Department of Water Resources Home

Water Quality Evaluation Section

North Central Region Office

Water Quality Evaluation Section


3500 Industrial Blvd.
West Sacramento, CA 95691
(916) 376-9600

Mailing Address:
P.O. Box 942836
Sacramento, CA 94236-0001



Untitled Document

 Water Quality Evaluation Section

Central Delta Monitoring Program Metadata


I. Contact Information
Program Manager: Rolf Frankenbach
Lead contact: Tyler Salman
Environmental Scientist
Department of Water Resources - North Central Region Office
3500 Industrial Blvd., West Sacramento 95691
Office: (916) 376-9645

DWR Bryte Chemical Laboratory
1450 Riverbank Road,
West Sacramento, CA 95605
(916) 375-6008
Sid Fong: Supervisor of Bryte Chemical Laboratory

II. Study Area and Sample Site Objectives

A.  General Information

The central delta network of water monitoring and flow stations is a joint effort between the Department of Water Resources and the US Geological Survey.  The DWR Water Quality Evaluations section acquired charge of the water monitoring stations beginning in 2005.  The primary objective behind this operation is gathering data that leads to a better understanding of how State and federal water projects alter flow patterns and water quality upstream and within the delta. Since the Sacramento River is the main source of higher quality (lower salinity) water that moves through the central delta into the export facilities, monitoring critical areas of the delta where water quality is was not well defined helps to characterize the path that water takes through the northern and central parts of the Delta to Clifton Court Forebay.  The movement of low salinity water directly impacts the ability to meet delta water quality compliance criteria as well as court mandated water quality standards. These stations help define the influence that the San Francisco Bay, the Sacramento and San Joaquin Rivers and Franks Tract have on the quality of the water entering into the south delta. 

The water quality component of this project consists of nine stations that continuously monitor water temperature, specific conductance, and, more recently, turbidity and chlorophyll levels. The USGS added turbidity sensors to all stations in November 2009 and Chlorophyll sensors to three stations in February 2010.  These stations acquire spatial and temporal water quality and flow information, allowing for increased understanding of the movement of temperature and salinity through the central delta system. Stations are also used for gathering historical data and may be useful in monitoring climate change events, water year changes, as well as tidal and seasonal changes.

In addition to continuously sampling for specific conductance and water temperature at these stations, a chloride/bromide and total suspended solids (TSS) grab sample taken at each site every three weeks for laboratory analysis. Times series data from each site is used as a surrogate to indirectly determine chloride and TSS concentrations on a continuous basis, through regression analysis with direct laboratory measurement results. A grab-sample is also taken at each chlorophyll monitoring station for laboratory analysis of chlorophyll a/pheophytin a. The relationship between continuous time-series chlorophyll data and lab-analyzed chlorophyll a datais used to estimate time-series chlorophyll a values in the central delta.

Data from this project is vital to project operations in the delta (i.e. reservoir releases, gate ops, barrier placements, and exports). Information and analysis gained from this established monitoring network expand knowledge on the ways in which project operations influence flows and help to determine the zone of influence of any management actions.  Programs that the Department is leading or involved with, such as the Bay-Delta Conservation Plan, the Frank’s Tract project, FloodSAFE, and the 2-Gates project center on developing sound environmental, ecological, economical and engineering solutions that allow sustainable use of the delta in meeting various demands. Data may also be used to help address the physical factors influencing salmon migrations and to prevent excessive salvage of delta smelt and other fishes. We are starting to learn and understand how tides, freshwater flow, diversions and discharges affect overall delta water quality especially in the central part of the delta.  Water temperature, specific conductance and turbidity are three important water quality variables in describing the habitat of three pelagic species of concern, delta smelt, longfin smelt, and striped bass, inhabiting the delta.

Compliance Stations:

In 2007, North Central Region Office (NCRO) began this program in response to Federal Judge Oliver Wanger’s ruling (case number 1:05-CV-01207-OWW-GSA), that the 2005 Long-Term State and Federal Water Projects' Pumping Operations Criteria Plan (OCAP) and Biological Opinion were unlawful and inadequate in regards to the protection of the threatened Delta smelt.

