Amanda R. Whaling Richard R. McDonald Travis M. Knight Jessica Z. LeRoy Andrea L. Creighton 20260116 v1.2.0 vector digital data https://doi.org/10.5066/P13PPXKN This data release presents a national cross-section (XS) database (NXSDB) of bathymetric cross-sections collected during water year (WY) 2023 (October 1, 2022, to September 30, 2023). The WY 2023 NXSDB has been approved by the U.S. Geological Survey (USGS) for publication and supersedes the provisional version. The data were collected by hydrographers during routine site visits to thousands of USGS streamgaging stations for river discharge measurements. The database comprises approximately 64,500 individual cross-sections from 8,556 stations across the United States and is provided as an Open Geospatial Consortium (OGC) GeoPackage file named NXSDB_WY2023_CONUS.gpkg. A rendered HTML README documentation file (README_Documentation.html) accompanies the release and provides essential context for data users, including information on the database’s development, structure, processing assumptions, and data inclusion criteria. The cross-section data have undergone automated quality checks to identify missing values, structural inconsistencies, and extreme outliers. However, unlike finalized discharge measurements published in the National Water Information System (NWIS), these data have not received the same level of manual review. Although approved for release by the USGS, no expressed or implied warranty is made regarding their use, display, or suitability for other systems or for general or scientific applications. Users are encouraged to consult the accompanying README to understand the quality control procedures and data cleaning methods applied prior to database integration, as well as the assumptions involved. This understanding is essential for appropriately interpreting the data and assessing its suitability for specific analytical purposes. The README also describes the naming conventions used in the WY 2023 database, which is organized into table layers within the GeoPackage. The file contains three main types of layers. The "geometry" layer is the only one that includes geospatial information and contains georeferenced and hydrographically addressed points representing station locations, along with summary information on the total number of cross-section measurements. Each station has a corresponding "record" layer that summarizes the available cross-section measurements. The third main layer type is the "measurement" layer, which represents individual cross-sections. This layer has two subtypes corresponding to the primary methodologies used to collect discharge measurements: midsection-method measurements, which are not georeferenced, and moving-boat acoustic Doppler current profiler (ADCP) measurements, which are sometimes georeferenced with real-world spatial coordinates (latitude and longitude). Due to the size of the database, programmatic access is recommended for efficient querying and analysis. To support users, demonstration packages in both R and Python are included in demo.zip. These packages contain custom functions designed to simplify interaction with the GeoPackage, since standard methods in these languages may be inefficient for files of this size. The database follows a structured versioning system aligned to changes the custom schema provided with this data release (NXSDB_schema_v1.2.0.json), which defines and validates the GeoPackage structure. This approved version of the database (v1.2.0) supersedes the previous version, which was provisionally published to incorporate improvements based on user feedback. A detailed record of changes is provided in the accompanying change_log.txt file. The previous version remains accessible as a tagged instance within a ZIP package so users can continue to access it if needed. The station master file (NXSDB_WY2023_CONUS_station_master.txt) lists all stations evaluated for inclusion. Stations that appear in this list but are absent from the GeoPackage had no cross-sections meeting the database’s inclusion criteria. Further processing details are available in the accompanying metadata XML file and README. In 2022, the National Hydrologic Geospatial Fabric (NHGF) initiated a project to retrieve cross-section data from Aquarius, the U.S. Geological Survey’s internal platform for managing hydrologic time-series data. The primary goal of this effort is to make these cross-sections publicly available in a standardized database to support a wide range of hydrologic modeling applications. Users are encouraged to consult the accompanying README, which describes the quality-control procedures and data-cleaning methods applied prior to database integration, as well as the assumptions made. Reviewing this information is essential for appropriately interpreting the data and evaluating its suitability for specific analytical purposes. 