Data Package Metadata   View Summary

Methods to produce BarrierTNS.csv:

The California Department of Fish and Wildlife’s Summer Townet Survey (STN) conducts surveys using a large conical, fixed-frame net which is pulled 2-3 times, obliquely, through the water for 10 minutes at each sampling station. STN samples at fixed locations from eastern San Pablo Bay to Rio Vista on the Sacramento River, and to Stockton on the San Joaquin River and a single station in the lower Napa River. These stations are sampled every other week June through August. The STN data was downloaded from CDFW’s website (https://www.wildlife.ca.gov/Conservation/Delta/Townet-Survey) and subset to include only stations in the vicinity of the EDB from 2014 to 2022.

Methods to produce Barrier DJFMP.csv:

The Delta Juvenile Fish Monitoring Program (DJFMP), run by the US Fish and Wildlife Service conducts beach seines at location across the Delta year-round. This program was designed to monitor distribution of juvenile salmonids, but is useful for tracking near-shore fishes of all types. Full information on their sampling methods is available on their EDI data publication: https://portal.edirepository.org/nis/mapbrowse?scope=edi&identifier=244&revision=11

The beach seine data was downloaded from the data package and subset to include only stations in the vicinity of the EDB from 2014 to 2022.

Methods to produce hydrodynamics data. The Fisherman’s Cut (FCT) continuous flow monitoring station has been operated by the California Department of Water Resources North Central Region Office (NCRO) since June 2015, just after the installation of the 2015 West False River Emergency Drought Barrier. All other stations are operated and maintained by the United States Geological Survey (USGS), California Water Science Center, Hydrodynamics group within the Estuarine Hydrodynamics and Sediment Transport Branch since April 1996. They followed the USGS Index-Velocity Method protocols to collect the stage, mean channel velocity, and flow data used in this report, as outlined in their publication: Levesque, V.A., and Oberg, K.A., 2012, Computing discharge using the index velocity method: U.S. Geological Survey Techniques and Methods 3–A23, 148 p. The net flow data used in this report was calculated using a Godin filter to remove the tidal signal from the data. The initial 15-minute data were treated to remove small gaps of less than or equal to three hours in the original data set using linear interpolation with surrounding good data. (This deviated slightly from USGS policy of interpolating for gaps of two hours or less.) Any gaps greater than three hours were not interpolated. After interpolating small gaps, the Godin filter applied a 24-hour moving average to the 15-minute data, then applied a 24-hour moving average to the first moving average and finished with a 25-hour moving average of the second moving average.

Turbidity methods Turbidity and settleable solids samples were taken three times a day during in-water construction activities, at approximately 9 a.m., 12 noon, and 3 p.m. Samples were taken either from the shore or by boat at a nominal distance of 300 feet upstream and downstream of construction activities. Samples were taken for 20 consecutive days during EDB installation from Thursday, June 3, 2021, to Tuesday, June 22, 2021. For each settleable solid test the technician transferred 1.00 liter into a graduated Imhoff cone and let the cone settle, undisturbed for one hour. The technician then recorded the time when the sample was captured and the amount of sediment in milliliters that has settled out of suspension. To measure turbidity, the technician used the remaining sample water to take a turbidity meter reading using a Hach 2100P instrument.

