A geological oceanography concept defining processes, measurements, or products used to describe the seafloor and its history. It applies when operational prerequisites and quality controls are satisfied and produces interpretable observations or derived maps. It does not yield reliable conclusions without validation, uncertainty handling, and integration of multiple evidence types. It materially affects knowledge of seafloor evolution, sediment transport, and marine hazards by improving the accuracy of interpretations. The concept is generally stable, though observational coverage and analytical methods improve over time.

Principle

Preserve true vessel motion while removing or annotating artifacts by combining statistical outlier detection, physical-motion filters, time-synchronization checks, sensor-fusion validation and provenance tracking, ensuring reproducibility and quantified uncertainty.

Demonstration

Illustrative scenario: a survey log contains intermittent GNSS outages during rough weather, INS drift during maneuvers, and occasional false RTK fixes. The cleaning pipeline reconstrues a continuous navigation track by (1) aligning GNSS and recorder clocks, (2) flagging and removing sub-second spikes and mono-epoch jumps, (3) applying constrained interpolation over short gaps, (4) fusing INS dead-reckoning for intervals without GNSS, and (5) updating covariance estimates for each epoch. The output includes flags for imputed segments and metadata describing the applied corrections.

Misapplication

Over-aggressive smoothing or long-gap interpolation that removes legitimate high-frequency vessel motion (turns, rapid heave changes) or replacing missing tracks with unrealistic straight-line assumptions, leading to biased source–receiver offsets and mispositioned features in the seismic image.

Consequence

Proper navigation cleaning reduces positioning noise, prevents propagation of erroneous fixes into shotpoint assignment and binning, improves alignment between navigation and sensor data, and reduces imaging artifacts caused by misplaced geometry, while providing documented data quality flags for later interpretation.

Reversal

Leaving navigation data uncleaned — accepting all raw fixes and timestamps without QC or synchronization — yields noisy, discontinuous time series with spikes and offsets that propagate into poor georeferencing, inconsistent offsets, and degraded imaging and interpretation.

Boundary

Applies to preprocessing of navigation time-series used in marine geophysical acquisition and includes GNSS, INS, DVL and auxiliary sensors; it does not encompass higher-level geodetic transformations that change coordinate reference frames for publication, nor does it cover interpretation-driven relocation of features based on seismic imaging outcomes.

Semantic Tension

There is a trade-off between aggressively eliminating apparent anomalies to obtain a smooth track and preserving genuine high-frequency motions that are geophysically relevant; deciding the appropriate aggressiveness depends on downstream use (e.g., binning vs. dynamic processing) and on documented uncertainty.

Synthesis

Navigation data cleaning is a disciplined pipeline of detection, correction, interpolation, fusion and flagging steps that converts heterogeneous raw sensor streams into a continuous, documented navigation product with quantified uncertainties, enabling reliable georeferencing for seismic acquisition and processing.