Continental-Scale Land Cover Mapping at 10 m Resolution Over Europe (ELC10)

Abstract

Land cover maps are important tools for quantifying the human footprint on the environment and facilitate reporting and accounting to international agreements addressing the Sustainable Development Goals. Widely used European land cover maps such as CORINE (Coordination of Information on the Environment) are produced at medium spatial resolutions (100 m) and rely on diverse data with complex workflows requiring significant institutional capacity. We present a 10 m resolution land cover map (ELC10) of Europe based on a satellite-driven machine learning workflow that is annually updatable. A random forest classification model was trained on 70K ground-truth points from the LUCAS (Land Use/Cover Area Frame Survey) dataset. Within the Google Earth Engine cloud computing environment, the ELC10 map can be generated from approx. 700 TB of Sentinel imagery within approx. 4 days from a single research user account. The map achieved an overall accuracy of 90% across eight land cover classes and could account for statistical unit land cover proportions within 3.9% (R2 = 0.83) of the actual value. These accuracies are higher than that of CORINE (100 m) and other 10 m land cover maps including S2GLC and FROM-GLC10. Spectro-temporal metrics that capture the phenology of land cover classes were most important in producing high mapping accuracies. We found that the atmospheric correction of Sentinel-2 and the speckle filtering of Sentinel-1 imagery had a minimal effect on enhancing the classification accuracy (< 1%). However, combining optical and radar imagery increased accuracy by 3% compared to Sentinel-2 alone and by 10% compared to Sentinel-1 alone. The addition of auxiliary data (terrain, climate and night-time lights) increased accuracy by an additional 2%. By using the centroid pixels from the LUCAS Copernicus module polygons we increased accuracy by <1%, revealing that random forests are robust against contaminated training data. Furthermore, the model requires very little training data to achieve moderate accuracies—the difference between 5K and 50K LUCAS points is only 3% (86 vs. 89%). This implies that significantly less resources are necessary for making in situ survey data (such as LUCAS) suitable for satellite-based land cover classification. At 10 m resolution, the ELC10 map can distinguish detailed landscape features like hedgerows and gardens, and therefore holds potential for aerial statistics at the city borough level and monitoring property-level environmental interventions (e.g., tree planting). Due to the reliance on purely satellite-based input data, the ELC10 map can be continuously updated independent of any country-specific geographic datasets.

The Third Generation of Pan-Canadian Wetland Map at 10 m Resolution Using Multisource Earth Observation Data on Cloud Computing Platform

“…The resulting pan-Canadian wetland map here extends on our previous work focusing on generating high-resolution wetland data, by which an overall improvement of about 10% and 5% in accuracy obtained relative to the CWIM1 [21] and CWIM2 [22], respectively. The accuracy of 10 m resolution CWIM3 produced here is 90.53% that are comparable with other Sentinel-based large-scale land cover mapping globally [32], [34]. However, general land cover classes (e.g., water, bareland, and cropland) are much easier to be delineated compared to ecologically similar wetland classes separated in this study.…”

“…This will also increase processing requirements due to a two-fold increase in data inputs. Future work should also integrate additional topographic variables proven to effectively detect wetlands and were not used in the development of the CWIM3, such as the topographic position index [32] and topographic wetness index [33].…”

“…In both cases, significant improvement was achieved for classifications through the ecozone-by-ecozone strategy. Although previous studies suggested that for large-scale land cover mapping either inclusion of auxiliary data or applying different classification models within a small area should be sufficient [32]. This does not hold for a country like Canada, where ecological and climatic features can vary even within a single ecozone.…”

MetaComNet: A random forest‐based framework for making spatial predictions of plant–pollinator interactions

“…As a proxy for climatic conditions, we used elevation above sea level, obtained from a digital elevation model with a 50 m resolution (Norwegian Mapping Authority, 2016). As proxies for landscape conditions, we used the European ELC10 land cover map (Venter & Sydenham, 2021) to calculate: the proportion of grassland area and Shannon landscape diversity within a 250 m radius. As a proxy for nesting conditions, we calculated the distance to sand‐dominated geological deposits, that is, soils with a high permeability (Geological Survey of Norway, 2011).…”

A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery

“…Land Use and Land Cover (LULC) maps are essential tools for documenting the changes in the environment and for quantifying the human footprint on the Earth's surface [1]. Due to the increasing availability of high resolution, frequently collected remote sensing data, there is a clear need for a highly automated data processing workflow to update the land cover and land use changes [2].…”