Cape Denison and McKellar Islands GIS dataset from Ikonos satellite imagery

Full description

A GIS dataset of around Cape Denison and part of George V land created from two IKONOS satellite images.
Layers created from digitising directly from the imagery include: mapping extent, continent, building, refuge, coastline, reef, offshore rocks, sea, snow, sheet, island, birds, rock, moraine, sea ice, lakes
- The mapping extent layer represents the edge of the IKONOS imagery.
- The continent layer represents the land mass shown in IKONOS imagery. It was generated using the digitised coastline and bounded by lines that represent the edge of the image.
- The snow spatial data represents the snow cover in January 2001
- The sheet ice spatial data represents the ice extent in January 2001
- The penguin spatial data represents the penguin colony extents, based on guano deposits.
- The rock spatial data represents the exposed bare rock

Lineage

Progress Code: completed

Statement: The data were digitised from two IKONOS 1 metre Pan sharpened Multispectral images. The image reference numbers were 2001013123204820000011627275 and 2001012623380410000011623143. The images were supplied as georeferenced band interleaved by line (BIL) multiband image files. Features were digitised in ArcMap version 8.1 as shapefiles and converted to ArcInfo coverages using ArcInfo version 8.1. Coverage clean up and editing was performed using ArcEdit and ArcInfo version 8.1 commands. Attribute assignment and management performed using ArcMap and ArcView 3.2. - The frame line represents the edge of the IKONOS image and was digitised directly from the image. - The Continent polygon was formed using the coastline digitised from the images and bounded by lines representing the edge of the images. - Buildings were located using spatial information provided by Ursula Ryan and visual identification by Henk Brolsma. - Refuges were visually identified by Henk Brolsma. - Coastline was tagged as being either rock or ice as identified from the imagery. - The reefs were difficult to confidently identify from the IKONOS imagery and in many cases it was difficult to distinguish between wave action and actual reef. An area bounded by 67 degrees 01 minutes South, 66 degrees 57 minutes South, 142 degrees 36 minutes East and 142 degrees 42 minutes East was scanned with care at 1:2,000 scale and anything that appeared to be a reef at that scale was captured as approximate reefs. There are an infinite number of small white objects that appear at larger scales. It was not possible to distinguish if these were rocks/reefs or waves. There was no value in capturing them all as this would indicate that the whole area was covered by regularly spaced reefs. Definite reefs were assigned a class value of DEFINITE and others were assigned a class value of APPROXIMATE. There is no guarantee that all the reefs in this area have been located or that the digitised approximate reef features are in fact reefs. - The sea polygon was formed using the coastline digitised from the images and bounded by lines representing the edge of the images. Island reef and offshore rock features were removed from this sea polygon to create holes. - Snow and ice polygons were digitised using image analysis techniques in Erdas Imagine. The highly detailed snow/ice shapefiles created by Erdas were converted to Arc/Info coverages and cleaned to remove the excess noise created by the image processing techniques. Fuzzy tolerances of up to 3 metres and dangle tolerances of up to 20 metres were used. Inland snow areas and islands of less than 1000 square metres were also removed to reduce the vast number of isolated drifts. The hand digitised coastline was integrated with the snow polygons to ensure consistency between layers. The edges of some large polygon areas were hand digitised to reduce the complexity of the automatically generated linework. The digitised data and imagery were georeferenced to UTM Zone 54 (WGS84) and converted to unprojected coordinates on WGS84 using the Arc/Info project command. - Penguin colony polygons were digitised from the images. Data were checked and corrected by Eric Woehler, Ornithologist, AAD. Additional colonies on the continent and on the Mackellar Islands were supplemented from aerial photography taken in 2001/02 by MacIntyre expedition. In April 2010 the coastline near Lands End was edited, a reinterpretation of the satellite image prompted by feedback from Chris Henderson, an expeditioner to Cape Denison with the Mawson's Huts Foundation in 2008/09 and 2009/10. This editing resulted in changed coastline, continent, island, offshore rock and rock data. Usually mapped to a Universal Transverse Mercator projection, Zone 54.