New Brunswick transitioning and above late-successional forest: 10 m probability surface, bootstrap uncertainty and extrapolation flag, trained on the licensed provincial inventory (2024 AlphaEarth embeddings)

A 10 m wall-to-wall probability surface for the transitioning and above late-successional condition across New Brunswick forestland, with a matching block-bootstrap standard deviation layer and a training-envelope extrapolation flag. Predictions come from a random forest over the 64 Google AlphaEarth Foundations 2024 satellite embedding bands and a Hansen v1.12 time-since-disturbance covariate, fitted on the 9,042 plot Configuration D basis of the provincial Continuous Land Inventory as delivered under license, labeled with the five-axis New Brunswick score at four or more (2,854 positives). The mosaicked forestland share is 32.15 percent (bootstrap percentile interval 30.93 to 33.37 across 20 forests), inside the independent design-based estimate of 31.47 percent (95 percent interval 29.17 to 33.31) on the same licensed frame, with no calibration step. The surface passed an eleven-gate acceptance suite, a plot-level assembly check against a permutation null, and a separate post-acceptance stress test. Intended use. This is a continuous probability surface, not a classified map, and the area above a fixed probability cut is not an estimate of class area. On H3 resolution 4 hexagons a Fay-Herriot small area model with the surface's block mean as covariate returns a slope of 1.39 (SE 0.19), so block means are compressed toward the middle, and a single cut matched to the provincial rate over-selects by 2.7 points in the richest quarter of blocks and under-selects by 2.1 in the poorest. Use block estimates, or a block-specific cut, rather than a nominal probability. Companion records: the New Brunswick CORE4 structural core surface on the same licensed labels (10.5281/zenodo.22967451) and the Maine transitioning and above surface (10.5281/zenodo.22967249), which maps a different, six-axis instrument. The fitted random forest and its 20-model block-bootstrap ensemble are included so the surface can be regenerated and audited, with a memorization measurement deposited beside them. No plot-level inventory records, plot coordinates, per-plot embeddings, per-holder or per-parcel outputs are included in this deposit. Release of aggregate products from the licensed inventory was permitted by the New Brunswick partners in September 2026.