Case Study: Marine data comparison

This example is a comparison between RCAHMS data of maritime losses off Scotland and primary survey data derived from a British Geological Survey multibeam echosounder survey of Loch Eribol, Durness , undertaken in 2009. The RCAHMS data is derived from a UKHO data dump in 1992, which comprises wrecks and losses. The Canmore data includes the position of known wrecks (in part derived from UKHO data) and also the approximate position of casualties or losses i.e. potential losses of shipping identified from extensive desk-based research.  Added to this, as part of the on-going maritime enhancement record project, are the locations of wrecks and losses, derived from both primary and secondary sources, including Whittaker’s ‘Off Scotland’.

Figure 34: RCAHMS database locations of wrecks are derived from many datasets, both primary and secondary. Accuracy is therefore an issue, with some of the records only be located within a few kms, while others may be surveyed to an accuracy of 10m. RCAHMS uses colour coding to highlight the level of accuracy.

Figure 34: RCAHMS database locations of wrecks are derived from many datasets, both primary and secondary. Accuracy is therefore an issue, with some of the records only be located within a few kms, while others may be surveyed to an accuracy of 10m. RCAHMS uses colour coding to highlight the level of accuracy.

RCAHMS’ database comprises a total of 20 records of wrecks and losses in Loch Eribol, compared with 4 known on the Seazone (UKHO derived) data. As a data cleansing exercise and as a pilot to develop different methods of presentation of RCAHMS marine data a series of polygons were placed around the site points, with the aim of demonstrating the accuracy of the site point geographical location. Some of the polygons were grid squares, ranging from 100m2 to 1km2, whilst others were ‘tailored’ polygons. Note some polygons cover land and sea. This polygon approach is designed to address the issue of points being located on land as a result of low resolution national grid references and to provide clarity over the potential location, and, in the case of ‘real’ polygons, the size of the site.

Figure 35: RCAHMS database points with polygons overlaid, to demonstrate the potential location of the site, ©RCAHMS.

Figure 35: RCAHMS database points with polygons overlaid, to demonstrate the potential location of the site, ©RCAHMS.

The BGS 2010 survey was undertaken using the White Ribbon, this relatively shallow draft vessel could not however, reach some of the potential sites in very shallow waters. In addition the survey did not go north of Rispond harbour and hence a number of potential locations were not surveyed. As a result of the BGS multibeam survey two wrecks were identified and two possible sites recognised that reconcile with the RCAHMS locations. Obviously, this demonstrates that there is a disparity between the datasets. This disparity between primary derived survey data and using secondary sources to populate the RCAHMS Canmore database, is a challenge that RCAHMS can address with third party datasets such as the data from the BGS Loch Eribol survey.

Return to Data issues