Methods to assess Impacts on Predation The EDBPS was conducted at the site of the EDB in False River. The study reach extended 500 meters east and 500 meters west of the EDB and included the entire channel width. The study implemented the use of drifting predation loggers to measure where and when predation events occurred. The drifting predation loggers were designed and constructed based on National Marine Fisheries Service’s Predation Event Recorders (PERs) (Demetras et al., 2016). Each PER was equipped with a GPS transponder and magnetic timer. A tethered Golden Shiner was attached to each PER as a surrogate bait species for Chinook Salmon. Sampling occurred at the study site 3 hours before sunset to 3 hours after sunset, to target a total of 600 minutes of cumulative float time on each sampling day. Sampling was conducted during 3 periods: pre-construction, during construction, and post-construction of the EDB. Before and during construction, PERs were deployed across the width of False River 500 meters up-current of the barrier and allowed to drift to 500 meters down-current of the barrier, or drift for 1 hour if the current was insufficient for the PERs to drift the entire 500 meters. During the post-construction period, PERs were deployed simultaneously on the east and west sides of the EDB. Depending on prevailing current and winds, PERs were allowed to drift either from the EDB to 500 meters away, or from 500 meters away towards the EDB, or for 1 hour within 500 meters of the barrier. If a PER intersected the buoy line associated with the EDB and became stuck, crews manually guided PERs under the buoy line to continue the duration of the drift. PERs were manually retrieved one by one. Upon retrieval, the status of the predation-triggered time and golden shiner was recorded. If triggered, the timer was reset and golden shiner replaced before subsequent deployments. A predation event was defined as a triggered timer where the fish was missing. Light and Vegetation Surveys During the 2021 sampling season, pilot surveys were conducted to measure light and vegetation within the study area. Results from these efforts are not included in the survival model and thus not discussed in detail in the results below. During the May, June, and July 2021 sampling months, a light meter was deployed subsurface from a boat and transects were driven across the PER deployment area. During November 2021 through June 2022 sampling months, light surveys employed a stationary light meter deployed at the study site during each day’s PERs sampling. Two vegetation surveys were completed during June and July, 2021. Surveys used side-scan sonar to visualize underwater structures and vegetation. During November 2021 through June 2022 sampling months, vegetation surveys consisted of monthly surveys of the shorelines using sonar and drone technology. Statistical Approach To evaluate effects of the barrier on predation risk, a Cox Proportional Hazard Model was used. This model evaluates the effects of spatial and temporal parameters during each construction period. The response variable, “predation”, is a Boolean value, which may occur at a single point in time for a trial (the trigger time) and after which, the trial ends. The predictor variables are a mix of linear parameters and categorical values. Predation risk models were run with each combination of all 10 parameters. The most parsimonious model was chosen via Akaike Information Criterion (AIC).

Methods to produce ContinuousWQ.csv (Daily averages and medians of continuous chlorophyll fluorescence, turbidity, electrical conductivity, DO, and temperature from 2020-2021): DWR and the U.S. Geological Survey (USGS) measures continuous real-time chlorophyll fluorescence, turbidity, electrical conductivity, dissolved oxygen, and water temperature data using Yellow Spring Instrument (YSI) EXO2 sondes deployed at water quality monitoring stations located throughout the San Francisco Estuary. The YSI sondes are configured to collect data at 15-minute intervals at a water depth of approximately 1 meter. For detailed information on YSI instrumentation, visit http://www.ysi.com/index.php. Field staff regularly maintain the monitoring stations by frequent calibration and validation of the YSI sondes using methods outlined in the USGS technical report Guidelines and Standard Procedures for Continuous Water-Quality Monitors: Station Operation, Record Computation, and Data Reporting (Wagner et al. 2006). Continuous water quality collected in 2020-2021 was acquired through direct data requests or downloaded from either the Water Data Library (WDL) or the National Water Information System (NWIS). Some of the data collected by USGS was considered provisional at the time of acquisition. Continuous data from 12 stations located near the Emergency Drought Barrier were integrated into one data set and aggregated into daily means and medians. Additional details on data integration and processing can be found in the EDBdata GitHub package: https://github.com/mountaindboz/EDBdata.

Methods to produce DiscreteWQ data: Discrete water quality data were collected by DWR"s North Central Region Office Water Quality Evaluation Section (WQES) and the ENvironmental Monitoring Program (EMP). Both programs collected water samples at stations throughout the Delta as described in the "stations" look-up table. Samples were collected either using a van-dorn sampler or a pump at 1-meter depth. Samples were filtered and processed as described in EMP's metadata: https://portal.edirepository.org/nis/metadataviewer?packageid=edi.458.5 . All samples were analyzed for a variety of water quality constituents as DWR's Bryte Laboratory. ==================== Data Sources ========================= https://portal.edirepository.org/nis/metadataviewer?packageid=edi.458.5 ===========================================================