20221001 20230930 ground condition Complete As needed -165.5100 -67.3181 70.2814 19.7121 ISO 19115 Topic Category geoscientificInformation inlandWaters transportation USGS Thesaurus bathymetry bathymetry measurement river reaches monitoring Geomorphology Hydraulic Geometry None cross-section cross section channel depth channel width cross-sectional area USGS Metadata Identifier USGS:689107cfd4be02693f467a60 Common geographic areas North America United States None. Please see 'Distribution Info' for details. The cross-section data have undergone automated quality checks to identify missing values, structural inconsistencies, and extreme outliers. However, unlike finalized discharge measurements published in the National Water Information System (NWIS), these data have not received the same level of manual review. Although approved for release by the U.S. Geological Survey (USGS), no expressed or implied warranty is made regarding their use, display, or suitability on other systems or for general or scientific applications. Users are encouraged to consult the accompanying README_Docementation.html to understand the quality control procedures and data cleaning methods applied prior to database integration, as well as the assumptions involved. This understanding is essential for appropriately interpreting the data and assessing its suitability for specific analytical purposes. Travis M. Knight U.S. Geological Survey, Water Resources Mission Area Hydrologist mailing and physical

Suite 704

Arden NC 28704 US 479-442-4888 x212 tknight@usgs.gov These data were collected at USGS streamgaging stations by dedicated hydrographers across the United States. Their commitment to measuring streamflow under challenging and remote conditions supports the near real-time delivery of discharge, water-level, and cross-section information included in this database. Amanda Whaling (Lower Mississippi-Gulf Water Science Center) drafted documentation, designed the database schema, created demonstration notebooks, and developed the software for GeoPackage construction. Richard McDonald (Water Resources Mission Area) developed software to pull Aquarius gage record attachments, extract cross-sections, and export results to various file formats. Travis Knight (Hydrologic Networks Branch) provided critical guidance on discharge measurement procedures and contributed QRev software updates that enabled the extraction of moving-boat cross sections. All authors of the NXSDB contributed to developing the conceptual framework of the database. In addition, Carl Legleiter (Water Resources Mission Area) provided valuable feedback as a database user, which led to clearer documentation and greater consistency for the benefit of all end users. Ubuntu 64-bit on Windows Subsystem for Linux 2 (WSL2); python 3.10.12 D. Phil Turnipseed Vernon B. Sauer 2010 publication Reston, VA US Geological Survey https://doi.org/10.3133/tm3A8 Timothy O Hodson Jayaram A Hariharan 2023 dataset https://www.sciencebase.gov U.S. Geological Survey https://doi.org/10.5066/P94I5TX3 Taher Chegini Hong-Yi Li L. Leung 2021 publication Journal of Open Source Software vol. 6, issue 66 ppg. 3175 https://doi.org/10.21105/joss.03175 David S Mueller 2020 dataset https://www.sciencebase.gov U.S. Geological Survey https://doi.org/10.5066/p9oz8qdl Open Geospatial Consortium (OGC) 2023 publication https://docs.ogc.org/is/12-128r19/12-128r19.html The cross-section data were not subject to accuracy tests and are provided as-is, without verification of positional or vertical accuracy. Cross-section measurements were processed using proprietary software (e.g., WinRiver, RiverSurveyor, SxS Pro, FlowTracker) for midsection measurements or USGS QRev software for moving-boat ADCP measurements. Valid measurements were identified by matching timestamps with published NWIS discharge values. Orientation consistency was ensured for midsection measurements, and georeferencing checks were performed for moving-boat ADCP measurements. See processing steps for more details. NXSDB data were accessed via the internal Aquarius Time Series API and processed for this standardized database; values should not be expected to exactly match the original source data. Data availability in the NXSDB reflects access at the time of download, so users should be aware that subsequent data queries could return different or additional values. Cross-sections with all missing data or not associated with a discharge measurement published in NWIS were removed. Some cross-sections may have missing data values. For attributes where missing or empty values are allowed in the GeoPackage, a "NULL" is listed in the sqlite_type field of the JSON schema included with this data release. Most cross-sections are not georeferenced. Additional checks were applied to georeferenced moving-boat measurements, however no tests were conducted to verify horizontal accuracy. The discharge measurement processing metadata was required to indicate that data from the GNSS antenna had been used as the navigation reference. In addition, all cross-section points were required to fall within a 10-mile radius of the station location. If either of these checks failed, the georeferencing data were omitted from the cross-section table, although depth and distance values were retained. All cross-sections in the database are not vertically referenced. The accuracy of depth measurements depends on the acoustic or mechanical methods used during the collection of the discharge measurement. A hydrographer makes a discharge measurement using either the midsection method or the moving-boat Acoustic Doppler Current Profiler (ADCP) method (Turnipseed and Sauer, 2010). 