ad3b585e-2ec2-43e8-bbf9-c0fd545019bf
English
dataset
Geographic Information Group
NatureScot
Data Supply
01463 725076
Great Glen House
Inverness
Highland
IV3 8NW
data_supply@nature.scot
2022-03-03T01:02:12
UK GEMINI
2.3
27700
EPSG
8.9.5(3.0.1)
Scotland Habitat and Land cover map (Change 2019 - 2020)
2021-06-04
HLCM19-20CHNG
Habitat and land cover maps created using AI to classify satellite data to EUNIS level 2 by Space Intelligence in partnership with NatureScot. This work was a response to the Can Do Innovation fund challenge AI for Good -How can we use Artificial Intelligence (AI) techniques to tackle the climate emergency? This dataset contains the changes in land cover in Scotland between the years 2019 and 2020. It is part of a series of 3 layers (raster datasets at ~20m resolution). The other layer provides the land cover classification for the year 2019 and a third layer provides the land cover for the year 2020.
Habitat and land cover maps created using AI to classify satellite data to EUNIS level 2. These data compare habitat and land cover datasets from 2019 and 2020 to show areas of change.
Maps and data created by Space Intelligence with input and support from NatureScot, © SNH
Philippa Vigano
NatureScot
Programme Manager: Innovative Technologies (IniT) Programme
Philippa.Vigano@nature.scot
Marina Gray
NatureScot
Geographic Information Officer
Marina.Gray@nature.scot
Ed Mitchard
Space Intelligence
07850 861130
ed@space-intelligence.com
GB-SCT
ISO 3166-2:2020
Codes for the representation of names of countries and their subdivisions — Part 2: Country subdivision code
2020-08-31
4
2020-08-31
Habitats and biotopes
GEMET - INSPIRE themes
2008-06-01
1
2008-06-01
habitat
GEMET - Concepts
2021-06-01
4.2.1
2021-06-01
Downloadable Data
Available under the Open Government Licence http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/
Maps and data created by Space Intelligence with input and support from NatureScot , © SNH
English
environment
Microsoft Windows 7 Version 6.1 (Build 7601) Service Pack 1; Esri ArcGIS 10.7.1.11595
true
-9.229737
-0.70499
54.513263
60.866098
GeoTIFF
MB
105
https://cagmap.snh.gov.uk/arcgis/services/snh_habitats_and_species/MapServer/WMSServer?request=GetCapabilities&service=WMS
OGC:WMS-1.3.0-http-get-capabilities
38
Scotland Habitat and Land cover map (Change 2019-2020)
https://cagmap.snh.gov.uk/natural-spaces/dataset.jsp?dsid=HLCM19-20CHNG
WWW:LINK-1.0-http--link
NatureScot - Natural Spaces download page
https://cagmap.snh.gov.uk/natural-spaces/download/HLCM19-20CHNG/tif
WWW:LINK-1.0-http--link
TIFF raster
Commission Regulation (EU) No 1089/2010 of 23 November 2010 implementing Directive 2007/2/EC of the European Parliament and of the Council as regards interoperability of spatial data sets and services
2010-12-08
Data Set Not Assessed
false
User’s Accuracy
2021-06-04
Philippa Vigano
NatureScot
Programme Manager: Innovative Technologies (IniT) Programme
Philippa.Vigano@nature.scot
The overall accuracy of the maps against our input data was high (95.8%). The User’s Accuracy, a measure of how likely a pixel of a particular class in the map is to actually be that class, ranged from 86.4% for class H3 (inland cliffs), to 99.9% (screes), with most classes well over 90% accuracy.
Commonly confused classes using remote sensing data, such as the four woodland classes (Deciduous, Coniferous, ‘Mixed’ and Small’ (patches and lines of disturbed woodland) were all classified with ~95% accuracy or higher. This suggests the classifier would be able, if repeated annually, to reliably detect changes between these classes, which is critical for the determination of the NCAI.
true
20 m resolution geotifs derived from
Ground Data
The ground data were collected through using a combination of the following sources, using a broad search that stretched beyond our Areas of Interest:
● Habitat Map of Scotland (ground polygons)1
● 2018 National Forest Inventory2
● Ordnance Survey3
● Global Forest Change v1.64
● High resolution imagery5
In all cases the ground data were not used naively: we used a careful combination of at least two data sources to create each polygon, and checking against recent high resolution imagery to ensure each polygon was ‘pure’ (i.e. included only one class) and up to date (for
example, if it was a forest polygon, the trees had not been cleared since the data were collected).
Satellite remote sensing datasets used for mapping
Optical Sentinel 2 (S2) (30/03/2019-10/11/2019)
Radar Sentinel-1, descending and ascending (01/01/2019-31/12/2019)
ALOS-PALSAR 2, 2018 annual composite
Topography Shuttle Radar Topography Mission (SRTM, 2000)
Process
Extensive training datasets, and derived features from remote sensing data, to implement a complex set of tuned machine learning algorithms to produce a Prediction Model, and ultimately a prediction of a class for each pixel. Through the project duration the sophistication of the models used increased, increasing accuracy and efficiency. For commercial reasons the details of the final algorithms used will not be revealed here.