Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The authors evaluated whether geospatial foundation model (GFM) embeddings (AlphaEarth, TESSERA v2) outperform standard Sentinel-1/2 features for old-growth forest mapping across 211,893 ha in Romania's Southern Carpathians using expert-annotated parcels. Rigorous evaluation is critical because standard validation schemes often conflate true geographic transferability with spatial autocorrelation artifacts. While TESSERA initially outperformed Sentinel-1/2 by 0.08 PR-AUC under unbuffered spatial cross-validation, enforcing a 10 km buffer between train and test sets collapsed this performance gap to a statistically insignificant 0.04, with pixel-based XGBoost matching 10 m CNN architectures.
Geospatial foundation models seem to handily beat standard satellite features on remote sensing benchmarks—until you enforce a 10 km spatial buffer, which virtually erases their statistical advantage.
Old-growth forests develop over centuries under minimal anthropogenic disturbance, producing structurally complex and biodiverse stands. In Europe, protecting them requires mapping that is accurate for individual forest parcels yet deployable continent-wide. Geospatial foundation model (GFM) embeddings enable label-scarce land classification, but their value for old-growth detection remains unknown. Here, we map old-growth forests across 211,893 ha of Romania's Southern Carpathians, a beech-spruce landscape typical of the Alpine Biogeographic Region. We construct high-confidence, expert-informed reference labels for old-growth and non-old-growth parcels. We add AlphaEarth, TESSERA v2 and Sentinel-1/2 features to a common baseline of topographic and human-access predictors, then compare them under spatially blocked validation with and without 10 km train-test buffers to limit residual autocorrelation. With buffering, GFM and Sentinel-1/2 predictors increase precision-recall AUC by 0.21-0.25 [95% CIs: 0.15-0.34] relative to baseline, indicating spectral data contain a spatially robust old-growth signal. With a PR-AUC of 0.84 [0.79-0.88], TESSERA outperforms Sentinel-1/2 (+0.08 [+0.05 to +0.11]) and AlphaEarth (+0.08 [+0.04 to +0.12]) under unbuffered spatial validation. At a 10 km buffer, however, this advantage narrows to +0.04 [-0.01 to +0.11] and +0.03 [-0.04 to +0.10], intervals consistent with no difference. At 10 m resolution, convolutional neural networks add no benefit over pixel-based XGBoost. Comparisons with four national- and continental-scale products show the importance of non-old-growth labels, and reveal 81% agreement between our predictions and a field-calibrated map. We conclude that buffered spatial validation is vital when transferring old-growth detection models to unseen landscapes, and provide our labels and predictions for future work.