19: Subantarctic and Antarctic Islands#

The region encompasses all glaciers on the islands in the periphery of mainland Antarctica, except those covered by ice rises within ice shelves. Some remote islands in the Southern ocean are also included. The region does not include glaciers on mainland Antarctica.

region map

Fig. 81 Regional glacier area. Download high resolution version.#

Changes from version 6.0 to 7.0#

Region 19 (previously “Antarctic and Subantarctic”) was split into two first-order regions. Region 19 in RGI 7.0 now solely includes the islands in the periphery of Antarctica, and was renamed to “Subantarctic and Antarctic Islands”. A new region 20, (“Antarctic Mainland”) was added to encompass the remaining subregion (“Antarctic Ice Sheet”, previously 19-31 and now 20-01).

Most RGI 6.0 outlines originated from Bliss et al. [2013] based on topographical maps from the Antarctic Digital Database [Scientific Committee on Antarctic Research, 2000] (see Bliss et al. [2013] for details). For RGI 7.0, glaciers in the following subregions were remapped using satellite imagery from around year 2000, and georeferencing issues were corrected (shifts, deformation, etc. See github discussion here and here):

  • 19-01: Subantarctic (Pacific)

  • 19-02: South Shetlands and South Orkney

  • 19-03: Subantarctic (Atlantic)

  • 19-04: Subantarctic (Indian)

  • 19-05: Balleny Islands

  • 19-11: E Queen Maud Land 7A

  • 19-12: Amery Ice Shelf 7B

  • 19-13: Wilkes Land 7C

  • 19-14: Victoria Land 7D

  • 19-15: Ross Ice Shelf 7E

  • 19-16: Marie Byrd Land 7F

  • 19-17: Pine Island Bay

  • 19-24: W Queen Maud Land 7K

Several wrongly mapped icebergs and other bodies were removed. Some larger ice caps were manually divided into individual glaciers, especially in subregions 19-15 and 19-16. Altogether, glacier area changes between RGI 6.0 and RGI 7.0 are small (+0.1%) but the quality of the inventory has considerably improved.

An important discussion about RGI outlines in region 19 has been triggered by two recent publications: Millan et al. [2022] and Hock et al. [2023], raising the question whether certain outlines in RGI belong to the ice sheet or the glacier category. We acknowledge the need for discussion and coordination within the scientific community to avoid double counting, but chose not to change the current separation in the RGI without consensus from both the glacier and ice-sheet research communities.

Kerguelen Islands (second-order region 19-04)

Glacier outlines in RGI 6.0 were from the 1960s and partly very roughly digitized. All outlines were replaced in RGI 7.0 based on a much improved and more recent (year 2001) data set in the GLIMS database (submissions 527 and 528). Since this data set does not include the south-western and northern part of the island, these regions were remapped. As suitable satellite images are not available for this region close to the target year 2000, images from 2010 (Landsat 7) and 2018 (Landsat 8) in the north, and images from 2018 (Landsat 8) and 2022 (Sentinel-2) in the southwest were used to map the glaciers. Furthermore, some missing smaller glaciers in the south-east of the main ice cap were added and some of the existing outlines were corrected. The “World imagery” layer of the ESRI Basemap was locally used to aid in the interpretation. The AW3D30 DEM was used to derive a flow direction grid and correct ice divides.

Attributes

RGI 6.0 included for each glacier a label for the terminus type attribute (term_type). However, in RGI 7.0 all glaciers received the label Not assigned, despite the prevalence of marine and shelf-terminating glaciers in this region. Updating this attribute is planned for the next version of the RGI.

Additional information#