The establishment and operation of two turbidity water quality monitoring stations (Holland Cut near Bethel Island and Victoria Canal near Byron) by the Water Quality Evaluations Section of NCRO were included as court-ordered compliance stations that would trigger restrictions to both State Water Project and Central Valley Project operations in the Sacramento-San Joaquin Delta in an effort to reduce salvage of delta smelt. The compliance period begins December 25 – January 15 and if the turbidity levels at either one of these stations exceeds 12 NTU, and flows measured at Sacramento River at Freeport do not exceed 80,000 cubic feet per second (cfs), the Federal and State Water Projects must reduce exports for 10 days to insure net upstream Old and Middle River flows are less than 2,000 cfs (Natural Resources Defense Council v. Dirk Kempthorne).
The Department of Water Resources (DWR) has been monitoring water quality as part of the South Delta Temporary Barriers project since 1991 to investigate water quality conditions in the South Delta that may be affected by temporary barrier installations and operations.  The North Central Region Office (NCRO) water quality sampling program consists of two components: 1) bi-monthly discrete sampling; and 2) continuous sampling.  NCRO staff analyzes the data and presents their findings in the water quality chapter of the Temporary Barriers Monitoring Report.  The information collected by this program is also required for compliance with a 401 Water Quality Certification issued by the Central Valley Regional Water Quality Control Board.  For detailed information on the South Delta Improvements Program and the Temporary Barriers Project please visit DWR’s Bay-Delta Office website at  

B.  Name and Location Information for Continuous Water Quality Sampling Sites

-Currently and Historically Sampled Sites

-Central Delta Stations Map –

C.  Station Period of Record, Instrumentation, and Parameter Collection History

-Central Delta Stations Table –

D.  CDEC and Water Data Library Station Codes

-CDEC and WDL Station Codes Table –

III. Sampling Frequency

A. Sonde Instrument Sampling

There are a total of 9 Central Delta water quality monitoring stations that collect time-series data at 15 minute intervals using a YSI 6600 internal data logging sonde.  All stations record a single measurement of temperature (Cº), specific conductance (µS/cm), and turbidity (NTU), within the 15 minute interval.  The following stations also measure chlorophyll (µg/L): Old River near Franks Tract, False River near Oakley, and Middle River Near Holt.  The San Joaquin River at Blind Point and Tuner Cut near Holt monitor the following constituents: water temperature (Cº), dissolved oxygen (%), pH, specific conductance (µS/cm), turbidity (NTU), and chlorophyll (µg/L).

B. Field Instrument Sampling


Upon each site visitation (can vary from weekly to every 3 weeks) , calibrated handheld field instruments are used to sample current water conditions and compare water quality  measurements to all the YSI 6600 or 600 continuous monitoring stations. Temperature (Cº) and specific conductance (µS/cm) measurements are field verified with an YSI 63 Handheld pH and Conductivity Instrument.  Dissolved oxygen measurements are field verified with a YSI Professional Optical Dissolved Oxygen Meter. Turbidity measurements are field verified with a Hach 2100P Portable Turbidimeter.

C. Discrete Sampling           

Staff collects a chloride/bromideand total suspended solids samples in separate plastic quart bottles at 1 meter depth with a Van Dorn sampling device at each site.  The chloride/bromide and total suspended solids samples and other field measurements are sampled at a 1 meter depth since the YSI 6600 sondes are also sampling at this depth.  During each field run, DWR staff also collects a duplicate chloride/bromide and total suspended solids sample at one of the stations to test for field and lab precision and repeatability.  Staff also collects a chlorophyll a/pheophytin a sample at the following stations: Old River near Franks Tract, False River near Oakley, Middle River Near Holt, Turner Cut near Holt, and San Joaquin River at Blind Point.

IV. Field Methods

Field measurements

During each site visit field readings are taken using portable and handheld instruments.  These handheld instruments are calibrated to comply with manufacturers or laboratory specifications.  Field readings are taken and recorded for comparison with YSI sonde data.  Field measurements are taken at each site at a 1meter depth to duplicate depth of deployed YSI sondes. 

YSI 63 Handheld pH and Conductivity Instrument

The YSI 63 Handheld Instrument measures temperature, conductivity, pH, and salinity.  The instrument is calibrated once a month for conductivity and pH is calibrated before each field run.


YSI Professional Optical Dissolved Oxygen Meter

The YSI Professional Optical Dissolved Oxygen meter measures percent saturation and mg/L of dissolved oxygen.  The instrument is checked and calibrated before each field run.

