Abstract:
This record consists of 5 different datasets covering the Alexander Island region of Antarctica: coastline (including grounding line and ice shelf fronts), contours, spot heights, rock outcrop and lakes. The data have been produced for a new topographic map of Alexander Island at 1:500,000 scale, produced by the Mapping and Geographic Information Centre, British Antarctic Survey.
The data are suitable for a 1:500,000 scale map but may be suitable for larger scales in certain areas. They have been created from source data ranging from 2022 - 2025. The data primarily cover Alexander Island, and also cover Rothschild, Charcot and Latady islands, as well as Wilkins and George VI ice shelves, and the Rymill Coast section of Palmer Land on the Antarctic Peninsula.
The datasets were created using a mixture of GIS software, primarily digitised from Sentinel-2 satellite imagery or extracted from high resolution, published elevation models. Exact details of each dataset can be found in the lineage statements.
Keywords:
Alexander Island, cartography, coastline, contours, elevation, grounding line, lakes, rock outcrop, spot heights, topographic
Gerrish, L., Skachkova, A., Ireland, L., & Field, E. (2025). Topographic data for Alexander Island and the surrounding region, including coastline, grounding line, contours, spot heights, rock outcrop and lakes (2022 - 2025) (Version 1.0) [Data set]. NERC EDS UK Polar Data Centre. https://doi.org/10.5285/98a92a0c-4fa3-4319-a554-f85d1564de40
| Access Constraints: | None |
|---|---|
| Use Constraints: | Data supplied under Open Government Licence v3.0 http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/. |
| Creation Date: | 2025-04-24 |
|---|---|
| Dataset Progress: | Planned |
| Dataset Language: | English |
| ISO Topic Categories: |
|
| Parameters: |
|
| Personnel: | |
| Name | UK Polar Data Centre |
| Role(s) | Metadata Author |
| Organisation | British Antarctic Survey |
| Name | Ms Laura Gerrish |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Aliaksandra C G Skachkova |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Ms Louise Ireland |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Name | Ms Elena Field |
| Role(s) | Investigator |
| Organisation | British Antarctic Survey |
| Parent Dataset: | N/A |
| Reference: | British Antarctic Survey, 2025. Alexander Island, 1:500 000 scale map. BAS Topographic Series, sheet 8, edition 1. | |
|---|---|---|
| Quality: | All data is subject to error and known potential errors have been detailed in the lineage statements. Elevation data accuracy statements can be seen in the source data publications. The coastline, rock and lakes have been interpreted from satellite imagery using the dataset creator's best knowledge of the region and available data. | |
| Lineage: | Coastline: The starting point for this data was version 7.8 of the Antarctic Digital Database coastline (doi.org/10.5285/98d7c98c-b223-4774-b20e-886630632b12). Using ArcGIS Pro version 3.3.0, the data was clipped to this area of interest and every section of line was checked against current satellite imagery, and updated where necessary. The grounding line has undergone a major revision, with the majority of new data being incorporated from Wallis et al, 2024, https://doi.org/10.5194/tc-18-4723-2024 . There ar 5 fields in the attribute table of the final dataset: * 'surface' - detailing whether the line is 'ice coastline', ''rock coastline', 'ice shelf and front' or 'grounding line' * 'revdate' - revision date * 'source' - the source of each section of line, mostly consisting of satellite name and ID * 'sourcedate' - the date of the source, mostly in YYYY-MM-DD format apart from 2 sources where a year is only available * 'lineage' - a brief description of how each section of line was created The coordinate reference system of the data is WGS84 Antarctic Polar Stereographic, EPSG 3031. Contours: Using ArcGIS Pro version 3.3.0, contours were extracted from two different Digital Elevation Models and then merged together into 1 file. The majority of contours were created from REMA v2 (Howat, Ian, et al., 2022, '''The Reference Elevation Model of Antarctica - Mosaics, Version 2'', https://doi.org/10.7910/DVN/EBW8UC, Harvard Dataverse, V1, accessed September 2024.) apart from Latady Island and Charcot Islands which were created from the Copernicus GLO-90 DEM (https://doi.org/10.5270/ESA-c5d3d65). For REMA, the 500m mosaic was downloaded and converted to EGM2008 geoid, and contours were extracted. For the Copernicus GLO-90 DEM, the 90m tiles were used and contours were