These methods, instrumentation and/or protocols apply to all data in this dataset:| Methods and protocols used in the collection of this data package |
|---|
| Description: |
Hydrology Data
We acquired hydrology (Delta inflow, Delta outflow, and water velocity) data from two sources. Delta inflow and outflow data from 1975-2021 are from the Dayflow model developed by the California Department of Water Resources (DWR) (CDWR 2022b). This model uses measured instantaneous flow and export data from stations throughout the Delta to calculate daily average Delta inflow and uses assumptions about within-Delta use to calculate a daily average net Delta Outflow index. Water velocity data collected from 2007-2021 at five stations in the Delta were obtained from the United States Geological Survey (USGS) National Water Information System using the dataRetrieval R package (De Cicco et al. 2022). The five water velocity stations operated by the USGS and used in this study are: Cache Slough at Ryer Island (11455350), Cache Slough above Ryer Island Ferry near Rio Vista CA (11455385), San Joaquin River at Jersey Point CA (11337190), Middle River at Middle River CA (11312676), and Old River at Bacon Island CA (11313405). The Cache Slough at Ryer Island station was discontinued in April 2019 and was replaced by the Ryer Island near Ryer Island Ferry station. Data for these two stations located on Cache Slough were combined to represent Cache Slough above its confluence with the Sacramento River.
The instantaneous water velocity data, collected at 15-minute intervals, were processed through a low-pass filter to remove tidal-period variation and calculate net velocity (Godin 1972). Before applying the filter, we imputed values for gaps up to 2 hours in the instantaneous data using linear interpolation with the 'imputeTS' R package (Moritz and Bartz-Beielstein 2017). The difference of the instantaneous velocity and net velocity resulted in the tidal velocity. We aggregated the net velocity data into weekly means, whereas tidal velocity was grouped into weekly minimum, maximum, and maximum absolute values.
| | Data Source | | | Data Source | |
| | | | Description: |
Water Quality, Nutrient, and Chlorophyll Data
We acquired the water quality (water temperature, salinity, and Secchi depth), nutrient (dissolved ammonium, dissolved nitrate + nitrite, and dissolved ortho-phosphate), and chlorophyll-a (or chlorophyll) data from the IEP integrated discrete water quality dataset available on EDI (Bashevkin et al. 2023). This dataset contains water quality, nutrient, and chlorophyll data collected by 16 long-term monitoring surveys that sample approximately monthly at locations throughout the estuary. We obtained data collected by 12 surveys that have at least 20 years of data between 1975-2021 within our study area in the estuary. All water temperature and salinity measurements were typically collected at a depth of one meter, and the nutrient and chlorophyll samples were collected at various depths less than five meters from the surface. For more information on data collection methods, refer to the integrated discrete water data publication on the Environmental Data Initiative repository (Bashevkin et al. 2022b).
Since some surveys collected more than one sample per day at a station, we filtered the water quality, nutrient, and chlorophyll samples keeping only one sample per station and day to ensure each sample had equal weight. Data were additionally restricted spatially and temporally by only including subregions that contained data for at least 35 of the 46 years (75%) between 1975-2021 for all four seasons. We used the deltamapr R package (Bashevkin 2021) to divide data into 34 subregions, and seasons were defined as follows: winter (December of the previous year, and January-February), spring (March-May), summer (June-August), and fall (September-November). The 34 Delta subregions were categorized into one of five broader regions including North, Confluence, South-Central, Suisun Marsh, and Suisun Bay. The Suisun Marsh region was excluded from the nutrient and chlorophyll analyses because of inconsistent long-term sampling.
A few anomalous water quality and nutrient values were removed before analysis. We removed values if they had a Z-score or modified Z-score greater than 15, grouped by each subregion for the water quality and nutrient parameters, respectively. Using the modified Z-score was more appropriate for the nutrient data set since it contains some values below the reporting limit (RL) for the laboratory method and the modified Z-score is more robust to this type of data since it is based on medians. No dissolved ammonium or chlorophyll values with modified Z-scores greater than 15 were removed as we determined they were within their expected ranges. Additionally, we excluded nutrient values that were below the RL with relatively high RL values defined as greater than the 75th percentile of the data set.
