PERSEUS libcbm vs GCBM cross-state intercomparison (CONUS)

48-state libcbm pool outputs under canonical B1.3 FIA EXPNS-anchored inventory stratification, plus 6-state libcbm vs GCBM density gap matrices (B1.1 v6 production parity and B1.3 FIA EXPNS), per-pool and per-DOM-pool decompositions, F3 Q10 mean annual temperature sensitivity sweeps for Georgia and Oregon, and the five publication figures supporting the PERSEUS multi-model forest carbon intercomparison methods finding.

Headline findings: The previously reported +24% libcbm-vs-GCBM engine gap is reproduced in 5 of 6 states with full GCBM aggregates (WA, MN, IN, ME, OR all between 0.74 and 0.80) under canonical B1.1 v6 parity, with Georgia the lone outlier. Replacing uniform-FT B1.1 inventory with B1.3 FIA EXPNS (the canonical FIA Total Area Estimator) adds +6,760 TgC to the CONUS-wide carbon stock at n=39 states (+14 percent shift). The +24% engine gap carries a regional DOM-pool fingerprint driven by cool-moist climate spinup conditions: Pacific Northwest mean slow-soil pool 184 Mg/ha, Atlantic Maritime Northeast 143 Mg/ha, Lake States and South + Southeast around 105 Mg/ha. An F3 Q10 MAT sensitivity sweep on Oregon (MAT 4 to 13 C) confirms the slow-soil overshoot is spinup-driven, not runtime Q10-driven.

Pipeline: All libcbm outputs were produced with the GCBM2hpc pipeline (github.com/holoros/GCBM2hpc) on the Ohio Supercomputer Center Cardinal cluster (allocation PUOM0008). 6-state GCBM aggregates were produced spatially via moja FLINT containerized GCBM at 1-degree WGS84 tiles. Auxiliary CONUS data: TreeMap 2022, LCMS v2024, FIA DataMart, NOAA 1991-2020 climate normals, EPA Level III ecoregions.

Companion methods note: holoros.github.io/perseus-forest-intelligence/methods/inventory-stratification/