Overview: Vegetation tree species sample points Traceability (lineage): This is an original dataset produced with a machine learning framework which used a combination of point datasets and raster datasets as inputs. Point dataset is a harmonized collection of tree occurrence data, comprising observations from National Forest Inventories (EU-Forest), GBIF and LUCAS. The complete dataset is available on Zenodo. Raster datasets used as input are: harmonized and gapfilled time series of seasonal aggregates of the Landsat GLAD ARD dataset (bands and spectral indices); monthly time series air and surface temperature and precipitation from a reprocessed version of the Copernicus ERA5 dataset; long term averages of bioclimatic variables from CHELSA, tree species distribution maps from the European Atlas of Forest Tree Species; elevation, slope and other elevation-derived metrics; long term monthly averages snow probability and long term monthly averages of cloud fraction from MODIS. For a more comprehensive list refer to Bonannella et al. (2022) (in review, preprint available at: https://doi.org/10.21203/rs.3.rs-1252972/v1). Scientific methodology: Probability and uncertainty maps were the output of a spatiotemporal ensemble machine learning framework based on stacked regularization. Three base models (random forest, gradient boosted trees and generalized linear models) were first trained on the input dataset and their predictions were used to train an additional model (logistic regression) which provided the final predictions. More details on the whole workflow are available in the listed publication. Usability: Probability maps can be used to detect potential forest degradation and compositional change across the time period analyzed. Some possible applications for these topics are explained in the listed publication. Uncertainty quantification: Uncertainty is quantified by taking the standard deviation of the probabilities predicted by the three components of the spatiotemporal ensemble model. Data validation approaches: Distribution maps were validated using a spatial 5-fold cross validation following the workflow detailed in the listed publication. Completeness: The raster files perfectly cover the entire Geo-harmonizer region as defined by the landmask raster dataset available here. Consistency: Areas which are outside of the calibration area of the point dataset (Iceland, Norway) usually have high uncertainty values. This is not only a problem of extrapolation but also of poor representation in the feature space available to the model of the conditions that are present in this countries. Positional accuracy: The rasters have a spatial resolution of 30m. Temporal accuracy: The maps cover the period 2000 - 2020, each map covers a certain number of years according to the following scheme: (1) 2000--2002, (2) 2002--2006, (3) 2006--2010, (4) 2010--2014, (5) 2014--2018 and (6) 2018--2020 Thematic accuracy: Both probability and uncertainty maps contain values from 0 to 100: in the case of probability maps, they indicate the probability of occurrence of a single individual of the target species, while uncertainty maps indicate the standard deviation of the ensemble model.

Overview: Potential Natural Vegetation (PNV): potential probability of occurrence for the Olive tree from 2018 to 2020 Traceability (lineage): This is an original dataset produced with a machine learning framework which used a combination of point datasets and raster datasets as inputs. Point dataset is a harmonized collection of tree occurrence data, comprising observations from National Forest Inventories (EU-Forest), GBIF and LUCAS. The complete dataset is available on Zenodo. Raster datasets used as input are: monthly time series air and surface temperature and precipitation from a reprocessed version of the Copernicus ERA5 dataset; long term averages of bioclimatic variables from CHELSA; elevation, slope and other elevation-derived metrics and long term monthly averages snow probability. For a more comprehensive list refer to Bonannella et al. (2022) (in review, preprint available at: https://doi.org/10.21203/rs.3.rs-1252972/v1). Scientific methodology: Probability and uncertainty maps were the output of a spatiotemporal ensemble machine learning framework based on stacked regularization. Three base models (random forest, gradient boosted trees and generalized linear models) were first trained on the input dataset and their predictions were used to train an additional model (logistic regression) which provided the final predictions. More details on the whole workflow are available in the listed publication. Usability: Probability maps are particularly useful when compared with existing products of potential distribution of species or when combined with maps of realized distribution: gaps in potential and realized distribution can be identified and used as information for future programs of tree planting or forest restoration. Uncertainty quantification: Uncertainty is quantified by taking the standard deviation of the probabilities predicted by the three components of the spatiotemporal ensemble model. Data validation approaches: Distribution maps were validated using a spatial 5-fold cross validation following the workflow detailed in the listed publication. Completeness: The raster files perfectly cover the entire Geo-harmonizer region as defined by the landmask raster dataset available here. Consistency: Areas which are outside of the calibration area of the point dataset (Iceland, Norway) usually have high uncertainty values. This is not only a problem of extrapolation but also of poor representation in the feature space available to the model of the conditions that are present in this countries. Positional accuracy: The rasters have a spatial resolution of 30m. Temporal accuracy: The maps cover the period 2018 - 2020 Thematic accuracy: Both probability and uncertainty maps contain values from 0 to 100: in the case of probability maps, they indicate the probability of occurrence of a single individual of the target species, while uncertainty maps indicate the standard deviation of the ensemble model.