The laboratory RLs for the nutrient parameters were not always provided by the data sources. Therefore, we needed to impute RL values when the result was reported less than the RL and its RL was unknown. For the EMP data, we used the most common historical RL value for all three nutrient parameters, 0.01 mg/L, for the records without reporting limits. In the SFBWQS data set, values below the RL were not explicitly documented; however, through personal communication with USGS investigators, we confirmed that if at least one of the three nutrient parameters had a value reported for a station and day, we could assume that the other parameters were sampled but below the RL (Erica Nejad, USGS, Dec. 16, 2021). We used 0.0007 mg/L, 0.0007 mg/L, and 0.0015 mg/L as the reporting limits for dissolved ammonium, dissolved nitrate nitrite, and dissolved ortho-phosphate, respectively, for the SFBWQS data (Erica Nejad, USGS, personal comm.).
To prepare the discrete water quality, nutrient, and chlorophyll data for analysis, we aggregated the data set for each parameter as seasonal-regional averages. To aggregate the data, we calculated monthly averages for each region, which we then used to calculate seasonal-regional averages for each year. We used an adjusted water year, December through November with the December of the previous calendar year included with the following year, to allow the entire “fall” season to be included in the same water year. Before aggregating the nutrient and chlorophyll data, we substituted the values below the laboratory reporting limit (RL) with simulated values between zero and the RL based on a uniform distribution. We ran one simulation for each parameter and set a seed prior to running the simulation to ensure reproducibility. Overall, 7% of the ammonium, 1% of the nitrate + nitrite, 0.7% of the ortho-phosphate, and 0.2% of the chlorophyll values were below the RL and required replacement with simulated values. R code used for the preparation of the data can be found in the WQ-LT-Publication GitHub repository archived on Zenodo (Bosworth et al. 2024).
| | Data Source | |
| | | Description: |
These methods were used to create the 'clams.csv' and 'GRTS_clams.csv' data.
Clams
Clams were sampled by two different surveys, both conducted by the California Department of Water Resources (CDWR). The first survey is the long-running benthic invertebrate survey of the Environmental Monitoring Program (EMP) (Wells and IEP 2022), which has sampled monthly at ten sites from San Pablo Bay upstream to endpoints at Clifton Court Forebay, near Stockton, and up the Sacramento River as far as Rio Vista. This survey has occurred at these ten sites from 1996 to present, and from 1975 to 1996 at a smaller subset of those sites along with others since discontinued. The ten EMP sites sampled since 1996 are hereafter referred to as “EMP core sites”, and the dataset encompassing sites sampled since 1975 as “EMP long-term data”. See Wells and IEP 2022 for full metadata (Wells and IEP 2022).
The second survey is a spatially extensive survey performed twice a year in May and October from 2007 through 2019 except for 2013 and 2016, designed to augment EMP data. CDWR staff who designed the survey chose sites in the same geographic range as the EMP survey, using a Generalized Random Tessellation Stratified (GRTS) sampling design (Stevens and Olsen 2004) which stratified site selection by water body type, ensuring adequate sampling effort in habitats with smaller total areas. Staff sampled 175 sites each year from 2007 through 2017 (sampling did not occur in 2013 and 2016), 100 sites in 2018, and only 50 core sites in 2019. Fifty of the sites sampled were the same year to year (core sites), and additional sites were newly selected each year using the same GRTS design. This survey is hereafter referred to as “GRTS”.
The two surveys are complementary in that the EMP survey is temporally intensive but spatially limited (up to ten sites chosen to be representative of main rivers and bays, done every month), while the GRTS survey was temporally limited but spatially intensive (sampling twice a year, at a minimum of 50 and at most 175 randomly chosen sites that extended into smaller water bodies, such as sloughs and canals). Data from these surveys were subset to include only stations within the area defined above for the jellyfish data, within the following regions: North Delta, Confluence, Suisun Marsh, Suisun Bay, and South-Central (Figure 2).
For each survey, staff used a Ponar dredge to collect 0.052 m2 of benthic area to a maximum sediment depth of 10 cm at each site. The sample was rinsed over 0.595 mm sieve and all P. amurensis and C. fluminea individuals were identified, enumerated, and binned by size (e.g., 0-1 mm, 1-2 mm). Clam densities were converted to individuals/m2.
