George, C.T.; Gerard, F.F.; Peyton, J.M.; Pescott, O.L.

Two-metre resolution land cover classification of the Akrotiri Peninsular, Cyprus from WorldView imagery 2018

https://doi.org/10.5285/95ab2233-ed8d-4c32-b82c-ae282ba567f7 Cite this dataset

This dataset contains land cover classification of the Akrotiri Peninsular of Cyprus, using WorldView Imagery at 2m resolution. Images from March 2018 and July 2018 were used. All the imagery was atmospherically corrected. The thematic detail of the map was generalised to the following main vegetation/cover types: bare, lake, temporary water, salt marsh, rush salt meadow, garrigue, sparse vegetation, grass, mixed woodland, Eucalyptus and Acacia. The cover classes salt marsh and rush salt meadow represent mosaics of vegetation communities. Due to Covid-19 restrictions, ground data collection was severely restricted, resulting in limited training and validation data.

This work was supported by the Natural Environment Research Council award number NE/R016429/1 as part of the UK-SCAPE programme delivering National Capability.

Publication date: 2021-09-29

Format

TIFF

Spatial information

Study area
Spatial representation type: Vector
Spatial reference system: WGS 84

Temporal information

Temporal extent: 2018-03-03 to 2018-07-06

Provenance & quality

Data was generated using the ENVI Example Based Feature Extraction tool, from the following satellite images: WorldView-3 (WV-3) 1.2m multispectral (8 wavelength bands) from 3rd March 2018 and WorldView-2 (WV-2) 2.0m multispectral (8 wavelength bands) from 6th July 2018. The July imagery was a single image covering the whole Akrotiri peninsular, while the March imagery consisted of two images (east and west of peninsular) originating from different swaths, each with a different viewing angle. All the imagery was atmospherically corrected using the ENVI FLAASH software. For both east and west, the July and March imagery were resampled to ensure consistency in spatial resolution and combined into a single 2m 16 wavelength band image stack. The classification was validated using google maps and local knowledge due to COVID-19 restricting ground data collection.