Weddell seal reporting grid of the Vestfold Hills, Antarctica

Brief description

This is a digital version of the grid reference map used to plot all sightings of Weddell seals in the Vestfold Hills. The point of origin is the same as the original map and each grid cell is numbered with the same numbering scheme. This can be used to plot any data using the same numbering scheme by joining (ArcInfo) or linking (ArcView) records to this coverage's polygon attribute table (pat) through the item GRIDREF.

The original map was a 1:100 000 map of the Vestfolds, provided by Harry Burton, with a grid drawn over it. The grid references were given as either six or four figure values on which field scientists are to plot their data.

This map has the following Antarctic Division drawing reference number:

M/75/05A

Some research with John Cox revealed that this grid was drawn up over a map digitised from another map with the following specifications:

Scale 1: 100 000
Date: 1958 (reprinted 1972)
Projection: Polyconic
Published by: Division of National Mapping, Canberra
Reference number: NMP/58/084

Data are referenced to a 'grid' of 1 minute spacing in x axis and 30 second spacing in y axis. The point of origin is apparently 68 20 S 77 48 E. There are 45 rows and 47 columns.

The 'grid reference' is in fact in geographic coordinates (but using arbitrary units) so the projection of the original map became irrelevant.

The procedure adopted to create a new digital grid was as follows:

(Carried out in Arc/Info)

1. Generate a coverage using the original 'grid references'.
2. Tics were also generated using the corners of the 'grid reference' system.
3. A new coverage was created with tics at the same locations but given the true latitude/longitude vales.
4. The original coverage was then transformed to the new coverage based on the new tic values.
5. The new coverage was then projected from geographic coordinates to UTM metres.

The data locations were then viewed in Arc/Info using a coverage of the coastline supplied by the Mapping Officer, Antarctic Division. This had previously been determined to be in the UTM projection.

An offset was clearly visible between the data locations and the coastline. In order to determine whether the offset was more or less uniform, ten locations were plotted from the original data onto the original map using the 'grid'. Finally a manual corrected was made by moving all the data locations by a uniform distance of 508 metres north and 68 metres west.

Information from John van den Hoff, February 2019:
The grid cells were originally labelled from 1 to 47 along the x axis and 1 to 45 along the y axis. The four digit values in the GRIDREF field of the attribute table are the x value followed by the y value.
To avoid confusion between x and y values, the grid was later revised so that the y values were prefixed with a ‘1’ so for example 01 became 101. The GRIDREF_X and GRIDREF_Y fields have the x and y values of the revised grid.
This needs to be kept in mind when data is sourced from field books.
The map shows the revised grid.

Lineage

Progress Code: completed

Statement: Positional Accuracy
1 m to 10 m

There are several sources of error:

1. Inaccuracies in the original map, both in the shape of the coastline and in the geodetic framework. In 1958, the geodetic control for the aerial photography would have been done using astronomical observations.
2. Inaccuracies in the 'modern' (UTM) coastline. There are known to be inaccuracies in the map from which this was digitised and there are question marks about the quality of the digitising.
3. Inaccuracies in plotting the location data in the field which were presumably done by estimating position from local landmarks rather than by measurement eg compass bearing.
4. The correction of the offset manually is only a 'best fit' for all points. Values provided in temporal coverage are approximate only.