Clam counts and sizes were converted into ash-free dry mass using length:biomass regression equations (see ClamRegressions.csv). These equations are constructed monthly from an additional separate sample taken at nearby EMP sites that contained large numbers of the relevant species of clam. The regressions from these ‘reference sites’ were used for all sites within a region for that year and month. Unique regressions were used for each region, year, and month, to account for differences in condition due to temperature, food availability, and other factors.
We combined data from GRTS and the EMP core sites sampled during May and October due to GTRS only sampling during these months, and from 2007-2019 (excluding 2013 and 2016, when GRTS data were not available) to assess effects of water year type on P. amurensis and C. fluminea population densities and their combined filtration rates across regions. We used combined filtration rates to test the hypothesis that total clam filtration increased during drought conditions.
| | Data Source | |
| | | Description: |
These methods were used to create the "Alljellyfish" dataset.
Sampling information
For an interactive map of sampling sites, see: https://deltascience.shinyapps.io/monitoring/
The CDFW Fall Midwater Trawl (FMWT) survey samples at fixed locations from eastern San Pablo Bay to the Cache Slough complex and Sacramento Deep Water Ship Channel, on the Sacramento River, and to Stockton on the San Joaquin River. This survey runs once per month during September, October, November, and December at 122 stations with a trawling net pulled obliquely through the water with one, 12-minute tow per station. The net has a mouth area of 13 m2 when stretched taut and mesh that graduates from 203.2mm (8-inch) stretch mesh at the mouth to 0.5 inch stretch mesh at the cod end. The FMWT survey primarily monitors young-of-the-year fishes, but it has recorded catch of gelatinous zooplankton in its net since 2000. Catch of gelatinous in the net is converted to catch-per-unit-effort (CPUE) by dividing by the volume of water sampled, as measured with a General Oceanics Flowmeter.
The CDFW Summer Townet (STN) Survey samples 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. The STN survey runs twice per month during June, July, and August and samples at 40 stations with a D-frame net which is pulled 2-3 times, obliquely, through the water for 10 minutes at each sampling station. The net is 4.6 long, has a mouth area of 1.5 m2, and 1.27 cm mesh tapering down to 0.32 cm mesh at the cod end. The survey primarily monitors young-of-the-year fishes, but has recorded catch of gelatinous zooplankton since 2007. Catch of gelatinous zooplankton in the net is converted to catch-per-unit-effort (CPUE) by dividing by the volume of water sampled, as measured with a General Oceanics Flowmeter.
The CDFW San Francisco Bay Study (Bay Study) samples fixed locations throughout the San Francisco Bay, Suisun Bay, the lower Sacramento River, and Lower San Joaquin River once per month year-round. The survey uses midwater trawls to target fish, shrimp, and crabs, and has recorded catch of gelatinous zooplankton in their midwater trawls since 2000. The midwater trawl net has the same specifications as the FMWT net, described above. Catch of gelatinous zooplankton in the net is converted to catch-per-unit-effort (CPUE) by dividing volume of water sampled, as measured with a General Oceanics Flowmeter.
The University of California, Davis, Suisun Marsh Fish Survey (Suisun) samples fixed locations within Suisun Marsh. The survey runs once per month year-round at 25 stations using otter trawls. The otter trawl has a mouth area of 6.45 m2, a length of 5.3 m, and body net mesh from 35 mm stretch at the mouth to 6 mm stretch at the cod end. The survey primarily monitors fishes but has recorded catches of gelatinous zooplankton since the survey began in 1980. Catch of Maeotias in the net is converted to catch-per-unit-effort (CPUE) by dividing by the volume of water sampled, which is calculated based on tow area (distance traveled times width of net, multiplied by the height of the net. Maeotias is the only jellyfish recorded by this survey, but other species do occur in the area occasionally, so they may be mis-identifying some.