331: Natural non-vegetated expanses of sand or pebble/gravel, in coastal or continental locations, like beaches, dunes, gravel pads; including beds of stream channels with torrential regime. Vegetation covers maximum 10%.

523: Zone seaward of the lowest tide limit.

Overview: 242: Mosaic of small cultivated land parcels with different cultivation types(annual and permanent crops, as well as pastures), potentially with scattered houses or gardens. Traceability (lineage): This dataset was produced with a machine learning framework with several input datasets, specified in detail in Witjes et al., 2022 (in review, preprint available at https://doi.org/10.21203/rs.3.rs-561383/v3 ) Scientific methodology: The single-class probability layers were generated with a spatiotemporal ensemble machine learning framework detailed in Witjes et al., 2022 (in review, preprint available at https://doi.org/10.21203/rs.3.rs-561383/v3 ). The single-class uncertainty layers were calculated by taking the standard deviation of the three single-class probabilities predicted by the three components of the ensemble. The HCL (hard class) layers represents the class with the highest probability as predicted by the ensemble. Usability: The HCL layers have a decreasing average accuracy (weighted F1-score) at each subsequent level in the CLC hierarchy. These metrics are 0.83 at level 1 (5 classes):, 0.63 at level 2 (14 classes), and 0.49 at level 3 (43 classes). This means that the hard-class maps are more reliable when aggregating classes to a higher level in the hierarchy (e.g. 'Discontinuous Urban Fabric' and 'Continuous Urban Fabric' to 'Urban Fabric'). Some single-class probabilities may more closely represent actual patterns for some classes that were overshadowed by unequal sample point distributions. Users are encouraged to set their own thresholds when postprocessing these datasets to optimize the accuracy for their specific use case. Uncertainty quantification: Uncertainty is quantified by taking the standard deviation of the probabilities predicted by the three components of the spatiotemporal ensemble model. Data validation approaches: The LULC classification was validated through spatial 5-fold cross-validation as detailed in the accompanying publication. Completeness: The dataset has chunks of empty predictions in regions with complex coast lines (e.g. the Zeeland province in the Netherlands and the Mar da Palha bay area in Portugal). These are artifacts that will be avoided in subsequent versions of the LULC product. Consistency: The accuracy of the predictions was compared per year and per 30km*30km tile across europe to derive temporal and spatial consistency by calculating the standard deviation. The standard deviation of annual weighted F1-score was 0.135, while the standard deviation of weighted F1-score per tile was 0.150. This means the dataset is more consistent through time than through space: Predictions are notably less accurate along the Mediterrranean coast. The accompanying publication contains additional information and visualisations. Positional accuracy: The raster layers have a resolution of 30m, identical to that of the Landsat data cube used as input features for the machine learning framework that predicted it. Temporal accuracy: The dataset contains predictions and uncertainty layers for each year between 2000 and 2019. Thematic accuracy: The maps reproduce the Corine Land Cover classification system, a hierarchical legend that consists of 5 classes at the highest level, 14 classes at the second level, and 44 classes at the third level. Class 523: Oceans was omitted due to computational constraints.

