February 2020

• Standard legends and metadata for the project outputs (i.e. land cover legends, environmental quality indicators etc),
• Automated harmonization/conversion of data,
• Easy access (for developers) to all layers produced by the Geo-harmonizer project (PostGIS DBs and Cloud-Optimized GeoTIFF),
• Optimized spatial overlay, Ensemble Machine Learning and prediction,
• Automated generation of mosaics and export of data,
• Tutorials for basic users and advanced developers,

• How to access data (for non-experts),
• How to use data for analysis and decision making (i.e. software package tutorial)
• How to use the REST API,
• How to get support and recommend additions / changes,

• Standardized metadata-service (OGC CSW)
• Standardized metadata profile (based on ISO 19115)
• Easy to use interface for search and access of data provided by the Geo-harmonizer platform

• OpenStreetMap: snapshot of the OSM layers (vectors) for EU (PENDING),
• Copernicus DEM – Global and European Digital Elevation Model for Europe,
• Copernicus/EEA layers 100 m: automatically downloaded (synced) and stacked to the same grid (bounding box, coordinate system, pixel size) layers responding to the target resolution of 100 m (PENDING),
• Copernicus/EEA layers 20 m: automatically downloaded (synced) and stacked layers (where necessary downscaled/upscaled to 30 m),
• EO Sentinel 1/2 and Landsat 6/7/8 layers 30 m: cloud-free mosaics (2000–2020) produced via the DIAS infrastructure or similar and using the 25% quantile and interquantile range (IQR) parameters (PENDING),
• Copernicus / Vito products: including CORINE land cover and similar,
• Other: other layers generated from global or regional projects such as the Global Surface Water Watch, Global Forest Watch and similar.

• For point data sets (training points) and lines: PostGIS DB,
• For complete spatial layers (rasters): Cloud-Optimized GeoTIFF,

• Land use classes being consistent and complete for the whole area,
• OSM-based density of green/protected areas (data digitized by citizens and not available at EEA),
• Administrative boundaries, transportation infrastructure (Veeckman et al., 2017) matching the open national maps (ensemble merge) i.e. no border problems or artifacts,

• Hard-class map of the land cover per year for .period 2000–2020,
• probability maps per land cover class,
• model errors (predicted uncertainty) per land cover class,
• accuracy assessment report and comparison using finer resolution local data sets.

• Air quality maps: e.g. PM2.5 based on automated mapping using Machine Learning, MODIS data and other auxiliary (meteorological) data,
• Maps of soil nutrients including heavy metals at 30 m: i.e. a more detailed version of the soil property maps produced by JRC (Ballabio et al., 2019),

The climate change indicators should include:

• Long-term trends 2000-2020 in temperatures based on MODIS LST products (Metz et al., 2017),
• Rainfall images (2000-2020; e.g. based on Brocca et al., 2013; 2019),
• Flood events images (2000-2020),
• Fires (2000-2020), (Lizundia-Loiola et al., 2018),

Detailed map of potential natural vegetation. This should include:

• PNV hard-class map based on the CORINE land cover classes at 30 m spatial resolution for years 2000 and 2020,
• PNV maps showing probability maps for the EU forest tree species (76; see: San-Miguel-Ayanz et al. 2016) with model errors (uncertainty),

• Easy searching / browsing of layers,
• Viewing layers including multiple layers at the same time,
• Metadata functionality,
• Download functionality,

Standardized interface and user-interface, that allow to search, browse and identify geodata sets based on their metadata.

• User segmentation report
• Use case initial developments based on user profiling
• List of contacts of potential users for the Geo-harmonizer products
• The quality assessment reports of the Geo-harmonizer map products
• Assessment of the European market for the Geo-harmonizer products
• Estimation of investment and return for the Geo-harmonizer map products

• How to access data (for non-experts),
• How to use data for analysis and decision making (package and the REST API tutorial),
• How to get support and recommend additions / changes,

On-line tutorials (wiki’s) including code tutorials on the project website (opendatascience.eu) and on the GitLab homepage.