| | Data Source | | | Data Source | | | Data Source | |
| | | | | Description: |
This methods were used to create the "Drought Zooplankton" dataset
Zooplankton abundance (organisms/m3) data for the CDFW EMP Zooplankton study were downloaded using the zooper package (https://github.com/InteragencyEcologicalProgram/zooper), an R package that synthesizes zooplankton data from multiple monitoring surveys (Bashevkin et al. 2022). We used abundance data from either the macro (500-505 μm mesh net tow), meso (150-160 μm mesh net tow), or micro (pumped into a 43 μm mesh net) gears, depending on which gear sampled each taxon most efficiently (Kayfetz et al. 2020). We then calculated adult biomass (carbon weight, μg/m3) utilizing the conversions in Bashevkin et al (2022) for meso (P. forbesi and Daphnia spp.) and microzooplankton (L. tetraspina) and the biomass for macro zooplankton (H. longirostris) using length to weight equations (Burdi et al. 2022). We focused on data from 1994-2021 since that is when all the examined taxa were present in the estuary and excluded winter months (December – February) due to inconsistent historical winter sampling. Sampling stations were assigned to regions (Suisun Marsh, Suisun Bay, the Confluence, the North Delta, and the South Central Delta using the deltamapr package (Bashevkin 2021). Data from the North Delta were subsequently excluded due to lack of consistent long-term zooplankton sampling in the region.
| | Data Source | |
| | | Description: |
These methods were used to create the "Dayflow_ResidenceTime" dataset.
To estimate residence time (the length of time a particle of water remains in the upper estuary), we replicated the methods used by Hammock et al. (2019). In brief, we obtained monthly residence times for the Sacramento and San Joaquin river corridors from Hammock et al. for 1991, 1996, 1998, 2005, and 2009 and then fit statistical models of residence time versus inflow, project exports, and in-delta agricultural diversions, then used the best model to predicted residence time for all years not modeled by Hammock et al.
Hammock et al. (2019) ran the Delta Simulation Model 2’s (DSM2 version 8.1(CDWR 2013)) particle tracking module to calculate monthly residence time for 1000 particles inserted near Sacramento on the Sacramento River (DSM2 node 331) or near Stockton on the San Joaquin River (DSM node 21). Residence time was measured as the number of days it took for 90% of particles to exit the estuary, either by passing Martinez or by being diverted into the Central Valley Project (CVP) or State Water Project (SWP). Monthly residence time for water years 1991, 1996, 1998, 2005, and 2009 was obtained from Bruce Hammock (pers. comm.) per the original paper. These years were chosen by Hammock et al. (2019) to encompass a range of hydrological conditions. We then fit statistical models using the flow parameters below to reconstruct residence time on the Sacramento River and San Joaquin River for 1975-2021.
Flow parameters:
1. Sac - Sacramento River Inflow, at Freeport, as included in the Dayflow model, log-ten transformed. This data is derived from USGS flow station 11447650 (Latitude 38.45601, Longitude -121.50134).
2. SJR - San Joaquin River Inflow at Vernalis, as included in the Dayflow model, log-ten transformed. This data is derived from USGS flow station 11303500 (Latitude 37.67611, Longitude -121.26528).
3. Exports - Combined Central Valley Project and State Water Project daily pumping rates as included in the Dayflow model (CDWR 2022b). This data is derived from DWR’s and the US Bureau of Reclamation’s operational records.
4. Ag - In-Delta agricultural diversions, as modeled by the Delta Evapotranspiration of Applied Water model aligned with the Delta Channel Depletion Model (CDWR 2021). This model is considered significantly more accurate than the estimate of in-Delta use provided by the Dayflow model.
We modeled the residence time on the Sacramento River (RTSAC) and San Joaquin River (RTSJR) as calculated by DSM2 in Hammock et al 2019 as:
RTSAC ~ Sac + Exports + Ag + Sac*Exports
RTSJR ~ SJR+ Exports + Ag + SJR*Exports
For each residence time parameter (RTSJR or RTSAC), we evaluated all possible combinations of the variables in the global model and ranked them using Akaike’s Information Criterion, adjusted for small sample size (AICc). Statistical models were run using the ‘lm’ function in R (R Core Team 2023).
Once we determined the best possible model of Sacramento and San Joaquin residence time, we used these models to predict residence time for all years not modeled by DSM2 (1975-1990, 1994-1995, 1997, 1999-2004, 2006-2008, and 2010-2021).
|
| | Description: |
These are the steps we used to create the integrated dataset.