Overview: ERA5-Land is a reanalysis dataset providing a consistent view of the evolution of land variables over several decades at an enhanced resolution compared to ERA5. ERA5-Land has been produced by replaying the land component of the ECMWF ERA5 climate reanalysis. Reanalysis combines model data with observations from across the world into a globally complete and consistent dataset using the laws of physics. Reanalysis produces data that goes several decades back in time, providing an accurate description of the climate of the past. Processing steps: The original hourly ERA5-Land air temperature 2 m above ground and dewpoint temperature 2 m data has been spatially enhanced from 0.1 degree to 30 arc seconds (approx. 1000 m) spatial resolution by image fusion with CHELSA data (https://chelsa-climate.org/). Subsequently, the temperature time series have been aggregated on a daily basis. From these, daily relative humidity has been calculated for the time period 01/2000 - 07/2021. Relative humidity (rh2m) has been calculated from air temperature 2 m above ground (Ta) and dewpoint temperature 2 m above ground (Td) using the formula for saturated water pressure from Wright (1997): maximum water pressure = 611.21 * exp(17.502 * Ta / (240.97 + Ta)) actual water pressure = 611.21 * exp(17.502 * Td / (240.97 + Td)) relative humidity = actual water pressure / maximum water pressure The resulting relative humidity has been aggregated to decadal averages. Each month is divided into three decades: the first decade of a month covers days 1-10, the second decade covers days 11-20, and the third decade covers days 21-last day of the month. Resultant values have been converted to represent percent * 10, thus covering a theoretical range of [0, 1000]. File naming scheme (YYYY = year; MM = month; dD = number of decade): ERA5_land_rh2m_avg_decadal_YYYY_MM_dD.tif Projection + EPSG code: Latitude-Longitude/WGS84 (EPSG: 4326) Spatial extent: north: 82:00:30N south: 18N west: 32:00:30W east: 70E Spatial resolution: 30 arc seconds (approx. 1000 m) Temporal resolution: Decadal Pixel values: Percent * 10 (scaled to Integer; example: value 738 = 73.8 %) Software used: GDAL 3.2.2 and GRASS GIS 8.0.0 Original dataset license: https://apps.ecmwf.int/datasets/licences/copernicus/ Processed by: mundialis GmbH & Co. KG, Germany (https://www.mundialis.de/) Reference: Wright, J.M. (1997): Federal meteorological handbook no. 3 (FCM-H3-1997). Office of Federal Coordinator for Meteorological Services and Supporting Research. Washington, DC Acknowledgements: This study was partially funded by EU grant 874850 MOOD. The contents of this publication are the sole responsibility of the authors and don't necessarily reflect the views of the European Commission.

Normalized Difference Water Index (NDWI) from MODIS data for Mauritania at 30 arc seconds (ca. 1000 meter) resolution (2019 - 2023). Source data: - MODIS/Terra Surface Reflectance 8-Day L3 Global 500 m SIN Grid (MOD09A1 v061): https://lpdaac.usgs.gov/products/mod09a1v061/ The Moderate Resolution Imaging Spectroradiometer (MODIS) Terra MOD09A1 Version 6.1 product provides an estimate of the surface spectral reflectance of Terra MODIS Bands 1 through 7 corrected for atmospheric conditions such as gasses, aerosols, and Rayleigh scattering. Along with the seven 500 meter (m) reflectance bands are two quality layers and four observation bands. For each pixel, a value is selected from all the acquisitions within the 8-day composite period. The criteria for the pixel choice include cloud and solar zenith. When several acquisitions meet the criteria the pixel with the minimum channel 3 (blue) value is used. For the time period January 2019 - December 2023, the NDWI has been calculated from the Terra MOD09A1 Version 6.1 product. The layers B02 (near infrared) and B06 (shortwave infrared) of the original data have been processed. Bad quality pixels or pixels with snow/ice and/or cloud cover have been masked using the provided quality assurance (QA) layer. The time series has been gapfilled with a temporal and a spatial approach. Gaps in the time series were filled with a harmonic analysis of time series using six frequencies to also model relatively short-term changes in NDWI. Only missing values were replaced by modelled values. NDWI was calculated as the normalized difference of the bands B02 (near infrared) and B06 (shortwave infrared) with: NDWI = (B02 – B06) / (B02 + B06). This NDWI represents vegetation water content. The 8-day data are then aggregated to monthly temporal resolution using the average and reprojected to Latitude-Longitude/WGS84. File naming: ndwi_monthly_YYYY_MM_30arcsec.tif e.g.: ndwi_monthly_2023_12_30arcsec.tif The date within the filename are year and month of aggregated timestamp. Pixel values: Raster values are in the range [0, 2000]. Real NDWI in the range [-1, 1] can be retrieved with: NDWI = raster_value * 0.001 - 1. Projection + EPSG code: Latitude-Longitude/WGS84 (EPSG: 4326) Spatial extent: north: 28N south: 14N west: 18W east: 4W Temporal extent: January 2019 - December 2023 Spatial resolution: 30 arc seconds (approx. 1000 m) Temporal resolution: monthly Software used: GRASS GIS 8.3.2 Format: GeoTIFF Original dataset license: All data products distributed by NASA's Land Processes Distributed Active Archive Center (LP DAAC) are available at no charge. The LP DAAC requests that any author using NASA data products in their work provide credit for the data, and any assistance provided by the LP DAAC, in the data section of the paper, the acknowledgement section, and/or as a reference. The recommended citation for each data product is available on its Digital Object Identifier (DOI) Landing page, which can be accessed through the Search Data Catalog interface. For more information see: https://lpdaac.usgs.gov/products/mod13a2v061/ Processed by: mundialis GmbH & Co. KG, Germany (https://www.mundialis.de/) Contact: mundialis GmbH & Co. KG, info@mundialis.de Acknowledgements: This study was partially funded by EU grant 874850 MOOD. The contents of this publication are the sole responsibility of the authors and don't necessarily reflect the views of the European Commission.

