EO-based area monitoring markers computed over the Lithuanian pilot region (2021)

In the context of the EU-funded project DIONE (No. 870378), the following EO-based monitoring marker maps were released over the two pilot regions, containing the results produced from a set of image analysis and machine learning techniques. The latest explores the benefits of Copernicus's multispectral high-resolution Sentinel-2 data acquired from 01-01-2021 until 26-07-2021 and provides tailored information on the needs of European paying agencies (e.g. CAPO and NPA), expressed with the following markers.

Mowing marker: used to detect mowing events on meadow/grass like Features Of Interest (FOI)

Mean-NDVI marker: used to detect erroneous claims with no vegetation

Homogeneity marker: used to determine if a parcel geometry really consists of a single crop or if there are multiple things growing on the parcel

Bare soil marker: used to detect observation where bare soil is present on the feature of interest. This indicates agricultural activity on the FOI (plowing, harvest)

Similarity and distance markers: used to give additional context to the crop classification and to detect erroneous claims

Land marker: used to detect the land type and non-productive EFAs of the FOI

Crop-type marker: used to detect the specific crop growing on the FOI

This dataset is comprised of the two subsequent geopackage files, (i) the "test_area_markers_summary.gpkg" and (ii) "crop_marker_summary_full_country.gpkg", which were computed for the Lithuanian pilot region.  Descriptions are given below.

Test area markers dataset: There is a total of 59995 FOIs in the test areas. Out of these, markers are computed on **54809** FOIs that contain more than 1 Sentinel-2 pixel.

Description of the information contained in the corresponding "test area markers" dataset

Attribute Table name
Description

POLY_ID 
Reference ID of the polygon  

LABEL
Declared crop group

N_VALID_OBS
Number of valid Sentinel-2 observations used for computing the markers

HOMOGENEITY_PROBA
The probability that the FOI is homogeneous

N_MW_OBS_MAY_OCT
Number of detected mowing observations between May and October

N_BS_OBS_APR_OCT
Number of detected mowing observations between April and October

SIMILARITY_SCORE
The normalised ([0, 100]) representation of the χ2 value of the NDVI time-series difference between the target FOI and the neighbouring FOIs

DISTANCE_SCORE
The normalised ([0, 100]) representation of the Median Euclidean distance between the NDVI time-series of the target FOI and the neighbouring FOIs having assumed crop type declared

MEAN_NDVI
Μean NDVI value of all valid observations of the  FOI

PREDICTED_CROP
Predicted crop group

CROP_PREDICTION_SCORE
The pseudo-probability of the crop group prediction

Crop marker full country dataset: There is a total of **1127455** FOIs for the full country GSAA dataset. Out of these, markers are computed on **1042671** FOIs that contain more than 1 Sentinel-2 pixel.

Description of the information contained in the corresponding "crop marker full country" dataset

Attribute Table name
Description

POLY_ID
Reference ID of the polygon

LABEL
Declared crop group

PREDICTED_CROP
Predicted crop group

CROP_PREDICTION_SCORE
The pseudo-probability of the crop group prediction