We developed an integrated dataset of environmental parameters derived from Interagency Ecological Program (IEP) long-term monitoring datasets from 1975-2021. Data from these surveys were averaged by region (Figure 3), and season (with Winter including December, January, February, Spring including March, April, May, Summer including June, July, August, and Fall including September, October, November). This seasonal definition meant that the ‘Fall’ season straddled two water years. Because ecological conditions in the fall are usually more dependent on conditions from the previous winter-spring than any early fall rain, we adjusted the water year to run from December 1st of the previous year to November 30th, instead of October-September. However, we still used the water year index assigned to the water year from October through September. We then took the regional average across seasons for chlorophyll and zooplankton biomass to get a single data point for each region and year. For other variables, we averaged across both season and region to get a single data point for the entire year. Some variables (such as Delta outflow) did not have a regional component, and some variables (such as the fish indices) used established summary methods instead of regional averages.
Delta Outflow, and Project Exports –
Net Delta outflow Index, and project exports (combined exports from the State Water Project and Central Valley Project) were from DWR’s Dayflow model (CDWR 2002). See supplemental info for data sources and how outflow is calculated. All data are published annually after input data have undergone quality control procedures on the California Natural Resources Agency Open Data Portal .
Water Quality, Nutrients and Chlorophyll
Water temperature, salinity, Secchi depth, dissolved ammonium, dissolved nitrate + nitrite, dissolved orthophosphate and chlorophyll-a were assembled as for Bosworth et al. (This Issue). In brief, data from 16 monitoring surveys were integrated into a discrete water quality dataset available on the Environmental Data Initiative repository (Bashevkin et al. 2023). This dataset contains water quality, nutrient, and chlorophyll data collected by long-term monitoring surveys that sample approximately monthly at locations throughout the upper estuary. We obtained data collected by 12 surveys that have at least 20 years of data between 1975-2021 within our study area in the upper estuary (Figure 3). All water temperature and salinity measurements were typically collected at a depth of one meter, and the nutrient and chlorophyll samples were collected at various depths less than five meters from the surface. For more information on data collection methods, refer to the metadata in Bashevkin et al. (2023).
We selected only one sample per station per day and restricted the dataset to include subregions that had data for at least 35 out of the 47 years for all four seasons to ensure spatial and temporal balance across the dataset. Subregions were defined by polygons from the deltamapr R package (Bashevkin 2021). Before aggregating the nutrient and chlorophyll data, we substituted the values below the reporting limit with simulated values between zero and the reporting limit based on a uniform distribution. We ran one simulation for each parameter and set a seed prior to running the simulation to ensure reproducibility. For more information on data cleaning and aggregation, see Bosworth et al. 2024.
Zooplankton
Zooplankton data were collated as for Barros et al. 2024. In brief, zooplankton Catch Per Unit Effort (CPUE, organisms/m3) for the Environmental Monitoring Program was downloaded using the zooper package, an R package that synthesizes zooplankton data from multiple IEP studies (Bashevkin et al. 2022a). We used CPUE data from either the macro (500-505 μm mesh), meso (150-160 μm mesh), or micro (43 μm mesh) nets, depending on which net sampled each taxon most efficiently. We focused on dominant taxa that are most frequently found in the diets of Delta Smelt and Longfin Smelt – the cladocerans Bosmina longirostris, Daphnia sp., Diaphanosoma sp.; the copepods Acartia sp., Acartiella sinensis, Eurytemora affinis (adults and copepodites), Limnoithona tetraspina (adults), Pseudodiaptomus forbesi (adults and copepodites), Tortanus sp.(adults), and the mysids Hyperacanthomysis longirostris, Neomysis kadiakensis, and Neomysis mercedis. We then calculated biomass per unit effort (BPUE, mgC/m3) utilizing the conversions in Bashevkin et al. (2022) for meso and micro zooplankton (copepods and cladocerans) and the BPUE for macro zooplankton (mysids) using length to weight equations (Burdi et al. 2021) for the years of 1975-2021. BPUE converts the zooplankton data into a ‘common currency’ relevant to estuarine productivity and food availability for fish consumers. We excluded winter months (December – February) due to inconsistent historical winter sampling and excluded data from the North Delta region due to lack of consistent long-term zooplankton sampling in the region.