extracted from these, and then smoothed using a 60m smoothing tolerance. The two areas were merged together, and there is a field in the attribute table called 'source' which details which source each line comes from. The main field is called 'Contour' and contains the height in metres of each contour line. There are also fields for 'surface' which states whether the contour falls on ice or ice-free ground, and a field called 'interval' which states 'yes' or 'no' depending on whether the contour is an interval of 1000 or not. These fields are to improve cartographic presentation of the data, when styled using the fields. The coordinate reference system of the data is WGS84 Antarctic Polar Stereographic, EPSG 3031. Spot heights: This file contains 905 spot heights for the Alexander Island region, each point identified as a local high point with the height extracted from the 2 metre mosaic tiles of REMA v2 (Howat, Ian, et al., 2022, ''The Reference Elevation Model of Antarctica - Mosaics, Version 2'', https://doi.org/10.7910/DVN/EBW8UC, Harvard Dataverse, V1, accessed September 2024.) Detailed methodology is given below. All processing was carried out in ArcGIS Pro version 3.3.0, apart from step 2 which was carried out in QGIS version 3.34.7. Only a selection of these heights are shown on the published Alexander Island map, but all of the points that were produced are published here. 1. 32 metre version tiles of REMA v2 were downloaded, merged together, and converted to use Mean Sea Level / EGM2008 heights. The 32 m version was used to start with because the 2 m version was too power and time intensive for processing. 2. Using the 'Spot Height Extractor' Plugin in QGIS (version 1.1 - https://plugins.qgis.org/plugins/Spot-Height-Extractor-Plugin-main/ ), spot heights were extracted from the 32 m mosaic. This plugin identifies local high points. A neighbourhood size of 101 was used, which gave an optimal number of output points. 3. Manual tidying of the output file was performed to delete points on ice shelves, and in other locations where a height did not look necessary (e.g. on fairly flat glaciers). This gave the highest points in the 32 m DEM, but we wanted the highest points in the more detailed 2 m version of the DEM. 4. 2 m version tiles of REMA v2 were downloaded and converted to use Mean Sea Level / EGM2008 heights. 5. A polygon buffer of the spot height points was created with a 40 m radius. 6. Then, using the ''Zonal Statistics, 'Raster Calculator' and 'Raster to Point' tools, the buffers were run over the 2 m tiles of REMA v2, to extract the highest points within each 40 m buffer, as detailed in this article: https://support.esri.com/en-us/knowledge-base/how-to-create-points-representing-the-highest-and-lowes-000030199. This process gave the true highest point according to the 2 m DEM, whereas the 32 m DEM just gave an indication of where the highest points were. The heights were also extracted from the original 'height above ellipsoid' versions of the REMA 2 m tiles, so that we had mean sea level heights as well as height above ellipsoid. 7. Manual tidying of the output file was performed to delete duplicate points in cases where 2 or 3 pixels has been identified as having the highest value, and in overlapping edges of the tiles. 8. Accuracy statements for REMA v2 are given in the associated paper and two separate files can be accessed to indicate the accuracy of a local area: 'count' details the number of contributing DEMs, and 'mad' details the median absolute deviation of contributing DEMs. Details here: https://www.pgc.umn.edu/guides/stereo-derived-elevation-models/pgc-dem-products-arcticdem-rema-and-earthdem/ 9. These figures are useful for assessing the usefulness of individual heights that we had created, so the necessary tifs were downloaded, and the values for each height was extracted. 10. Finally, a field for place name was added to the resultant shapefile, where the spot height appeared to be associated with a known place name in the British Antarctic Gazetteer, accessed November 2024 (https://apc.antarctica.ac.uk/ ). Where a height was not clearly identified with a place name, the field was left empty. Final shapefile attributes: * 'msl_egm08' - the height of each point, extracted from the EGM2008 versions of the 2 m DEMs * 'hae_wgs84' - the height of each point, extracted from the original height above ellipsoid versions of 2 m DEMs * 'QA_count' - the 'number of contributing DEMs' to the REMA v2 tiles * 'QA_mad' - the 'median absolute deviation of contributing DEMs' to the REMA v2 tiles * 'Name' - the associated