MD DNR Water Quality Mapping Project 2020

WATER QUALITY CALIBRATION SAMPLES: At each calibration station, "grab" water quality samples were collected from the outflow of the DATAFLOW unit. "Grab" samples were collected at the same time as the Hydrolab surface sample was recorded. Numbered two quart bottles were triple-rinsed and filled with water for chlorophyll and total suspended solids samples. Chlorophyll and suspended solid water samples were filtered on station or shortly thereafter. Sample waters and filters were placed on ice immediately after filtration. HYDROLAB PROFILE: The first reading of the Hydrolab water-column profile at each calibration station was recorded at the same time the water quality bottle sample was collected. The first Hydrolab record logged was for the 0.5 m depth. The sonde was then lowered to the bottom. A reading was taken at 0.3 m above the bottom. The sonde was raised and measurements were recorded at 0.5 m or 1.0 m increments until it reached the surface. (In cases where station depth was greater than 3 m, the sonde was raised in 1 m increments). SECCHI DEPTH: Secchi disk depth was measured at each calibration station. Readings with the Secchi disk were made in situ without the aid of sunglasses. The Secchi disk was lowered into the water, on the shady side of the boat, and the depth at which it was no longer visible was recorded. The Secchi depth reading was taken near the stern of the vessel, and the time at which the reading was taken was noted (to the second) from the Global Positioning System. This facilitated later matching of Secchi depth readings with turbidity probe data. PAR MEASUREMENT: Underwater Photosynthetically Active Radiation (PAR, 400-700nm) At each calibration station, down-welling light penetrating the water column (PAR) was measured underwater at several depths to calculate the light attenuation coefficient, Kd. Simultaneous deck and submersed PAR intensity measurements were taken to account for variability in incident surface irradiance due to changes in cloud cover. Data collected using this procedure were used to estimate the depth of the photic zone. The equipment used was manufactured by LI-COR, Inc. and consisted of a LI-192SA, flat cosine Underwater Quantum Sensor, a LI-190SA air (deck) reference sensor and a Data Logger (LI-1000 or LI-1400). Deck and underwater readings were recorded simultaneously. Readings were allowed to stabilize before being recorded. If the station depth was less than 3 meters, readings were taken at 0.1 meter and at 0.25 m intervals until 10% of the 0.1 m reading was reached. If the station depth was greater than 3 m, a reading was taken at 0.1 m and at 0.5 m intervals until 10% of the 0.1 m reading was reached. SONDE PROBES: YSI 6600 data sondes equipped with a 6560 conductivity/temperature probe, a 6136 turbidity probe, a 6025 chlorophyll probe, a 6561 pH probe and a 6150ROX (Optical) dissolved oxygen probe were maintained and calibrated before and after each deployment in accordance with YSI recommendations <http://www.ysi.com/resource-library.php>. SMALL BOAT SURVEYS: DATAFLOW is a compact, self-contained surface water quality mapping system, suitable for use in a small boat operating at planing speeds of about 25 knots. The system collects water through a pipe ("ram") deployed on the transom of the vessel, pumps it through an array of water quality sensors, and then discharges the water overboard. Orientation of the sonde vertically, with probes upward, ensures that no air bubbles are conveyed to the sensors, preventing errors that might be caused by such bubbles. Water quality instrumentation consisted of a YSI 6600 Sonde equipped with a flow-through chamber. The system was configured with conductivity/temperature, turbidity, chlorophyll, pH and dissolved oxygen probes. Positioning and depth instrumentation consisted of a Raymarine A70D Chartplotter/Sounder. The data logger matched the position data with water quality sensor data for each observation. The Raymarine A70D GPS transmitted NMEA data to a Panasonic ToughBook(tm) or Dell Latitude Rugged Extreme. A DATAFLOW/LabVIEW program was used to merge position and depth data with data collected by the logger and create an output file. The system was equipped with an inline flow meter. Although the flow rate did not affect sensor readings, decreased flow was an indication of either a partial blockage or an interruption of water flow to the instrument. Flow data were used in the field as a diagnostic tool to ensure that the system was working properly and, later, as a quality assurance tool to verify that water flow was uninterrupted. A boat horn was wired to the flow meter. If the flow-rate fell below 3.0 L/s, the horn sounded and warned operators that a problem needed to be corrected. Cruise tracks varied depending on the water body being mapped. In general, a square-wave pattern was followed by alternately sampling shallow shoreline areas, and open, deeper waters while traveling up and down river. Alternative cruise paths were followed if water body size, shape impediments, or obstructions dictated otherwise. Cruise patterns were selected to obtain representative coverage of shallow water habitats and open waters so that segment-wide criteria could be assessed as accurately as possible. Navigational issues and placement of representative calibration sites also determined ultimate cruise tracks. LARGE VESSEL SURVEYS: Maryland Department of Natural Resources has used large vessels to collect monthly water quality samples at fixed Chesapeake Bay mainstem stations since 1984. When the water quality mapping program began surveying Chesapeake Bay segment CB4MH in 2017, it was decided that DATAFLOW sonde data could be collected as the mainstem survey sampling vessel proceeded from station to station. The larger survey vessels are built with through-hull fittings (sea-cocks) located 0.5m below the waterline. The fittings are means of safely controlling the flow of water from outside the vessel hull into the vessel. At the beginning of each survey, the DATAFLOW sonde was activated. Bay water was flowed from the sea-cock across the DATAFLOW sonde sensors in the YSI Flow-cell. A vessel navigation log file was started and vessel coordinate and speed values were acquired from the vessel geographic positioning system. At the end of the survey, the navigation file was copied and uploaded to a data server. After sonde post-calibration, the sonde file was copied and uploaded to a data server. Larger vessels served as survey platforms during the 2020 water quality mapping season, RV Kerhin and RV Rachel Carson. Vessel length was 49 feet. A program named GPSU was used to log navigation data from a Simrad HS70 GPS compass, and depth readings from a Furuno FCV 1100 L echo sounder.

