Pressure Sensor Offset After a Freshwater Flood
A PORTS water-level station in a brackish estuary uses a vented pressure transducer as primary and a co-located acoustic gauge as backup. Over 10 days, the primary reads 3–5 cm lower than the backup at low slack, but the two agree within 1 cm at high tide. The pressure sensor was calibrated 8 months ago; the acoustic gauge 2 months ago. Six days ago, a tropical storm dropped 18 cm of rain and reduced surface salinity from 12 to 3 PSU. The station's public display shows a single green verified status for all water-level data. The pilots use the display for under-keel clearance, and a dredging contractor uses it for bucket-depth control. The technician has conductivity data at the station but no independent water-level measurement since the storm. What is the likely cause, and what QC and communication actions are appropriate?
1. What is the most likely cause of the offset, and how would you test that hypothesis?
Read the worked answer
The pattern points to a water-density effect rather than simple sensor drift. A pressure transducer measures pressure and converts it to water level using an assumed density. After the storm, low-salinity water at low slack has a lower density than the assumed value, so the pressure for a given true level is lower and the computed level underreads. At high tide, saltier water may dominate and the assumed density is closer to reality, so the two gauges agree. Test this by comparing the offset with conductivity and temperature at the same times, and compute the density correction. If the corrected pressure record matches the acoustic gauge, the hypothesis is supported. If the offset remains after density correction, inspect the pressure port and vent tube for fouling, sediment, or partial blockage. The 8-month calibration interval alone does not explain a tide-correlated pattern, and the green verified status is not justified without this check.
2. What QC flags and display changes should be made while the investigation is underway?
Read the worked answer
Change the station status from verified to provisional or suspect for the affected period. Do not leave a single green status that implies all values are equally trustworthy. The acoustic gauge can serve as the operational reference if it is independently checked, but it should also carry a note that the primary is under review. Apply a suspect flag to the pressure record from the onset of the offset and retain the raw values. If density correction is applied, document the method, the conductivity source, and the time period, and use a corrected or estimated flag rather than overwriting the original data. Notify the pilots and dredging contractor directly because both use the display for safety and depth control. A web banner alone is insufficient for under-keel clearance users. Do not apply a constant offset to make the curves match; the error varies with tide and salinity.
3. What corrective and follow-up actions should be taken before returning the primary to verified status?
Read the worked answer
First, complete the density analysis using the station conductivity and temperature data. If density explains the offset, either implement a documented density correction in processing or designate the acoustic gauge as primary until the pressure processing is revised. Second, inspect and service the pressure sensor, vent tube, and intake for fouling or sediment after the flood. Third, perform a local comparison with a tide staff or independent reference over several tidal cycles. Fourth, review the calibration history and consider recalibration if inspection finds no cause. Fifth, update metadata to describe the event, the correction, and the QC flags. Only after these steps should the record be considered for verified status. The follow-up should also check whether the same density issue affects other stations in the estuary and whether the public display's status logic needs to distinguish verified, provisional, and suspect data.