place name for each point, if a name is present The coordinate reference system of the data is WGS84 Antarctic Polar Stereographic, EPSG 3031. Rock outcrop: This file details the extent of rock outcrop in the Alexander Island region. The data largely consists of data automatically extracted from Sentinel-2 imagery, with further data automatically extracted from Landsat 8 imagery and manually added from Sentinel-2 imagery. The majority of this dataset was created using a mosaic of Sentinel-2 imagery. This mosaic was generated with Google Earth Engine from cloud-free pixels of Sentinel-2 images, acquired during the 2023-2024 Antarctic summer season. The ResNet18 model, pre-trained with Sentinel-2 MOCO weights and fine-tuned with manually created labels, was used for processing (see Wang et al. 2022, https://doi.org/10.48550/arXiv.2211.07044 for details regarding the pre-trained model). The model training was conducted using Python TorchGeo library. The mosaic has a spatial resolution of 50 metres. The output rock outcrop data was then generalised using ArcGIS tools: polygons smaller than 10 pixels were deleted, polygon parts less than 2% of a polygon area were eliminated, and the data was smoothed with the threshold of 40 metres. Due to the 50 m pixel size of the input satellite imagery, and the nature of this automatic extraction methodology, small areas of rock outcrop were identified as missing from the data. The most important of these locations were identified by comparing against named nunataks and other sources of rock outcrop data. A published rock outcrop dataset was used to fill in some of these gaps (Burton-Johnson, A. et al, 2016, doi.org/10.5194/tc-10-1665-2016 ). Where data was missing from both sources, data was manually added by looking at full-resolution (10 m) Sentinel-2 imagery from autumn 2024. The final shapefile contains a field called 'source' which details which of the 3 sources each section of rock outcrop comes from. The coordinate reference system of the data is WGS84 Antarctic Polar Stereographic, EPSG 3031. Lakes: The lakes in this dataset have been digitised from Sentinel-2 satellite imagery in ArcGIS Pro version 3.3.0. The file contains 8 lakes: 4 named lakes and 4 unnamed. All of the lakes are covered by ice and difficult to distinguish accurately from satellite imagery. They have been digitised to the best of the dataset creator's ability, but the extents may not be completely exact. The 4 unnamed lakes were visually interpreted by looking at the imagery and identified as of the same appearance as other known/named lakes, but their likely existence might be proven wrong in future. Furthermore, there might be further lakes that were not identified in this dataset, especially smaller lakes. The shapefile contains two fields: 'source' contains the ID of the Sentinel-2 image that the lake has been digitised from, and 'Name' which contains the name of the lakes, from the British Antarctic Territory, accessed January 2025. The coordinate reference system of the data is WGS84 Antarctic Polar Stereographic, EPSG 3031. |
|
| Temporal Coverage: | |
|---|---|
| Start Date | 2022-01-01 |
| End Date | 2025-04-20 |
| Spatial Coverage: | |
| Latitude | |
| Southernmost | -72.61944 |
| Northernmost | -68.6888 |
| Longitude | |
| Westernmost | -77.47722 |
| Easternmost | -67.4997 |
| Altitude | |
| Min Altitude | 200 |
| Max Altitude | 2800 |
| Depth | |
| Min Depth | N/A |
| Max Depth | N/A |
| Location: | |
| Location | Antarctica |
| Detailed Location | Alexander Island |
| Location | Antarctica |
| Detailed Location | Wilkins Ice Shelf |
| Location | Antarctica |
| Detailed Location | George VI Ice Shelf |
| Location | Antarctica |
| Detailed Location | Fossil Bluff |
| Location | Antarctica |
| Detailed Location | Latady Island |
| Location | Antarctica |
| Detailed Location | Rothschild Island |
| Location | Antarctica |
| Detailed Location | Charcot Island |
| Location | Antarctica |
| Detailed Location | Rymill Coast |
| Location | Antarctica |
| Detailed Location | Bach Ice Shelf |
| Data Collection: | ArcGIS Pro (version 3.3.0) QGIS (version 3.34.7) QGIS Spot Height Extractor plugin version 1.1 |
|---|
| Distribution: | |
|---|---|
| Distribution Media | Online Internet (HTTP) |
| Distribution Size | 9.7 MB |
| Distribution Format | SHP |
| Fees | N/A |
| Data Storage: | This dataset consists of 5 zipped shapefiles: - Zipped shapefile (coastline) 537 kb - Zipped shapefile (contours) 2281 kb - Zipped shapefile (spot heights) 41 kb - Zipped shapefile (rock outcrop) 6969 kb - Zipped shapefile (lakes) 11 kb |