20221001 The hydrographer processes the measurement using proprietary software associated with the instrument used (e.g., WinRiver, RiverSurveyor, SxS Pro, FlowTracker) for measurements made using the midsection method, or uses USGS QRev software for measurements made using the moving-boat ADCP method. 20221001 The processed discharge data is loaded into SVMAQ (Site Visit Mobile Aquarius), generating an XML file that can be combined with field notes collected in SVMAQ. 20221001 The output file from SVMAQ is uploaded to AQUARIUS Time-Series, the internal USGS platform for storing and processing hydrologic time-series data. 20221001 The discharge value and any rating curve shifts undergo a thorough review process. The discharge measurement values are published to the National Water Information System (NWIS), but the underlying cross-section data remains in Aquarius and is not publicly accessible. 20221001 USGS Water Services (https://waterservices.usgs.gov/) was used to generate a list of potential streamgage station IDs there were active in water year 2023. The dataretrieval (Hodson and others, 2023) and HyRiver (Chegini and others, 2021) Python packages were used to add station metadata and hydrographic addressing information, respectively, for each station. The resulting station master table, 'NXSDB_WY2023_CONUS_station_master.txt', is provided with this data release. Please see the 'demo.zip' archive for R and Python functions for efficiently reading the text file. 20230101 Python is used to programmatically access and download the cross-section data from the XML output from SVMAQ for all stations in the generated station list ('NXSDB_WY2023_CONUS_station_master.txt'), compiling the results into two JSON files organized by station, one for measurements made with the midsection method, and the other for measurements made with the moving-boat ADCP method. 20250801 A Python script processes the stations listed in the station list by parsing the output JSON files into station dictionaries. The first step in building the station dictionary is to check measurement entries for each station and from each JSON file for valid cross-sections. A valid cross-section must (1) contain data in the cross-section table within the measurement entry and (2) have a measurement timestamp that corresponds to a discharge value published in NWIS. To support this validation, discharge measurements are retrieved from NWIS using the get_discharge_measurements() function in the Python dataRetrieval package (Hodson and Hariharan, 2023). The timestamps ("mmt_Timestamp") recorded in the cross-section metadata are then compared to the NWIS records, and the corresponding discharge value is stored in the record table as "NWIS_Discharge_cfs". This safeguard prevents potentially unreliable cross-sections from being included in the database and resulted in the exclusion of a relatively small number of measurements (approximately 600) where discharge values were not available in NWIS. Additional validation checks exclude measurements that lack measurable channel width or cross-sectional area values, as determined by the "NXSDB_Width_ft" and "NXSDB_Area_sf" attributes in the record layer. In most cases, these failures were related to having fewer than two valid distance and depth points within the cross-section. 20250925 After validating the cross-sections, the second step in building the station dictionary is cleaning any remaining valid cross-section tables and adding the cleaned table to the station dictionary. For midsection measurements, the cleaning involves reversing the orientation of profiles collected by the hydrographer that were initiated on the right side of the stream (when viewed facing downstream, in the direction the water is flowing) and ended on the left; reversing right-oriented midsection measurements ensures all cross-sections in the final database are provided in the same left-to-right orientation. For moving-boat ADCP measurements, the mean cross-section output from QRev (Mueller, 2020) is always oriented from left to right. Cleaning for these measurements involves removing invalid NULL depths, projecting profile endpoints outward by half a grid-cell to ensure consistent width calculations while retaining original depth values, discarding the original grid-centered endpoints, and converting all distances and depths from meters to feet. If latitude and longitude coordinates are present, the new endpoints are recalculated, and the data undergoes a completeness check and a spatial verification to ensure the cross-section points fall within 10 miles of the station location. If either check fails, the latitude and longitude columns are removed, though the depth and distance profile is retained; if both pass, the measurement is flagged as a georeferenced cross-section. 