osm: Industrial building aggregated and rasterized from OSM polygons, first to 10m spatial resolution and after downsampled to 30m by spatial average.

Preview of the hillshading map for EU.

Overview: 313: Vegetation formation composed principally of trees, including shrub and bush understory,where neither broad-leaved nor coniferous species predominate. Traceability (lineage): This dataset was produced with a machine learning framework with several input datasets, specified in detail in Witjes et al., 2022 (in review, preprint available at https://doi.org/10.21203/rs.3.rs-561383/v3 ) Scientific methodology: The single-class probability layers were generated with a spatiotemporal ensemble machine learning framework detailed in Witjes et al., 2022 (in review, preprint available at https://doi.org/10.21203/rs.3.rs-561383/v3 ). The single-class uncertainty layers were calculated by taking the standard deviation of the three single-class probabilities predicted by the three components of the ensemble. The HCL (hard class) layers represents the class with the highest probability as predicted by the ensemble. Usability: The HCL layers have a decreasing average accuracy (weighted F1-score) at each subsequent level in the CLC hierarchy. These metrics are 0.83 at level 1 (5 classes):, 0.63 at level 2 (14 classes), and 0.49 at level 3 (43 classes). This means that the hard-class maps are more reliable when aggregating classes to a higher level in the hierarchy (e.g. 'Discontinuous Urban Fabric' and 'Continuous Urban Fabric' to 'Urban Fabric'). Some single-class probabilities may more closely represent actual patterns for some classes that were overshadowed by unequal sample point distributions. Users are encouraged to set their own thresholds when postprocessing these datasets to optimize the accuracy for their specific use case. Uncertainty quantification: Uncertainty is quantified by taking the standard deviation of the probabilities predicted by the three components of the spatiotemporal ensemble model. Data validation approaches: The LULC classification was validated through spatial 5-fold cross-validation as detailed in the accompanying publication. Completeness: The dataset has chunks of empty predictions in regions with complex coast lines (e.g. the Zeeland province in the Netherlands and the Mar da Palha bay area in Portugal). These are artifacts that will be avoided in subsequent versions of the LULC product. Consistency: The accuracy of the predictions was compared per year and per 30km*30km tile across europe to derive temporal and spatial consistency by calculating the standard deviation. The standard deviation of annual weighted F1-score was 0.135, while the standard deviation of weighted F1-score per tile was 0.150. This means the dataset is more consistent through time than through space: Predictions are notably less accurate along the Mediterrranean coast. The accompanying publication contains additional information and visualisations. Positional accuracy: The raster layers have a resolution of 30m, identical to that of the Landsat data cube used as input features for the machine learning framework that predicted it. Temporal accuracy: The dataset contains predictions and uncertainty layers for each year between 2000 and 2019. Thematic accuracy: The maps reproduce the Corine Land Cover classification system, a hierarchical legend that consists of 5 classes at the highest level, 14 classes at the second level, and 44 classes at the third level. Class 523: Oceans was omitted due to computational constraints.