Methods for SCHISM modeling To assess the effectiveness of the EDB in reducing salinity intrusion, three-dimensional simulations were carried out using the Bay-Delta SCHISM three-dimensional circulation model (Ateljevich et al. 2014). Bay-Delta SCHISM is an application of the Semi-implicit Cross-scale Hydroscience Integrated System Model (Zhang and Baptista 2008; Zhang et al. 2016). Results shown here focus on historical period of June 1–December 31, 2021, with the full structure in place. To obtain the effect of the barrier, salinity difference maps were generated comparing salinity from historical operations and geometry (EDB) to a hypothetical case with no drought barrier (No EDB). Field data were used for flow boundary conditions and exports through the end of the calendar year. To investigate the impact of Delta Cross Channel (DCC) Gate operations, additional hypothetical/sensitivity simulations were performed that modeled water quality under enhanced operation of the DCC. The enhancement was to add two tidal days (50 hours) at the beginning of each weekly open-close cycle in September and October. The Bay-Delta SCHISM model was used with its usual historical data sources for the year 2021. Assumptions are reported in Ateljevich (2014) and the mesh and associated preprocessing templates are avilable on the project GitHub repository https://github.com/CADWRDeltaModeling/BayDeltaSCHISM as tag v1.1.4. Emergency Drought Barrier was modeled as a barrier with culverts representing a small leakage flow observed over monitoring campaigns by DWR North Central Regional Office. This files represent the difference in specific conductance in uS/cm between a simulation with the drought barrier and a no action case without the barrier under the same hydrology for the months of . Locations (x,y) are at schism nodes, reported in meters in the UTM Zone 10 projection (EPSG:32610). Conductance data have been depth and time averaged over the first 14 days of the indicated months. Ateljevich E, Nam K, Zhang Y, Wang R, Shu Q. 2014. “Bay Delta Calibration Overview.” In: Methodology for Flow and Salinity Estimates in the Sacramento–San Joaquin Delta and Suisun Marsh. 35th Annual Progress Report to the State Water Resources Control Board in Accordance with Water Right Decisions 1484 and 1641. June 2014. Sacramento (CA): California Department of Water Resources. Zhang Y, Baptista AM. 2008. “SELFE: A Semi-implicit Eulerian-Lagrangian Finite-Element Model for Cross-Scale Ocean Circulation." Ocean Modelling Volume 21 (Issues 3–4), Pages 71–96. Zhang Y, Ye F, Stanev EV, Grashorn S. 2016. “Seamless cross-scale modeling with SCHISM.” Ocean Modelling Volume 21 (Issue 3): Pages 71–76.

For the analysis of fish salvage data, the salvage database was downloaded from the CDFW website: https://filelib.wildlife.ca.gov/Public/salvage/

The total catch was calculated for all listed fish species (Chinook Salmon [Oncorhynchus tshawytscha], Delta Smelt [Hypomesus transpacificus], Longfin Smelt [Spirinchus thaleichthys], Green Sturgeon [Acipenser medirostris], and Steelhead [Oncorhynchus mykiss]) for 2014-2022.

Methods to produce zooplankton data

Zooplankton were monitored primarily using four existing IEP surveys, including the CDFW 20mm Survey, STN and FMWT, as well as Environmental Monitoring Program (see Figure 5).

Zooplankton sampling by STN and FMWT are described in the fish methods setp. EMP conducts water quality, phytoplankton, and zooplankton sampling on a monthly basis throughout the upper estuary at 17 stations. At each station, they collect a 10-minute stepped oblique trawl using the same zooplankton sled used by FMWT (see above). Additionally, they collect microzooplankton using a vertically-integrated pump sample (https://wildlife.ca.gov/Conservation/Delta/Zooplankton-Study). Two of these stations are not fixed, but instead follow the salinity field and sample where the bottom specific conductance reaches 2000 uS/cm and 6000 uS/cm, respectively.

All three surveys have similar zooplankton processing methods. In brief, samples are concentrated in the laboratory by pouring them through a sieve screened with 150-micron mesh wire and reconstituted to organism densities of 200-400 per milliliter. The sample is stirred to distribute the animals homogeneously and a 1-milliliter subsample is extracted with an automatic pipette and placed in a Sedgewick-Rafter cell (slide). All animals on a slide are identified and counted under a compound microscope to the lowest possible taxonomic classification. This procedure is repeated until 6% of the sample, or between 5 and 20 slides, are analyzed.

Data from the three surveys were combined using the zooper package in R (Bashevkin et al 2022). This package combines datasets by standardizing variable names and accounting for differences in taxonomic resolution. We only used mesozooplankton data (150 micron net) and removed any taxa not quantitatively sampled by this mesh size. We aggregated taxonomic groups into larger categories for analysis.