20250925 The third step in building the station dictionary is creating a table summarizing the metadata for each measurement and adding it to the station dictionary. The resulting table includes site visit information noted by the hydrographer in SVMAQ. 20250925 The fourth step involved in making the station dictionary is generating station summary information based on the validated and cleaned cross-sections in the station dictionary and adding the results to the station dictionary. The station information includes the first and last cross-section timestamps, cross-section totals, and the number of georeferenced cross-sections for that station. 20250925 A Python script reads and merges station dictionaries created from each JSON file, if available, as some stations might have measurements from both methods. These merged dictionaries are used to create a final GeoPackage with three types of tables or layers, each assigned specific names. First, station summary information from the dictionaries is merged with station metadata stored in the 'NXSDB_WY2023_CONUS_station_master.txt' file, which lists all stations checked for cross-sections in Aquarius. Stations without valid cross-sections from either measurement method are excluded from the merged table. The station location information from NWIS is used to create a layer named 'geometry,' storing the stations as point vector objects referenced to WGS84, resulting in a single geometry layer. Second, a 'record' layer is created for each station, containing combined cross-section metadata from both measurement methods with each record layer named according to the station ID. A unique measurement ID is assigned to each cross-section based on the station number, date, measurement method, and measurement number on the reported day, and stored as an attribute in the record layer called 'mm_ID'. Finally, measurement layers are added for each valid cross-section indicated in the record layer with each measurement layer named according to the assigned measurement ID. All layers are written incrementally, with thread pool processing used to expedite creation of the comprehensive GeoPackage file provided in this data release as 'NXSDB_WY2023_CONUS.gpkg.' The writing software enforces the schema defined in the accompanying JSON schema file ('NXSDB_schema_v1.2.0.json') to ensure consistency across all layers. As part of this enforcement, each attribute is stored in one of three data types that include TEXT, INTEGER, or REAL, with the exception of the 'geom' attribute, which is stored in GeoPackage binary format as a BLOB, in compliance with the OGC standard. The schema file further specifies floating-point precision for REAL-type attributes according to the 'precision' field. For more details on the schema, see the README documentation. 20250925 A Python function was written that checks the final GeoPackage against the target JSON schema for the NXSDB provided with this data release (NXSDB_schema_v1.2.0.json). The function outputs a text file with validation warnings if any are encountered, or fails if required data or tables are missing. Validation warnings include the existence of tables in the GeoPackage that are not anticipated by the target schema used by the authors to subsequently inspect and validate manually. 20250925 Vector Entity point 7451 0.0197262852 0.0277477663 Decimal seconds WGS_1984 WGS_84 6378137.0 298.257223563 NXSDB_WY2023_CONUS.gpkg GeoPackage file containing stations with cross-section data in water year 2023. The described metadata is for the 'geometry' layer within the GeoPackage and contains station attribute information for the 8,556 gages with at least one cross section in the database. Definition sources labeled as 'NWIS' (National Water Information System) indicate that the attribute was downloaded via the Python dataretrieval package (Hodson and others, 2023). NWIS-sourced attribute definitions can be found at https://help.waterdata.usgs.gov/codes-and-parameters. Definition sources labeled as 'NHD' (National Hydrography Dataset) indicate that the attribute was downloaded via the HyRiver package (Chegini and others, 2021). NHD-sourced attribute definitions can be found at https://www.usgs.gov/ngp-standards-and-specifications/national-hydrography-dataset-nhd-data-dictionary-feature-classes. 