Probabilistic Data Generating Process-based Crop Type Map for the EU 2010-2020
General Description This dataset consists of probabilistic crop type maps for the EU-28 for the years 2010-2020 that distinguish 28 crop types at 1km resolution (EPSG:3035). The maps were generated using the Data Generating Process-based procedure developed by Baumert, Heckelei and Storm (2024) [htt...
Gespeichert in:
Hauptverfasser: | , , |
---|---|
Format: | Dataset |
Sprache: | eng |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | General Description
This dataset consists of probabilistic crop type maps for the EU-28 for the years 2010-2020 that distinguish 28 crop types at 1km resolution (EPSG:3035). The maps were generated using the Data Generating Process-based procedure developed by Baumert, Heckelei and Storm (2024) [https://doi.org/10.1016/j.ecoinf.2024.102836]. We refer to this paper for details on the generation and validation of the maps. The code used to create the maps including a detailed list of the input data can be found here: GitHub - JoBaumert/Probabilistic_Crop_Mapping_EU .
Downloadable Data
The file “EU_expected_crop_shares.zip” consists of 11 raster files, one for each year from 2010 – 2020. The raster files indicate the expected shares for each of the 28 distinguished crop types in a grid cell for the entire EU-28 (see readme.txt contained in the zipped folder). Note that this raster file does not contain uncertainty information.
The other 28 zip files contain the entire crop map ensemble (i.e., including uncertainty information), each for one of the EU countries and the United Kingdom. Each of those zip files contain 11 raster files, one for each year from 2010 – 2020. Each raster file has 2830 bands: the first two bands indicate the weight of the cell (proportional to the utilized agricultural area in a cell) and the estimated number of agricultural fields in a cell, respectively. The next 28 bands indicate the expected shares for each of the 28 crops in the respective cell. The remaining 2800 bands compose the crop type map ensemble, i.e., 100 simulated crop shares for each of the 28 crops. The zipped country folder also includes a csv file named “bands” that describes which band refers to which crop. Note that all crop shares were multiplied by 1000 when writing them to the raster files (saving them as integers requires less storage capacity), i.e., if a crop share is 0.325 or 32.5% it will appear as 325 in the raster files.
The distinguished crops are (with abbreviation used in "bands.csv"):
Apples and other fruits, nuts and berries (APPL+OFRU)
Barley (BARL)
Citrus fruits (CITR)
Durum wheat (DWHE)
Flowers and ornamental plants (FLOW)
Grassland (GRAS)
Maize (both green maize as well as grain maize, LMAIZ)
Rape and turnip (LRAPE)
Nurseries (NURS)
Oats (OATS)
Other cereals (OCER)
Other permanent crops (OCRO)
Other forage plants (OFAR)
Other industrial plants (OIND)
Olives (OLIVGR)
Rice (PARI)
Potatoes (POTA)
Pulses (PULS)
Fodder roots and brassicas (ROOF)
Ry |
---|---|
DOI: | 10.5281/zenodo.13819663 |