Fall Midwater Trawl Indices
Data on population abundance of pelagic fishes (age-0 Striped Bass (Morone saxatilis), Threadfin Shad (Dorosoma petenense), American Shad (Alosa sapidissima), Longfin Smelt (Spirinchus thaleichthys), and Delta Smelt (Hypomesus transpacificus) came from the Fall Midwater Trawl Survey (FMWT) conducted by CDFW (White 2021). Annual abundance indices were obtained from the California Department of Fish and Wildlife (CDFW) website, (https://filelib.wildlife.ca.gov/Public/TownetFallMidwaterTrawl/, date: 10/11/2022). The FMWT conducts stepped-oblique trawls at 100 index stations throughout the upper estuary once per month in September, October, November, and December of each year. The annual index for each species is calculated based on weighted catches from each of 17 areas over the four-month sampling period (September-December). Details for the calculations, as well as the complete catch data from this program, are available from the CDFW website.
Salmon Cohort Replacement Rate
The combined cohort replacement rates for fall-run Central Valley Chinook Salmon and spring-run Central Valley Chinook Salmon (Oncorhynchus tshawytscha) were calculated as per Nelson et al. (This Issue). In brief, CDFW calculates annual escapement data based on watershed-wide surveys for salmon throughout their life cycle (Grandtab; Azat 2021). This dataset can be used to estimate a particular returning year class’s contribution to the population’s overall abundance. Year-over-year escapement can be used to calculate a population’s “cohort replacement rate” (CRR), which is the measure of the number of spawners produced by the parental generation of spawners three years prior (see supplemental information for details). We calculated the CRR for spring-run and fall-run Chinook Salmon and looked for effects of drought or water year index on CRR for the year that the cohort out-migrated.
| | Data Source | | | Data Source | | | Data Source | | | Data Source | |
| | | | | | Description: |
These are the methods used to create the "ContWaterQualityDaily" dataset used for Bouma-Gregson et al, 2024.
CDWR and USGS maintain a network of water-quality sondes and streamgages that collect data continuously (i.e., every 15 minutes) across the Delta. These stations collect data on water temperature, specific conductance, flow, DO, chlorophyll fluorescence (fCHL), turbidity, and pH using Yellow Springs Instruments (YSI) EXO2 sondes . Quality-control data were requested from CDWR personnel when available, and provisional data were queried from the California Data Exchange Center (CDEC) if finalized data were not available. To assess how cyanoHABs affect water-quality parameters, we plotted the daily mean of data collected at stations in the South and Central Delta that experienced cyanobacteria blooms in 2021 versus day of the year for the past seven years (2015–2021). 2014 was not included in this analysis because the water quality station at Frank’s Tract (FRK) was not installed until 2015.
| | Data Source | |
| | | Description: |
References
Azat J. 2021. GrandTab 2021.06.30 California Central Valley Chinook Population Database Report. California Department of Fish and Wildlife. [accessed 2023 Jan 4], https://nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=84381.
Barros, A., R. Hartman, S. Bashevkin, and C. Burdi. 2024. Years of drought and salt; decreasing flows determine the distribution of zooplankton resources in the estuary. San Francisco Estuary and Watershed Science. [accessed 2024 Mar 29]. 22 (1). https://doi.org/10.15447/sfews.2024v22iss1art3
Bashevkin, S. M., R. Hartman, M. Thomas, A. Barros, C. E. Burdi, A. Hennessy, T. Tempel, and K. Kayfetz. 2022. Five decades (1972–2020) of zooplankton monitoring in the upper San Francisco Estuary. Plos ONE. [accessed 2022 May 17]. 17 (3):e0265402. https://doi.org/10.1371/journal.pone.0265402
