============================================================
Dataset: European Forest Disturbance Atlas (EFDA) v3.0
Source: Landsat 4-9
Period: 1985 to 2024
Spatial resolution: 30 m
Projection: EPSG:3035
Version: 3.0
Release: 09.2026
============================================================
[Layers]
============================================================

Name: annual_disturbance_<country>_v30.tif
Description: Annual forest disturbance occurrence.
Type: Multi-band GeoTIFF, Uint8
Period: 1985 (band1) to 2024 (band40)
Values: 1 = Disturbed
NoData: 255

Name: disturbance_probability_<country>_v30.tif
Description: Annual probability associated with forest disturbance
detection from the deep learning-based disturbance detection model.
Type: Multi-band GeoTIFF, Uint8
Period: 1985 (band1) to 2024 (band40)
Values: 0 to 100
NoData: 255

Name: disturbance_severity_<country>_v30.tif
Description: Annual spectral severity of forest disturbances, based on changes in 
the Normalized Burn Ratio (NBR) relative to pre-disturbance conditions.
Type: Multi-band GeoTIFF, Int16
Period: 1985 (band1) to 2024 (band40)
Values: -1,000 to 1,000 (original_dNBR * 1000)
NoData: -32768

Name: disturbance_agent_<country>_v30.tif
Description: Annual attribution of the causal agent associated with
detected forest disturbances.
Type: Multi-band GeoTIFF, Uint8
Period: 1985 (band1) to 2024 (band40)
Values: 1 = Wind / bark beetle complex; 2 = Fire; 3 = Harvest
NoData: 255

Name: disturbance_agent_aggregated_<country>_v30.tif
Description: Summary of disturbance-agent attribution over the complete
1985 to 2024 time series. Pixels affected by different disturbance agents
in different years are classified as mixed agents.
Type: GeoTIFF, Uint8
Values: 1 = Wind / bark beetle complex; 2 = Fire; 3 = Harvest;
4 = Mixed agents (different agents occurred at different times (e.g., harvest in 1995 and fire in 2022)
NoData: 255

Name: number_disturbances_<country>_v30.tif
Description: Total number of distinct forest disturbance events detected
at each pixel during the complete observation period.
Type: GeoTIFF, Uint8
Values: 1 = One disturbance; 2 = Two disturbances; N = Number of detected disturbance Events
NoData: 255

Name: latest_disturbance_<country>_v30.tif
Description: Year of the most recent forest disturbance detected at each
pixel during the observation period.
Type: GeoTIFF, Int16
Values: 1985 to 2024 = Year of latest disturbance
NoData: -9999

Name: forestlanduse_mask_<country>_v30.tif
Description: Forest land-use mask defining the spatial domain in which
EFDA forest disturbances are mapped.
Type: GeoTIFF, Uint8
Values: 1 = Forest land
NoData/Non-forest: 0

============================================================
[Methods and References to be cited]
============================================================

Disturbance-agent attribution, forest land use in previous and current version 
of EFDA follows the methodology described in:
- Viana-Soto, A., & Senf, C. (2025). The European Forest Disturbance Atlas: a forest disturbance monitoring system using the Landsat archive. 
Earth System Science Data, 17(6), 2373-2404. https://doi.org/10.5194/essd-17-2373-2025

Disturbance detection:
In EFDA v3.0, the disturbance detection algorithm has been updated to a deep learning-based approach. The disturbance detection methodology 
and validation are described in:
- Viana-Soto, A., Mense, J. A., Kowalski, K., Pauls, J., Gieseke, F. & Senf, C. (2026). Deep learning-based forest disturbance detection for 
Europe using Landsat time series. Remote Sensing of Environment, In press

============================================================