'Producer Defined' attributes were constructed from Python code. The 'geom' attribute is defined by the Open Geospatial Consortium (OGC). See processing steps for more information. In addition to the 'geometry' layer, there are additional record and measurement layers containing the cross-section metadata and data, respectively. See the 'README_Documentation.html' provided with this data release for a full description of the GeoPackage contents. Producer Defined station A string of numeric characters representing the station number. NWIS Numeric station number. station_nm Name of the station. NWIS Alphanumeric station name. total_mmt_count Total number of midsection and moving-boat measurements for the given water year. Producer Defined 1 139 ms_mmt_count Number of midsection measurements for the given water year. Producer Defined 0 55 ms_first_mnnt_ts Timestamp of the first midsection measurement for the specific station in ISO 8601 format with the time in UTC. Producer Defined A measurement timestamp in ISO 8601 format with the time in UTC. ms_last_mmt_ts Timestamp of the last midsection measurement for the specific station in ISO 8601 format with the time in UTC. Producer Defined A measurement timestamp in ISO 8601 format with the time in UTC. mb_mmt_count Number of moving-boat ADCP measurements for the given water year. Producer Defined 0 139 mb_first_mmt_ts Timestamp of the first moving-boat ADCP measurement for the specific station in ISO 8601 format with the time in UTC. Producer Defined A measurement timestamp in ISO 8601 format with the time in UTC. mb_last_mmt_ts Timestamp of the last moving-boat ADCP measurement for the specific station in ISO 8601 format with the time in UTC. Producer Defined A measurement timestamp in ISO 8601 format with the time in UTC. num_georeferenced Number of georeferenced moving-boat ADCP measurements for the specific station. Producer Defined 0 36 reachcode Unique identifier for a reach, which is a segment of a waterbody that is part of the NHD network. It is used to identify the location of a reach within the NHD surface water drainage network. Reach Codes are essential for linking water-related entities and analyzing their upstream/downstream relationships. The first eight digits are the Watershed Boundary Dataset HUC8 code. The next six digits are randomly assigned, sequential numbers that are unique within a HUC8 (USGS, 2024). NHD NULL No Data Producer defined A string of numeric values. measure Measure along the reach, in percent from downstream end, where a point event is located (USGS, 2024). NHD 0.0 100.0 comid A unique identifier for a waterbody through which a flowline flows. It is used to identify the waterbody that contains a specific flowline feature. ComIDs are important for understanding the flow direction and flow relationships within the NHD. NHD NULL No Data Producer defined 3109 948100400 huc_cd Eight-digit hydrologic unit code (HUC) the station is located in. NHD NULL No Data Producer defined A string of numeric values. state_cd State code (numeric), as defined in NWIS. NWIS 0 56 state Two-letter state abbreviation. NWIS Two-letter state abbreviation. dec_lat_va Decimal latitude from NWIS. NWIS 19.71213889 70.2813799 dec_long_va Decimal longitude from NWIS. NWIS -165.5099745 -67.3180556 coord_datum_cd Latitude/longitude (horizontal) coordinate datum from NWIS. NWIS NULL No Data Producer defined NAD83 North American Datum of 1983 Producer defined NAD27 North American Datum of 1927 Producer defined coord_meth_cd Indicates the method used to determine latitude and longitude coordinates from NWIS. NWIS NULL No Data Producer defined X GNSS2 - Level 2 Quality Survey Grade Global Navigation Satellite System. Producer defined M Interpolated from MAP. Producer defined N Interpolated from Digital MAP. Producer defined Y GNSS3 - Level 3 Quality Survey Grade Global Navigation Satellite System. Producer defined G Mapping grade GPS unit (handheld accuracy range 12 to 40 ft). Producer defined D Differentially corrected Global Positioning System. Producer defined F Survey-grade GPSF. Producer defined U Unknown. Producer defined S Transit, theodolite, or other surveying method. Producer defined R Reported. Producer defined L Long range navigation system. Producer defined W GNSS1 - Level 1 Quality Survey Grade Global Navigation Satellite System. Producer defined C Calculated from land net. Producer defined coord_acy_cd Code for decimal coordinate accuracy from NWIS. NWIS F Accurate to + or - 5 sec. Producer defined M Accurate to + or - 1 min. Producer defined S Accurate to + or - 1 sec. Producer defined 1 Accurate to + or - .1 sec (Differentially-Corrected GPS). Producer defined U Unknown or unspecified. Producer defined T Accurate to + or - 10 sec. Producer defined H Accurate to + or - .01 sec (Differentially-Corrected GPS). Producer defined 5 Accurate to + or - .5 sec (PLGR/PPS GPS). Producer defined C Level 2 survey-grade GPS. Producer defined R Accurate to + or - 3 sec (SPS