SONDE FILE AND NAVIGATION FILE MERGE: The output of the DATAFLOW/LabVIEW system used on small boat surveys contains merged time, coordinate and sonde results. When the large vessels were used, the LabVIEW component was not used and additional post-processing work was required to merge navigation log files with the sonde files into a single file. The process was complicated because the navigation log file and the water quality data sonde files were each started at different times and configured with different time increments. While some navigation and sonde records aligned at precisely the same hour:minute:second timestamp, other navigation and sonde record timestamps were close but differed by one or more seconds. In order to align the sonde timestamps with the navigation timestamps, time offset expressions were used. For example if the sonde and navigation timestamps were 1 second off, a 1 second timestamp offset was used to align the records. Offset values were documented. Often, timestamps matched and no adjustment was required. In other instances, 1, 2 or 3 second adjustments were employed to align the navigation and sonde data. During a survey lasting approximately 7 hours, a 3 second time offset was considered to be sufficiently geographically-precise for the purposes of the project. Each month, the raw large vessel navigation log file, the raw water quality sonde data file and timestamp bridge file were referenced in a Microsoft Access(tm) database. The bridge timestamp file contained one second interval time records from 06:00:00 through 23:59:59. Database queries were used to populate a table that joined navigation records with timestamp bridge records, and a table that joined water quality sonde records with timestamp bridge records. Then, a final query was used to join the navigation and sonde tables and the result set was saved as a Microsoft Excel(tm) workbook. A series of steps were executed on Excel(tm) workbook worksheets that selected a subset of the record set generated using the Access(tm) database. The first worksheet served to document the raw, merged navigation-bridge-sonde records set. The second, based on a copy of the first, used sorting to select sonde and navigation data time stamps that matched exactly. The third worksheet, also based on a copy of the first, prepared records for the fourth worksheet. In the fourth worksheet, sonde values were shifted up one row, in relation to navigation record timestamp values. The number of seconds-shifted (offset) needed to align navigation values and sonde values in worksheet rows, was documented in the fourth worksheet. Finally, the last worksheet, combined water quality sonde and navigation values from the second and fourth worksheet results, including time-shift offset values ranging from 0 seconds to 3 seconds. At this point, the larger vessel merged time, coordinate and sonde results had been transformed to a stage where they were ready for processes described under the heading: DATAFLOW FILE POST-PROCESSING. DATAFLOW FILE POST-PROCESSING: Each file was opened in Microsoft Excel(tm) and renamed using a survey-date naming convention. Rows of data acquired before and after mapping were deleted. Records (if any) were also deleted if they did not have associated GPS values. A macro was executed that rearranged columns and inserted error-tracking columns and headings. Next, negative values were flagged, and values outside each parameter's normal range were highlighted. The macro also returned a form summarizing exceedances. Finally, mapping cruise event and instrument information were appended to each record. Flagged values were evaluated for common anomalies, including spikes in fluorescence and turbidity, dips in specific conductance, and unusually high dissolved oxygen readings. Instrument post-calibration results, in situ comparisons with HydroLab, LI-COR readings, historical data from nearby locations, and survey crew remarks were used to determine whether sensor values were acceptable. In cases where data were determined to be unreliable, the reason(s) values were determined to be erroneous were documented with error codes and comments. Unreliable data were masked. No data were discarded. All DATAFLOW data for each mapping cruise, both valid and erroneous, were retained in an archival file. Only data considered reliable were published in reports. VERIFICATION AND DATA MANAGEMENT: At the end of the monitoring season, DNR Tawes Office and Field Office personnel conducted additional data QA/QC procedures. All of the water quality calibration "grab" sample data were plotted, and outliers and anomalous values were thoroughly researched. Staff compared unusual values to historic values from the site and values from nearby sites in the Bay. Weather events were considered, event logs were reviewed and field staff members were consulted regarding possible legitimate causes for outlying values. In cases where values were not considered to be legitimate, they were masked from the published dataset with the approval of the field staff and the Quality Assurance Officer.

LABORATORY ANALYSIS - CBL University of Maryland's Chesapeake Biological Laboratory (CBL), Nutrient Analytical Services Laboratory (NASL) analyzed chlorophyll, phaeophytin and total suspended solids. Further information about laboratory analytical procedures may be obtained from the "Process_Contact".