Continuous monitoring of water-level, stream velocity, and specific conductance combined with periodic discharge measurements and collection of water-quality samples were required to compute the discharge and nutrient flux from the five estuarine river stations. Stream velocity (index-velocity) data were used to estimate the mean cross-section channel velocity (mean velocity) that was measured using a vessel-mounted acoustic Doppler discharge measurement system. Stream velocity data were used with the measured mean channel velocity to develop index-to-mean velocity relations. Each station was equipped with similar instrumentation. Data collected at each station were reviewed for accuracy and entered into the USGS NWIS database. Data for each station were then used to develop regression equations that were used to estimate water discharge and nutrient flux.

Water-Level and Velocity Measurements

Water level was measured at each station using Design Analysis Associates H-310 vented-submersible pressure sensors that were mounted within 2-in. inside-diameter polyvinylchloride (PVC) pipes with end caps, which functioned as stilling wells. Quarter-inch holes were drilled around the perimeter of the PVC pipe within 6 in. of the bottom. Pressure sensors were programmed to average over an 8-second period every 15 minutes. Distances to water-level surfaces were measured from fixed reference points every 4 to 6 weeks and compared to the water levels measured by the pressure transducers. Water-surface reference measurements varied less than 0.02 ft from the pressure transducer water levels during the study.

Two types of vertically oriented acoustic Doppler velocity systems were used to measure 2-min averages of velocity through the water column every 15 minutes:

(1) SonTek ADP 1.5 and 3 MHz velocity profilers, and

(2) SonTek Argonaut-XR 3 MHz depth-averaging velocity sensors.

Vertically oriented velocity profilers were used because vertical stratification of flow was possible, and because these sensors allow variations in the velocity profiles in the water column to be measured during all conditions. It has been suggested (Chiu and Said, 1995) that if the maximum velocities in an open channel can be measured, then all other velocities in a cross section can be theoretically estimated using channel geometry. Therefore, the use of vertically oriented velocity systems should allow the systems to be placed in regions of maximum velocities, where the maximum amount of information about open-channel flow is contained (Chiu and Said, 1995). Theoretically, the most reliable relations between index velocity and measured-mean velocity could be acquired using this method. Unfortunately, maximum flood and maximum ebb velocities often do not occur in the same region of the river cross section. However, repeatable relations between index and mean velocities can be calculated despite this characteristic.

One drawback to using a vertically oriented index-velocity system is the limited amount of across-channel volume that is measured. In rivers and streams with variable cross-channel flow distribution, this could prove to be a problem. If water flow is relatively uniform and flow patterns are always the same across the channel, vertically oriented sensors can provide an alternative to horizontally oriented index-velocity systems.

Discharge Measurements

River discharge was measured directly using a vessel-mounted acoustic discharge system (Simpson and Oltmann, 1992). An RD Instruments 1.2 MHz acoustic Doppler discharge measurement system was used to measure discharge for calibration and validation of the index-to-mean velocity relation at all stations every 4 to 6 weeks. The discharge section edges were marked with buoys and unmeasured edge section distances were measured using optical and laser range finders. Discharge values were calculated using manufacturer's software. Individual discharge measurements took between 4 to 10 minutes to complete.

Discharge measurements were collected over varying ranges of tidal level and tidal phase for about one year to develop an index-to-mean velocity relation for each station. Multiple measurements were made during 3- to 6-hour periods to better characterize variations in flow caused by variations in tide, inflow, and wind effects, and to reduce serial correlations between measurement sets. Discharge data collected after the calibration period were used to check the accuracy of the water-level/velocity/discharge relations for the duration of the study and to apply corrections or shifts to the index-to-mean velocity relations if required. Corrections to the velocity regressions were not required at four of the five stations during the study. The Harney River station index-velocity data were corrected for June 1998 because of an electronics problem that caused the index-velocity data to be biased low.

Water-Quality Sampling

Water-quality samples also were collected every 4 to 6 weeks during trips for discharge measurement and equipment maintenance. Daily fluctuation in water quality was assumed to be represented by samples collected near flood maximum and ebb maximum based on previous studies in west-central Florida (Stoker and others, 1995, 1996). Water-quality samples were collected for the determination of total nitrogen and total phosphorus concentrations and to estimate total nitrogen and total measured phosphorus flux at the southwest coast stations. Samples were collected at each station at least once daily, and typically were collected during flood, ebb, and slack tides.

Depth-integrated samples were collected at three cross-stream locations using a stainless-steel-weighted Teflon^TM bottle. A modified downrigger was used to regulate descent and ascent speed of the weighted Teflon^TM bottle. Depth-integrated water samples from the three cross-stream locations were combined and mixed in a polyethylene churn and then distributed to individual sample bottles. Water sample bottles were then bagged in plastic and placed in a cooler with ice. All water-quality samples were analyzed at the USGS laboratory in Ocala, Fla. Water samples were analyzed for total and dissolved ammonia, ammonia-plus-organic nitrogen, nitrate-plus-nitrite, nitrite, phosphorus, and orthophosphate. Analytical methods used in this study are documented in Fishman and Freidman (1989). Field measurements of water temperature, specific conductance, dissolved oxygen, and pH were collected concurrently during water-quality sample collection at each location to identify cross-stream and vertical variability.

Approximately 20 percent of the water samples collected were field quality-assurance samples. Two types of quality-assurance samples were collected: (1) duplicate samples and (2) equipment blanks. Field quality-assurance samples were sent to the laboratory with routine samples. Field measurement sensors were calibrated at the beginning of each day for each parameter.

In-situ temperature and specific conductance sensors were used to collect 15-minute-interval data near the water surface and at the bottom of the water column at each station. Near surface temperature and conductance sensors were removed after 3 years at the Broad, Harney, and Shark River stations because data indicated the rivers to be well-mixed from the top to the bottom of the water column. In-situ temperature and specific conductance sensors were checked for accuracy every 4 to 6 weeks both before and after cleaning using three conductance standards. Any adjustments to data from the sensors based on field calibrations were made using variable shifts in the USGS Automated Data and Processing System (ADAPS).