Assessing Credibility of Observers and Data Quality in Maritime Volunteer Bathymetry Surveys
In the pursuit of ensuring the reliability of Volunteered Bathymetric Information (VBI) for authoritative hydrographic data, a robust approach has been proposed. This approach combines data quality assessment, validation against existing authoritative sources, and clear classification methods.
The assessment of VBI data quality involves evaluating key factors such as position accuracy, depth accuracy, collection date, and sea floor coverage. These factors help define zones or categories of confidence for bathymetric data, similar to the classification approaches used in official hydrographic surveys.
Validation and cross-checking VBI against authoritative survey data or satellite, remote sensing, or modeled data can highlight discrepancies or confirm reliability. This might involve overlaying VBI on existing charts or consulting in-situ measures for alignment.
Metadata requirements for VBI submissions should include detailed metadata on data collection methods, sensor/system characteristics, time/date, and environmental conditions to facilitate assessment of accuracy and fitness for purpose.
Incorporation of uncertainty estimates is crucial. VBI should be accompanied by uncertainty or confidence estimates derived from known sensor accuracies and sampling methodologies. This supports more informed use in navigation and resource management.
Establishing automated and manual quality control steps, including filtering of outliers, consistency checks, and expert review, helps prevent erroneous VBI from impacting authoritative datasets.
Adopting community standards and frameworks, such as the IHO (International Hydrographic Organization) standards or the Zone of Confidence (ZOC) framework, provides a systematic way to integrate VBI within established hydrographic data quality paradigms.
Stakeholder involvement and feedback mechanisms are essential. Engaging hydrographic offices, maritime users, and scientists in evaluating VBI contributions and setting thresholds for acceptance enhances credibility and usability.
The methods outlined apply equally to data made by observers and to all observers, not just volunteers. NOAA supports the IHO Crowd-sourced Bathymetry Working Group and hosts the IHO Data Center for Digital Bathymetry (DCDB).
Changepoints in bias were identified using a first run of the bias parameter algorithm, and a second run computed a final model for each group of transits. Vertical correctors for water level can generally be evaluated from a nearby tide gauge or harmonic constituents, and in some instances, sound speed correctors can be derived from forecast models or oceanographic atlases.
Data were selected from the DCDB, using only data from their "Crowd-sourced" database, which consists almost exclusively of single-beam echosounder observations from recreational mariners submitted using Rose Point Coastal Explorer through a NOAA-led pilot project.
The work has demonstrated that it is possible to estimate biases for VBI observers, and to assess reputation, given a sufficient amount of data in an area with a reference depth dataset. Few, if any, VBI observers have corrections applied to their data or, if they do, document what was done. Comparisons against authoritative data are therefore likely to always fail without bias assessment.
Tools to assist in the assessment of VBI data are required, and a reputation metric would allow the HO to determine how to treat the observation for authoritative purposes. Each observer is initialized with a reputation of (1500, 350), reflecting an uninformative prior. The reputation is then updated as data are contributed, breaking the sequence of observations into 60-second batches.
Data were subsequently filtered to a time range of 2016-01-01 to 2019-08-10, and a depth limit of less than 11 km in order to remove clearly erroneous data. All observers modelled demonstrate a positive offset value and negative slope, indicating shallower observed depths compared to reference depths.
The paper focuses on evaluating dynamic observer reputation from archive observations, using the IHO DCDB dataset for the area of Puget Sound, WA around Seattle. In smaller, or less frequented regions, sound speed corrections may be more difficult to obtain, and the vertical correction from the echosounder to the waterline is almost always missing.
The Canadian Hydrographic Service has used DCDB data to address dangers to navigation in the Inside Passage from Seattle, WA to Juneau, AK. The proposed solution for assessing the credibility of VBI is an adaptation of a pairwise ranking system originating in chess to the problem of assessing observer reputation for VBI. This system makes demands of the Hydrographic Office, particularly that there is a willingness to set thresholds for reputation and evidence required to trigger changes to the chart.
A transitive trust relationship allows the reputation of the observer to be assessed from the observations, but observers may change over time, initially providing poor data and improving, or forgetting to accommodate a modified configuration of the boat, and suddenly (or gradually) generating poorer data. Reputation is not a static concept, and a principled method for database age management is required.
The suggested system could also be applied to authoritative databases, allowing for principled management of hydrographic databases. Most hydrographic offices agree that Crowdsourced Bathymetry or Volunteered Geographic Information can have potential for authoritative charting. However, the majority of publicly available VBI does not have refinements like pre-capture quality assurance, installation parameters corrections, or corrections for sound speed and other factors. Consequently, an estimate of bias must be generated from the data.
In summary, the management of VBI for authoritative use relies on structured assessment criteria applied to positional, temporal, and depth accuracy, careful validation exercises, and classification within established confidence zones to ensure that voluntary data augments rather than undermines hydrographic products. This combined approach aligns with current hydrographic practices for data quality and integrates VBI into operational frameworks.
- The application of technology in ocean mapping, such as data-and-cloud-computing, plays a crucial role in the quality assessment and validation of Volunteered Bathymetric Information (VBI), which is essential for authoritative hydrographic data.
- In the field of environmental-science and education-and-self-development, the integration of VBI data can contribute significantly to general-news articles and sports reporting, providing valuable insights into marine environments and the impact of human activities on them.
- To facilitate the use of VBI data in navigation and resource management, it's important to employ clear classification methods, incorporating uncertainty estimates and metadata requirements, as well as establishing automated and manual quality control steps.
- Collaboration among stakeholders, including hydrographic offices, maritime users, and scientists, is key to the credibility and usability of VBI, as they can provide valuable feedback and help set acceptance thresholds for VBI contributions.