Bashevkin, S. M., S. E. Perry, E. B. Stumpner, and R. Hartman. 2023. Six decades (1959-2022) of water quality in the upper San Francisco Estuary: an integrated database of 16 discrete monitoring surveys in the Sacramento San Joaquin Delta, Suisun Bay, Suisun Marsh, and San Francisco Bay ver 7. Environmental Data Initiative. [accessed 2023 Sep 7]. https://doi.org/10.6073/pasta/8dbd29c8c22f3295bbc5d3819fb51d00
Bashevkin, S. M., S. E. Perry, E. B. Stumpner, and R. Hartman. 2023. Six decades (1959-2022) of water quality in the upper San Francisco Estuary: an integrated database of 16 discrete monitoring surveys in the Sacramento San Joaquin Delta, Suisun Bay, Suisun Marsh, and San Francisco Bay ver 7. Environmental Data Initiative. [accessed 2023 Sep 7]. https://doi.org/10.6073/pasta/8dbd29c8c22f3295bbc5d3819fb51d00
Bosworth DH, Hartman R, Bashevkin SM, Stumpner EB, Bouma-Gregson K. 2024. InteragencyEcologicalProgram/WQ-LTPublication: Manuscript-Resubmittal 1, v2.0.0. Zenodo. [accessed 2024 Jan 11]. https://doi.org/10.5281/zenodo.10493268
Bosworth, D. H., S. M. Bashevkin, K. Bouma-Gregson, R. Hartman, and E. B. Stumpner. 2024. The anatomy of a drought in the upper San Francisco estuary: water quality and lower-trophic responses to multi-year droughts over a long-term record (1975-2021) San Francisco Estuary and Watershed Science. [accessed 2024 Mar 29]. 22 (1). https://doi.org/10.15447/sfews.2024v22iss1art1
Bouma-Gregson, K., D. Bosworth, T. Flynn, A. Maguire, J. Rinde, and R. Hartman. 2024. Delta Blue(green)s: The Impact of Drought and Drought Management Actions on Microcystis in the Sacramento–San Joaquin Delta. San Francisco Estuary and Watershed Science. [accessed 2024 Mar 29]. 22 (1). https://doi.org/10.15447/sfews.2024v22iss1art2
Burdi, C. E., S. B. Slater, T. L. Bippus, and J. A. Jimenez. 2022. Mysid and Amphipod Length-Weight Relationships in the San Francisco Estuary. IEP Newsletter. [accessed 2024 Mar 11]. 40 (1):15-25. https://cadwr.app.box.com/v/InteragencyEcologicalProgram/file/860495889310
California Department of Water Resources (CDWR). 2002. Dayflow: and estimate of daily average Delta Outflow. Dayflow documentation 1997 through present., California Natural Resources Agency Open Data. [accessed 2023 Jan 4]. 25. https://data.cnra.ca.gov/dataset/dayflow/resource/776b90ca-673e-4b56-8cf3-ec26792708c3
California Department of Water Resources (CDWR). 2022. Dayflow: and estimate of daily average Delta Outflow. California Natural Resources Agency Open Data. [accessed 2023 Feb 6]. https://data.cnra.ca.gov/dataset/dayflow/resource/776b90ca-673e-4b56-8cf3-ec26792708c3
De Cicco, L. A., R. M. Hirsch, D. Lorenz, and W. D. Watkins. 2022. dataRetrieval: R packages for discovering and retrieving water data available from Federal hydrologic web services. v. 2.7.11. [accessed 2023 Jan 6]. https://doi.org/10.5066/P9X4L3GE
Godin, G. 1972. The Analysis of Tides. University of Toronto Press, Toronto.
Hammock, B. G., S. P. Moose, S. S. Solis, E. Goharian, and S. J. Teh. 2019. Hydrodynamic Modeling Coupled with Long-term Field Data Provide Evidence for Suppression of Phytoplankton by Invasive Clams and Freshwater Exports in the San Francisco Estuary. Environmental Management. [accessed 2023 Jan 4]. 63:703-717. https://doi.org/10.1007/s00267-019-01159-6
Hartman, R., E. Stumpner, C. Burdi, D. Bosworth, A. Maguire, and IEP Drought Synthesis Team. 2024. Dry me a river: Ecological effects of drought in the upper San Francisco Estuary. San Francisco Estuary and Watershed Science. [accessed 2024 Mar 21]. 22 (1). https://doi.org/10.15447/sfews.2024v22iss1art5
Hartman, R., L. Twardochleb, C. Burdi, and E. Wells. 2024. Amazing graze: Shifts in distribution of Maeotias and Potamocorbula during droughts. San Francisco Estuary and Watershed Science. [accessed 2024 Mar 29]. 22 (1). https://doi.org/10.15447/sfews.2024v22iss1art4
Moritz, S., and T. Bartz-Beielstein. 2017. imputeTS: Time Series Missing Value Imputation in R. R Journal. [accessed 2023 Jan 6]. 9.1. https://doi.org/10.32614/RJ-2017-009
| | Data Source | |
| |
|
|
