GLAD Deforestation Alerts

Biodiversityv02.00 credits/per_km2

Responsible:Niccolò Calandri

Near-real-time deforestation alerts from the UMD GLAD system (Global Land Analysis and Discovery), distributed through Resource Watch (WRI). The layer flags recent forest-disturbance events detected on Landsat imagery, classified by confidence level (1 = low, 2 = medium, 3 = high) at 30 m resolution, so users can spot potential canopy loss shortly after it occurs.

Data Sources & References

Sources

WRD_GLAD_01

Methodology

The layer reads the UMD GLAD near-real-time deforestation alerts from the Google Earth Engine ImageCollection projects/resource-watch-gee/for_003_nrt_rw1_glad_deforestation_alerts. GLAD (Global Land Analysis and Discovery, University of Maryland) detects forest-disturbance events from Landsat time series and publishes them as alerts distributed through Resource Watch (WRI).

Processing

The most recent image in the collection is selected by sorting on system:time_start in descending order and taking the first element.
Band b1 is selected and clipped to the polygon of interest.
Pixel values encode the alert confidence level, rendered on a graduated palette normalised over the range 0-3 (1 = low, 2 = medium, 3 = high confidence).
The image is reprojected to Web Mercator at a native scale of 30 m before download.

Metadata reports the alert date, the maximum alert value found in the area, and the mean confidence over the polygon.

Data Source

Code	Provider	Resolution	Year
WRD_GLAD_01	UMD GLAD via Resource Watch (WRI), Landsat-based	30 m	2024 (near-real-time)