HACH 2100P Portable Turbidimeter

The HACH 2100P turbidimeter measures turbidity optically using the nephelometric principle.  The portable instrument is calibrated every three months using manufactures StablCal standards to assure accuracy.

Sample Collection

Staff collects discrete chlorophyll a/pheophytin a, chloride/bromide, and total suspended at 1 meter depth with a Van Dorn at each monitoring station.  One sample field run is duplicated, on a rotating basis, for quality assurance/quality control testing at the lab.

      C. Sample containers and holding times

The DWR Bryte laboratory supplies all necessary sampling materials to the NCRO Water Quality Evaluation field staff. Requirements for sample containers, preservation techniques, and holding times are found in one of the following references (or later editions): Standard Methods for the Examination of Water and Waste Water, American Public Health Association, et al., 21st Edition; Handbook for Sampling and Sample Preservation of Water and Wastewater, 2005, and 1979 US EPA Manual entitled “Methods for Chemical Analysis of Water and Wastes,” EPA-600/4-79-020.

V. Laboratory Methods

All water quality analyses are conducted by staff at the DWR Bryte Laboratory. Located in West Sacramento, the lab’s primary function is to analyze drinking water, surface water, groundwater, and wastewater. Bryte lab has maintained certification by the Environmental Protection Agency and the California Department of Health Services for water analysis since 1978. It also provides quality assurance, and related technical services.
Bryte Laboratory Method Chart for Discrete Sample Analyses

-Lab Method Chart –

VI. Data Management

A. Sonde Data

Post-deployment Quality Assurance
After the YSI 6600 sondes are removed from the field, DWR staff perform the following two procedures to check whether the sondes are still operating properly and measuring accurately:
A post-deployment accuracy check on the day the sondes are removed and before the instruments are cleaned
A comparison between the data measured by the handheld field instruments and the data collected by the sonde at the closest 15 minute time interval

The accuracy of sonde probes while deployed in the field can be negatively affected by probe malfunction, drift away from initial calibration, and/or fouling caused by biological growth on the probe reading surface.  DWR staff perform the post-deployment accuracy check by the following procedure prior to cleaning the sonde probes:

  • Placing the sonde probes in fresh calibration standards with known values

Operating the sondes in the standards and recording the values the sondes are reading
Rating the values collected during the accuracy check for each constituent as excellent, good, fair, or poor based on their deviation from the calibration standard according to the USGS technical report “Guidelines and Standard Procedures for Continuous Water Quality Monitors-Station Operation, Record Computation, and Data Reporting” (Wagner et al., 2006)
The ratings obtained from the accuracy check indicate the quality, accuracy, and reliability of the data that the sonde collected while in the field. 

In addition to the post-deployment accuracy check, DWR staff compare the water temperature, specific conductance, pH, dissolved oxygen, and turbidity data measured in the field by the handheld instruments (the YSI-63, YSI Pro-ODO, and HACH 2100P) to the sonde data that is closest in time.  While taking the field measurements, DWR staff attempt to collect the field readings at the same depth that the sonde probes are measuring at (1 meter) and as close to the sonde pipe as possible.  Since the field instruments are calibrated regularly, a large difference between the sonde and field readings could indicate inaccuracy of the sonde data during the deployment period.  DWR staff consider these comparisons between the field and sonde readings and the ratings obtained from the post-deployment accuracy check while assessing data quality when entering the continuous data into the Hydstra database. 


Data Quality Assurance/Quality Control (QA/QC)
DWR staff import data files from sondes into the North Central Region Office (NCRO) Hydstra database where additional QA/QC procedures are performed.  In addition to documenting the results of quality assurance procedures discussed in the previous section, staff use the results of these procedures to flag any suspect or unreliable data.  Any obvious outliers in the continuous dataset due to fouling or other factors are flagged as unreliable.  No data that has been determined by DWR staff as suspect or unreliable were used in this chapter; only reliable data of known good quality were used.  The reliable and good quality data in Hydstra are used to populate the Water Data Library where the data for all the continuous sites are available online at

Chlorophyll a estimation
After all of the chlorophyll data was compiled, DWR staff then used the Minitab statistical software to analyze regression relationships for the matched chlorophyll data pairs.  Each of the 13 continuous monitoring locations was analyzed individually since the relationship between lab and sonde data is specific to location.  Each regression analysis generated an equation describing the relationship between sonde and lab chlorophyll data for the particular location.  DWR staff then used these equations to adjust the chlorophyll concentrations from the sonde to more closely represent chlorophyll a concentrations.  The regression analysis procedure is described in the following steps:

