Hydrodynamic time-series data from two marshes and adjacent shallows in Northern San Francisco Bay, California, 2022-2023

Two quadpod platforms were deployed in Northern San Francisco Bay, one in the subtidal shallows of Corte Madera Bay, and a second in the subtidal shallows of San Pablo Bay National Wildlife Refuge. Platforms were turned roughly every 60 days, equating to eight total deployments, named A through H. Each of the two subtidal stations (named PS and CS) included a Nortek single point acoustic doppler velocimeter each with a Seapoint turbidity meter, and an RBR CTD. Both turbidity meters were equipped with a Zebra-Tech hydro-wiper to prevent biofouling.

Two weightstand (flat weight with center pole) platforms were deployed in Northern San Francisco Bay, one in the intertidal shallows of Corte Madera Bay, and a second in the intertidal shallows of San Pablo Bay National Wildlife Refuge. Platforms were turned roughly every 60 days, equating to nine total deployments, named A through I. Each of the two intertidal stations (named PI and CI) included an RBR bursting pressure sensor and a Seapoint turbidity meter logged by an RBR Virtuoso single-channel data logger. Both turbidity meters were equipped with a Zebra-Tech hydro-wiper to prevent biofouling.

Two marsh-top platforms, CM and PM, were deployed at Corte Madera Marsh and San Pablo National Wildlife Refuge Marsh, both were positioned directly in line with the intertidal and subtidal sites, at approximately 25 m back from the bay edge. The bay edge was defined as the end of marsh vegetation and start of mudflat. Both platforms included a Seapoint turbidity meter logged by an RBR Virtuoso single-channel data logger, as well as an RBR logger measuring pressure attached to rebar above the marsh surface. Both Seapoint turbidity meters were equipped with Zebra-Tech hydro-wipers to prevent biofouling. Platforms were turned roughly every 60 to 90 days, equating to eight total deployments for CM, named A through H, and five total deployments for PM, named A, C, D, F, and H.

One platform, CC, consisting of a Nortek Aquadopp acoustic doppler profiler with a Seapoint turbidity meter and a RBR pressure logger were attached to a grate that was deployed directly onto the bottom surface of the channel of a tidal creek located in Corte Madera Marsh. The turbidity meters were equipped with a Zebra-Tech hydro-wiper to prevent biofouling and had a wire cage placed around it, outside of the sensors sample volume, to keep floating debris from catching or damaging the sensor. The platform was turned roughly every 60 days, equating to eight total deployments, named A through H.

All recovered data from the retrieved instruments were post-processed using manufacturers software packages and USGS MATLAB codes or USGS Python codes.

Data were accumulated into NetCDF files. During post-processing, data collected before deployment and after recovery were removed. Bad data points identified by 1) the deployment and recovery times, 2) data flagged as erroneous by the manufacturer, 3) amplitude or correlation values outside of predetermined quality thresholds for the velocimeter were removed, as were data where pitch and roll exceeded 25 degrees. The files include data of intermittent poor quality due to various reasons which may include but are not limited to biofouling, movement of platform, and exposure during low tides. Despite removal of bad data, some are still present in the files and therefore judgment should be applied when using them, as with all other data files. Use of all data requires careful assessment of data quality.

For each turbidity sensor, a linear least-squares regression with intercept of zero was made between the turbidity values (in NTU or volts) and sample suspended sediment concentration (SSC) values (in milligrams per liter) from in situ measurements coincident to water sample collection. These constants are available as a support file in this data release NBM_Turbidity_to_SSC_calibration_constants.csv. Users should also note the gain setting used when applying a calibration coefficient to the turbidity data, all turbidity to SSC calibration constants were based on 5x gain setting.

All pressure data were given a new variable P_1ac, that contains atmospherically corrected pressure data. For this correction, the residual difference between the in-air reading of the pressure sensor and the NOAA station was removed, and is noted in the netCDF file as atmospheric_pressure_correction_offset_applied.

Wave statistics were calculated from the bursts of velocity and pressure measured by the ADV and from the bursts of pressure measured by the RBR bursting pressure logger. Calculations were completed using routines available on https://code.usgs.gov/cmgp/stglib. Additional information on frequency cutoffs used for the calculations can be found in the metadata of the nc data file.

Files were named with a convention that uses a 12- to 15-digit alpha-numeric code. The first three characters for this dataset are all 'NBM' indicating the experiment name; the fourth and fifth positions represent the calendar year in which the project commenced (22, 2022); the sixth character is a letter representing the embayment (S, San Pablo; C, Corte Madera); the seventh character is a letter representing the station's location (M, marsh; C, tidal creek; I, intertidal; S, subtidal); the eight character is a letter corresponding to the deployment (A through I, where A is the first deployment); the ninth and tenth characters represent the instrument position on the platform, where 01 is the top-most. The eleventh, twelfth, and sometimes thirteenth characters are abbreviations for the instrument type (vec, Norteks Acoustic Doppler Velocimeter; aq, Nortek aquadopp acoustic doppler current profile; pt, RBRs bursting pressure sensor with temperature; pr, RBRs pressure sensor; ctd, RBRs conductivity, temperature, and depth logger; tu, RBRs Virtuoso turbidity logger). There are an additional 1 to 4 characters with a hyphen used to convey information about the data contained in the file. Filenames containing (-a) were used for instruments with singular/averaged observations, the (-a) also appears in the vector filenames to indicate that the file includes the burst data, while filenames with (-s) indicate the statistics/burst averaged observations of the data for bursting instruments. Similarly, (s-a) indicates the file contains wave statistics. For the bursting pressure sensors only (b-cal) indicates the file contains the burst data, and that those pressure data were corrected for atmospheric pressure. Information on the files available for each station and the data each file contains is listed in the NBM_time_series_file_table.csv.

The netCDF files were compressed and combined in zipped files. Files were grouped by station (San Pablo Subtidal, Intertidal, Marsh, Corte Madera Subtidal, Intertidal, Marsh or Tidalcreek) and then by instrument type. Instrument type suffixes follow the nomenclature defined above for filenames including a two- or three-character code for the instrument type (vec, Norteks Acoustic Doppler Velocimeter; aqd, Nortek aquadopp acoustic doppler current profile; pt, RBRs pressure sensor with temperature; pr, RBRs pressure sensor; ctd, RBRs conductivity, temperature, and depth logger; tu, RBRs Virtuoso turbidity logger).