GPS). Producer defined D Level 3 survey-grade GPS. Producer defined B Level 1 survey-grade GPS. Producer defined alt_va Altitude of Gage/land surface from NWIS. NWIS NULL No Data Producer defined -32808.0 10986.0 alt_datum_cd Datum for altitude from NWIS. NWIS NULL No Data Producer defined NAVD88 North American Vertical Datum of 1988. Producer defined NGVD29 National Geodetic Vertical Datum of 1929. Producer defined LMSL Local Mean Sea Level. Producer defined alt_meth_cd Method used to determine altitude from NWIS. NWIS NULL No Data Producer defined X GNSS2 - Level 2 Quality Survey Grade Global Navigation Satellite System. Producer defined L Level or other surveyed method. Producer defined Y GNSS3 - Level 3 Quality Survey Grade Global Navigation Satellite System. Producer defined Z GNSS4 - Level 4 Quality Survey Grade Global Navigation Satellite System. Producer defined M Interpolated from topographic map. Producer defined N Interpolated from Digital Elevation Model. Producer defined G Definition unavailable (see https://help.waterdata.usgs.gov/code/alt_meth_cd_query?fmt=html). Producer defined D Differentially corrected Global Positioning System. Producer defined F Survey-grade GPSF. Producer defined W GNSS1 - Level 1 Quality Survey Grade Global Navigation Satellite System. Producer defined U Definition unavailable (see https://help.waterdata.usgs.gov/code/alt_meth_cd_query?fmt=html). Producer defined R Reported method of determination. Producer defined J Light Detection And Ranging, airplane. Producer defined A Altimeter. Producer defined alt_acy_va Accuracy of altitude in feet from NWIS. NWIS NULL No Data Producer defined 0.0 700.0 wkb Well-Known Binary (WKB) string representation of each station location. Necessary for efficiently reading in the GeoPackage programatically (see 'help_fxs.py' and 'help_fxs.R' within 'demo.zip' to see recommended reading algorithms.). Producer Defined Well-Known Binary (WKB) string. geom Required GeoPackage geometry column representing the station location. In addition to the same coordinate values stored in the attributes 'dec_lat_va' and 'dec_long_va', the geometry BLOB header encodes a spatial reference identifier, which ties to the CRS definition in the GeoPackage. This allows GeoPackage drivers to correctly interpret station point geometries in a spatial context. OGC GeoPackage binary NXSDB_WY2023_CONUS_station_master.txt Tab-delimited text file containing station IDs used to query Aquarius for cross-section data in water year 2023. Generated using USGS Water Services (https://waterservices.usgs.gov/). Producer Defined station Unique identifier for the station. Producer Defined Numeric station numbers. station_nm Name of the station. Producer Defined Alphanumeric station names. README_Documentation.html Contains general background information about the cross-section table and a schema description of the 'NXSDB_WY2023_CONUS.gpkg' GeoPackage schema description. Producer Defined demo.zip A zip archive that contains R and Python notebooks intended to show users how to programatically access and query the 'NXSDB_WY2023_CONUS.gpkg' GeoPackage dataset. Producer Defined change_log.txt Text file containing versioning information relevant to updates reflected in the GeoPackage and or documentation. Producer Defined NXSDB_schema_v1.2.0.json Schema file in JSON format used to validate the final GeoPackage file NXSDB_WY2023_CONUS.gpkg. Producer Defined NXSDB_WY2023_CONUS_v1.1.0_PROVISIONAL.zip A ZIP archive of the initial [provisional] NXSDB release (v1.1.0, August 2025) conformant with the JSON schema NXSDB_schema_v1.1.0.json. This deprecated version was archived to allow users to access thi previous iteration for reference or reproducibility. A detailed record of all changes is included in the accompanying change_log.txt. Producer Defined NXSDB_WY2023_CONUS_v1.2.0_APPROVED.zip A ZIP archive of the approved NXSDB release (v1.2.0, September 2025) conformant with the JSON schema NXSDB_schema_v1.2.0.json. This stable version is the primary downloadable GeoPackage, while the earlier version remains archived as a ZIP file named after its schema version, allowing users to access previous iterations for reference or reproducibility. A detailed record of all changes is included in the accompanying change_log.txt. Producer Defined GS ScienceBase U.S. Geological Survey mailing address

Denver Federal Center, Building 810, Mail Stop 302

Denver CO 80225 United States 1-888-275-8747 sciencebase@usgs.gov Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data on any other system or for general or scientific purposes, nor shall the act of distribution constitute any such warranty. Digital Data https://doi.org/10.5066/P13PPXKN None 20260116 Travis M. Knight U.S. Geological Survey, Water Resources Mission Area Hydrologist mailing address

Suite 704

Arden NC 28704 US 239-579-6068 tknight@usgs.gov FGDC Content Standard for Digital Geospatial Metadata FGDC-STD-001-1998