Groundwater profiler data quality checks for pumping-induced drawdown artifacts

Groundwater profilers can reveal sharp changes in turbidity, suspended solids, conductivity, temperature, redox potential, and other indicators with depth. The same instruments can also record misleading features when a bore is being pumped, purged, sampled, or developed. A pressure response caused by abstraction may be mistaken for a natural aquifer boundary, while disturbed sediment can appear to be a discrete contaminant layer.

Reliable interpretation depends on treating pumping conditions as part of the measurement record. Flow rate, pumping history, stabilisation time, profiler movement, sensor response, and well construction should be reviewed alongside the water-quality data. This is especially important in Australia, where monitoring programs may span shallow urban aquifers, irrigation bores, coastal groundwater systems, and remote Great Artesian Basin sites.

Why pumping creates apparent drawdown features

Pumping lowers hydraulic head around the bore and establishes a transient flow field. Water moves towards the screened interval, while water from other depths may enter the well through the screen or open-hole section. In a long-screened bore, this vertical flow can redistribute water before the profiler is lowered. A concentration peak measured at a particular depth may therefore reflect mixing inside the bore rather than the formation at that depth.

The problem becomes more pronounced when pumping begins shortly before profiling. Wellbore storage can delay the hydraulic response, and the initial discharge may carry fines released from the screen, gravel pack, or bore wall. Turbidity and suspended-solids readings can rise quickly, while conductivity or temperature changes more gradually. If the pump intake is close to the profiling interval, the instrument may record a local drawdown signature, a moving interface, or a sediment pulse generated by the pumping operation.

Pumping can also alter the physical conditions around an optical sensor. Increased velocity may entrain air, create bubbles at fittings, or dislodge deposits from the bore casing. These effects can produce short-lived spikes or step changes that resemble a groundwater plume. A profiler moving through a strong vertical current may also experience cable drag and unstable depth, causing measurements to be assigned to the wrong elevation.

Establish a defensible baseline

Before interpreting a profile, record the bore identification, coordinates, screened interval, casing diameter, total depth, static water level, and recent pumping history. The field sheet should include pump type, intake depth, discharge rate, start and stop times, purge volume, sampling method, and any changes in flow. A water-level logger positioned independently of the profiler can provide a useful reference for drawdown and recovery.

The baseline should include at least one profile collected under stable conditions, where practical, together with repeat measurements at selected depths. A downcast and upcast comparison can show whether a feature follows the water column or is tied to instrument movement. If the profiler is lowered too quickly, sensor lag may shift a sharp boundary downward on the downcast and upward on the upcast. That hysteresis is a measurement effect, not evidence of two separate layers.

Depth control deserves the same attention as sensor calibration. Confirm the zero reference, cable markings, winch encoder, and water-level datum. In a deep production bore, the apparent depth may differ from the elevation used in a groundwater model. Tide, barometric pressure, and nearby abstraction can also affect water levels. In Perth’s coastal aquifers, for example, seasonal pumping and recharge can change the hydraulic gradient over relatively short periods, so a profile should be time-stamped precisely.

Recognise the signature in profiler data

A pumping-related artefact often has a recognisable temporal pattern. It may begin soon after the pump starts, intensify while the flow rate increases, and decay during recovery. A natural stratigraphic boundary usually remains at a consistent elevation across repeated passes. A drawdown-induced feature may move, broaden, or disappear as the bore returns towards static conditions.

Compare all available channels rather than relying on turbidity alone. A genuine change in suspended solids may coincide with a conductivity, temperature, dissolved oxygen, or redox response. An optical spike without a corresponding hydrochemical response is more likely to reflect bubbles, sensor fouling, cable movement, or a transient sediment disturbance. Conversely, a coherent change across several sensors can indicate that water from a different zone has entered the bore.

Look for relationships between the water-level trace and the profile. If a sharp conductivity break appears only during maximum drawdown, treat it as provisional. If the feature is reproduced at the same depth after recovery, confidence increases, although well mixing can still preserve a temporary signal. A useful quality-control plot places sensor value, depth, elapsed time, pump status, and water level on aligned axes. This often exposes an apparent vertical trend that is actually a time-dependent response.

The shape of the response also matters. A broad, smooth shift may reflect mixing or sensor equilibration. Narrow spikes may indicate particles or bubbles passing the optical path. Repeated oscillations can result from pump cycling, generator fluctuations, or movement of the profiler in a turbulent bore. The instrument’s sampling interval and response time should be considered before assigning significance to features only a few centimetres thick.

Check the optical measurement chain

Optical turbidity and suspended-solids sensors are sensitive to particle size, colour, shape, and concentration. A pumping event can change all four. Fine clay may remain suspended and produce a strong optical response, while coarser sand settles quickly and causes a brief pulse. The same nominal concentration can produce different readings when the particle population changes, so turbidity should not automatically be converted to mass concentration using a universal relationship.

A field audit using representative water samples can test whether the sensor response remains credible under the site’s actual conditions. The process described in this field accuracy audit is relevant when pumping has produced an unusual profile and grab samples can be collected safely. Samples should be linked to exact depths and times, with preservation, transport, laboratory method, and chain-of-custody details recorded.