  • A simple linear regression analysis is performed with the sonde data as the explanatory variable (x-variable) and the laboratory data as the response variable (y-variable).  Three assumptions of this parametric regression procedure are: the data follow a linear pattern, the underlying distribution of the data follows a normal or bell-shaped curve distribution, and the variance of the residuals from the regression is constant.  If these three assumptions are met, then the equation from the linear regression analysis can be used to adjust chlorophyll a concentrations.  If not, move on to step #2. 

If the data do not follow a linear pattern, the explanatory variable (sonde data) needs to be transformed so that this assumption is met.  If the variance of the residuals is not constant or the underlying data is not normally distributed, then the response variable (laboratory data) needs to be transformed.  A typical transformation that is effective for this data is the natural logarithm.  Once the response variable is transformed, the regression equation no longer predicts the mean chlorophyll a concentration; it predicts the geometric mean or median.  In order to correct for this, either of two methods can be used:  the Maximum Likelihood Estimator (MLE) or Smearing.  These methods are described more in Step #4.  If it is not possible to transform the data so that the three assumptions of a linear regression are satisfied, then a nonparametric regression needs to be used, which is described in Step #5.
If the equation from either the simple linear regression (Step #1) or the linear regression with transformed data (Step #2) is going to be used, then the seasonal terms, sine and cosine, are added to the regression analysis to determine if they are good predictors of the seasonality of chlorophyll concentrations.  If one or both of the seasonal terms are statistically significant in the analysis, then both terms are added to the regression equation in order to incorporate seasonality into the equation.
MLE is one of the bias correction methods used to estimate the mean concentration when using a regression equation with the response variable transformed to a natural logarithm.  The MLE is calculated by the following equation:

  • MLE = e(0.5*MSE) 
  • The MSE is the mean squared error in logarithmic units, which is a quantification of the difference between the true data and the data estimated by the regression equation.  The adjusted chlorophyll a data generated by the regression equation with a natural logarithm transformation is then corrected by multiplying the adjusted value by the MLE.  For most of the stations that had a regression equation with data transformed to natural logarithms, the MLE was used as the bias correction factor to estimate the mean chlorophyll a concentrations.  However, for one of the stations the Smearing correction factor was used, since it was the better estimator .  The Smearing factor is calculated by first using the natural exponent (e) to “back-transform” all of the residuals from the regression.  The factor is then the average of all of the “back-transformed” residuals.  Like with the MLE, the adjusted chlorophyll a data is multiplied by the Smearing correction factor to correct the values to the estimated mean concentrations.

If transforming the data doesn’t allow for the assumptions of a linear regression to be attained, then the Theil-Sen line, a nonparametric regression procedure, can be calculated.  Like with the linear regression when the response variable is transformed to the natural logarithm, the Theil-Sen line equation predicts the median chlorophyll a concentration.  However, there is no bias correction factor available to estimate the mean concentrations when using a nonparametric regression procedure.  Therefore, when summarizing chlorophyll a concentrations adjusted with the Theil-Sen equation, averages such as daily or monthly averages cannot be reliably calculated.

B. Field Data

Field measurements are recorded on field data sheets upon each site visitation, and if the station has a discrete grab sample component, the data is then entered into the field module of DWR Bryte Laboratory’s “Flims” database.  Laboratory analyses are performed by DWR Bryte Laboratory on the discrete grab samples and the results are entered into the “FLIMS” database at the lab. For those stations without a discrete sample component, the field data sheets are used for sonde data comparison and archived accordingly.   On a regular basis the data from “Flims” database is loaded into DWR’s Water Data Library (WDL) database for public viewing and access.  Lab results are further archived and used for data report analyses.

  The decision to use the MLE or Smearing correction factor was determined by the following procedure.  The sonde chlorophyll data that was matched to the lab chlorophyll data was plugged into the regression equation and then corrected by the MLE factor to provide a predicted sonde chlorophyll concentration.  These predicted values were then matched with their corresponding lab chlorophyll values, and a linear regression was performed on the matching pairs.  The same procedure was used with the Smearing correction factor.  The correction factor that gave a regression equation with a slope closest to one was then used.