GNSS and total station data, point clouds in the form of lidar data exchange (las) files from the ground lidar data and aerial imagery produced via Structure from Motion (SfM).

Bathymetric LiDAR technology was used to collect riverbed elevation data along the Potomac River. In support of this effort, a bathymetric survey with a boat-mounted acoustic Doppler current profiler (ADCP) was conducted in the study area during October 4-7, 2021. The study area consisted of four verification reaches on the Potomac River including: 1) Williamsport accessed through the Williamsport Park boat ramp below Conococheague Creek and RTE 11 (Williamsport), 2) Big Slackwater above C&O Canal Dam #4 accessed through the Big Slackwater Boat Ramp (Dam4), 3) Four Locks above C&O Canal Dam #5 accessed through the Four Locks Boat Ramp (Dam5), and 4) Little Tonoloway Recreation Area accessed through the Hancock Boat Ramp below RTE 522. Global Navigational Satellite Systems (GNSS) were used to concurrently collect survey grade real-time kinematic (RTK) horizontal and vertical coordinates of the ADCP transducer face. The riverbed elevations were collected using the ADCP with WinRiverII to export for post-processing in Microsoft Excel and RStudio. The GNSS equipment was programmed to continuously collect an observation every 1 to 2 seconds and the ADCP was programmed to continuously collect an observation every 1 second to 2 seconds. The corrected depths from the 4 ADCP beams were averaged and then subtracted from the GNSS derived elevation of the ADCP transducer face to compute the elevation of the riverbed. All spatial data is referenced horizontally to the North American Datum of 1983 (2011) and vertically to the North American Vertical Datum of 1988 (NAVD88). Grid coordinates are projected in Universal Transverse Mercator Zone 18 North and are represented in meter units. This data release consists of four (4) comma-delimited (*.csv) files with fifteen columns each: GNSS_ID, Time_hh_mm_ss, GNSS_Northing_M, GNSS_Easting_M, Computed_Elevation_M, GNSS_Transducer_Elevation_M, Computed_Mean_Depth_M, GNSS_PDOP, GNSS_Vertical Precision_M, GNSS_Satellites, ADCP_Ensemble_ID, ADCP_Temp_C, ADCP_Pitch_Degrees, ADCP_Roll_Degrees, and Type. This data release supersedes a previous version (https://doi.org/10.5066/P9EA0IKM) which contained a constant error of +0.344 meters in the GNSS antenna height reference elevations.