Nine experimental annual maps. Two bounded mountain studies. Every footprint is a question for investigation.
What the colors mean
ForTy v2 beta assigns one of seven native classes to each 10 m cell. Mature and regrowing primary forest remain separate. “Other land” can include temporarily unstocked forest. The annual series is experimental; no changes are inferred by comparing it with the separate v1 map.
What makes a candidate persistent?
A beta 2020 natural-forest cell (classes 1, 2 or 7) must enter the same destination class (3, 4, 5 or 6) in two adjacent years during 2021–2025. Baseline and both years need scores ≥0.7. An event remains historical after apparent recovery. A 2025 departure without such a pair is provisional.
Why the totals differ
The list retains polygon components ≥0.5 ha; smaller screened pixels remain included in the pixel totals.
Independent evidence has different support
Candidate fractions are averaged to an aligned 30 m equal-area grid and compared with Hansen GFC 2025 v1.13 canopy loss in 2021–2025. Per-patch percentages use that comparison support. This does not validate 10 m boundaries, timing, cause, permanent conversion, or legal compliance. The comparison layer always shows 2021–2025, whatever ForTy year is selected.
A map is not ground truth
Model scores are uncalibrated, rather than probabilities. No field verification, species inference, carbon estimate or legal determination is provided. Each rectangle is a defined study extent covering the main massif and surrounding slopes; it is not an official mountain or protected-area boundary. Terrain is real, at 1× vertical scale.
Geographic names
Peaks, passes, ridges, valleys, lakes, rivers and settlements come from a cached GeoNames Romania snapshot. Labels become more detailed as you zoom in. Available peak elevations are source values; names and positions can be incomplete or approximate. These labels provide geographic context and do not represent historical land cover or hiking guidance.
Sources & attribution