Multi-year harmonized land cover samples based on LUCAS and CORINE datasets

Harmonized training samples based on LUCAS (Land Use and Coverage Area frame Survey) and CLC (CORINE Land Cover) Maps, according to the following classes:

• 111 - Urban fabric: the aggregated continuous and discontinuous urban fabric class is classified when urban structures are dominating the surface area. The impermeable features like buildings and artificially surfaced areas range from 30 to 100% land coverage.
• 122 - Road and rail networks and associated land: motorways and railways, including associated installations.
• 123 - Port areas: infrastructure of port areas, including quays, dockyards and marinas.
• 124 - Airports: airports installations as runways, buildings and associated land.
• 131 - Mineral extraction sites: areas of open-pit extraction of construction materials (sandpits, quarries) or other minerals (open-cast mines).
• 132 - Dump sites: public, industrial or mine dump sites.
• 133 - Construction sites: spaces under construction development, soil or bedrock excavations, earthworks.
• 141 - Green urban areas: areas with vegetation within urban fabric.
• 211 - Non-irrigated arable land: cultivated land parcels under rain-fed agricultural use for annually harvested non-permanent crops, normally under a crop rotation system.
• 212 - Permanently irrigated arable land: cultivated land parcels under agricultural use for arable crops that are permanently or periodically irrigated, using a permanent infrastructure (irrigation channels, drainage network and additional irrigation facilities).
• 213 - Rice fields: cultivated land parcels prepared for rice production, consisting of periodically flooded flat surfaces with irrigation channels.
• 221 - Vineyards: areas planted with vines.
• 222 - Fruit trees and berry plantations: cultivated parcels planted with fruit trees and shrubs, including nuts, intended for fruit production.
• 223 - Olive groves: cultivated areas planted with olive trees, including mixed occurrence of vines on the same parcel.
• 231 - Pastures: meadows where dispersed trees and shrubs may occupy up to 50% of surface. These meadows are characterized by rich floristic composition.
• 311 - Broad-leaved forest: vegetation formation composed principally of trees, including shrub and bush understorey, where broad-leaved species predominate. 
• 312 - Coniferous forest: vegetation formation composed principally of trees, including shrub and bush understorey, where coniferous species predominate. 
• 321 - Natural grasslands: grasslands under no or moderate human influence. Low productivity grasslands. Often situated in areas of rough, uneven ground, frequently include rocky areas, or patches of other (semi-)natural vegetation. 
• 322 - Moors and heathland: vegetation with low and closed cover, dominated by bushes, shrubs (heather, briars, broom, gorse, laburnum, etc.
• 323 - Sclerophyllous vegetation: bushy sclerophyllous vegetation in a climax stage of development, including maquis, matorral and garrigue. 
• 324 - Transitional woodland-shrub: transitional bushy and herbaceous vegetation with occasional scattered trees. Can represent either woodland degradation or forest regeneration / re-colonization 
• 331 - Beaches, dunes, sands: unvegetated expanses of sand or pebble/gravel, in coastal or continental locations, like beaches, dunes, gravel pads.
• 332 - Bare rocks: scree, cliffs, rock outcrops, including areas of active erosion.
• 333 - Sparsely vegetated areas:  Areas with sparse vegetation, covering 10-50% of the surface.
• 334 - Burnt areas:  Areas affected by recent fires.
• 335 - Glaciers and perpetual snow:  Land covered by ice or permanent snowfields
• 411 - Inland wetlands: low-lying land usually flooded in winter, and more or less saturated by water all year round; and wetlands with accumulation of considerable amount of decomposed moss and vegetation matter. 
• 421 - Maritime wetlands: vegetated low-lying areas in the coastal zone, above the high-tide line, susceptible to flooding by seawater; salt-pans for extraction of salt from salt water; and coastal zones under tidal influence between open sea and land.
• 511 - Water courses: natural or artificial water courses serving as water drainage channels
• 512 - Water bodies: natural or artificial water surfaces covered by standing water most of the year
• 521 - Coastal lagoons: stretches of salt or brackish water in coastal areas which are separated from the sea by a tongue of land or other similar topography
• 522 - Estuaries: the mouth of a river under tidal influence within which the tide ebbs and flows.
• 523 - Sea and ocean: zone seaward of the lowest tide limit.

The samples were obtained from the geographic location of LUCAS (in-situ source) and the centroid of all polygons from CLC maps, harmonized according to the above classes and organized by year, where each unique combination of longitude, latitude and year was considered as a independent sample. Some specific CORINE samples (i.e. 111, 122, 131, 141, 211, 221, 222, 223, 231, 311, 312, 321, 411, 512) were filtered according to convergence with existing mapping products (OSM roads, railways and buildings; Copernicus High-Res. Layers - HRL), where, for example, “111 - Urban fabric” samples located in low density building areas (> 50% according to Copernicus-OSM building layer) were removed. The LUCAS points with a unique land-cover class received a confidence rating of 100%, while CORINE received 85%. Using these filtered samples a spacetime overlay was performed (check the code demonstration in eumap library) considering several raster layers for Continental Europe: four season quantiles for GLAD Landsat ARD (spectral bands and indices - Potapov, 2020), DTM-based elevation and slope (Hengl, 2020), VIIRS/SUOMI NPP night lights (Hillger, 2013), Global surface water frequency (Pekel, 2016), and geometric minimum and maximum temperature derived according to Kilibarda, 2014.

The provided samples has 5,362,229 rows and 263 columns, including the geographic location, the metadata of land cover harmonization, the mapped classes according to OSM and HRL, and the result of the spacetime overlay (178 covariates). 

Use the following Python/R code to open the files lcv_landcover.hcl_lucas.corine_harm.samples.overlaid.*

import joblib
samples = joblib.load('lcv_landcover.hcl_lucas.corine_harm.samples.overlaid')

readRDS('lcv_landcover.hcl_lucas.corine_harm.samples.overlaid.rds')

These samples were used to train a spatiotemporal model, which predicts the land cover for continental Europe over 20 years (2000 - 2019). To access the predictions results (dominant class, probabilities and uncertainties) use the following services:

• Open Data Science Europe viewer: https://maps.opendatascience.eu
• S3 Cloud Object: https://medium.com/swlh/europe-from-above-space-time-machine-learning-reveals-our-changing-environment-1b05cb7be520

A publication describing, in detail, all processing steps, accuracy assessment and general analysis of land-cover changes in continental Europe is under preparation. To suggest any improvement/fix use https://gitlab.com/geoharmonizer_inea/spatial-layers/-/issues