Abstract
Widely used European land cover maps such as CORINE are produced at medium spatial resolutions (100 m) and rely on diverse data with complex workflows requiring significant institutional capacity. We present a high resolution (10 m) 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 8 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. We found that atmospheric correction of Sentinel-2 and speckle filtering of Sentinel-1 imagery had minimal effect on enhancing 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 conversion of LUCAS points into homogenous polygons under the Copernicus module 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%). 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).
Abstract (translated)
URL
https://arxiv.org/abs/2104.10922