Inspect the sensor before and after each run. Check the optical windows, wiper, protective cage, cable connection, battery level, and internal diagnostics. Rinse with site water between depths when the operating procedure permits, while avoiding a cleaning action that removes the very material being investigated. A zero-water check and a stable reference standard can identify drift, but they cannot reproduce every field effect caused by bubbles, flow velocity, or particle type.

Sensor selection also affects how a pumping response should be interpreted. Point-source measurements represent conditions near the sensing volume, whereas integrating arrangements average across a longer path. The practical distinction is explained in this discussion of sensor measurement choices. A narrow point-source spike may be real at the sensor location, while an integrating measurement may smooth it into a modest broadening of the profile.

Separate hydraulic response from instrument artefact

A controlled sequence is usually more informative than a single survey. Where site conditions allow, collect a static or low-disturbance profile, record the pumping rate, profile during the planned pumping period, and repeat during recovery. Keep the profiler speed, depth increments, sensor orientation, and logging interval consistent. The sequence should be designed around the bore’s expected recovery time rather than an arbitrary waiting period.

For a monitoring program in an irrigation area of the Murray–Darling Basin, pumping may be intermittent and demand-driven. A profile collected after a nearby irrigator starts a pump could show a genuine regional head response, local inflow, or bore-specific disturbance. The field team should identify nearby abstraction where possible and compare the hydrograph with pump telemetry, operational records, and regional groundwater data.

A similar issue occurs in Great Artesian Basin work, where remote logistics can encourage teams to profile immediately after purging to save a site visit. That approach may be necessary, but the resulting dataset should be labelled as pumped or disturbed rather than combined uncritically with static profiles. In a managed monitoring program, the metadata can preserve the value of the observation without overstating its representativeness.

Use recovery behaviour as an additional discriminator. A hydraulic response generally follows a physically plausible recovery curve and may be observed by an independent pressure logger. A fouling event, bubble passage, or cable snag often produces an abrupt sensor-specific change with no matching water-level behaviour. If the signal remains in one channel while pressure, conductivity, and temperature remain steady, flag it for investigation rather than treating it as a formation property.

Apply a repeatable quality-control workflow

Data validation should begin before the profiler enters the bore. Confirm calibration status, clock synchronisation, firmware settings, depth datum, battery capacity, and available memory. Review the risk of sediment mobilisation from the pump, previous development, or a poorly sealed casing. The product and technical information available through D & A Instruments can support decisions about optical sensing, suspended-solids monitoring, and deployment in marine or freshwater environments.

During the survey, log every operational change at the time it occurs. Mark pump start, rate adjustment, discharge interruption, profiler pause, reversal of direction, sensor cleaning, and sample collection. A short note such as “pump cycled at 14:32” can explain a feature that would otherwise be misread months later. If a sensor saturates, loses communication, or exceeds its expected range, preserve the raw record and document the event rather than deleting the affected interval.

Use the following checks before accepting a drawdown-related feature as evidence of a discrete groundwater zone:

• Compare the feature across downcast, upcast, repeat, and recovery profiles.
• Align sensor values with pump status, flow rate, water level, elapsed time, and depth.
• Check whether conductivity, temperature, pressure, and optical channels respond coherently.
• Review calibration checks, grab samples, sensor diagnostics, bubbles, fouling, and cable movement.
• Assign a quality flag and retain the original raw data when the cause remains uncertain.

The final dataset should distinguish measured values from interpreted values. A profile may be technically valid for describing conditions during pumping while being unsuitable for estimating static aquifer chemistry. Quality flags such as “pumped,” “transient,” “possible bubble interference,” or “depth uncertain” are more useful than a single pass-or-fail label.

Report drawdown effects with appropriate confidence

A clear report should state whether the bore was static, purged, pumped, recovering, or affected by nearby abstraction at the time of profiling. Include the pump intake and screen intervals, pumping rate, water-level change, recovery duration, profiler speed, and any deviation from the standard method. These details allow a later analyst to distinguish a hydraulic event from an equipment problem.

Interpretation should describe the evidence supporting each feature. For example, a conductivity transition repeated after recovery and supported by temperature data may be reported as a probable inflow boundary. A turbidity spike that occurred during a pump-rate change, appeared on only one pass, and lacked a corresponding pressure or chemistry response should be reported as a potential pumping or optical artefact.

Australian projects may need to communicate results to state groundwater regulators, water utilities, consultants, landholders, and laboratory teams with different expectations. Plain language helps: identify what was measured, what was inferred, and what remains unresolved. A concise data-quality statement can prevent a transient pumped-bore profile from being used later as a static baseline in a conceptual site model.

The most defensible interpretation preserves the disturbance history instead of trying to remove it from the record. Pumping-induced drawdown can be valuable evidence of connectivity, inflow, storage, or bore performance when it is measured deliberately. When it is incidental, careful timing, independent water-level data, repeated profiles, optical checks, and transparent quality flags keep the artefact from becoming a false groundwater boundary.