CONUS forest carbon projections from a calibrated Forest Vegetation Simulator (FVS) for PERSEUS

Continental (CONUS, 49 states) forest carbon projections to 2125 from a modernized, Bayesian-calibrated Forest Vegetation Simulator (FVS) initialized from the full USDA Forest Inventory and Analysis database (FIADB). Three growth/mortality engines (default native mortality, Bayesian-calibrated, and a national density-dependent gompit mortality) are run under four management scenarios (reserve/no-harvest; extensive partial harvest; a realistic managed scenario with intensive clearcut harvest confined to plantations; and an all-intensive bound), with harvest and disturbance regimes derived from the CONUS-HCS decision surfaces and sampled at FIA plot locations. Per-state live aboveground carbon and biomass are anchored to FIADB carbon. Bayesian posterior-draw parameter uncertainty and bootstrap sampling uncertainty are quantified; a FIADB-vs-TreeMap multi-scale area-expansion comparison and landowner/ecoregion/state trend breakdowns are included. These data drive the FVS engines on the PERSEUS Forest Intelligence dashboard. Includes per-state series (CSV), the merged dashboard series (JSON), management-scenario outputs, plantation and harvest-rate inputs, the full reproducible pipeline code, and findings documentation including a handoff. Derived from the fvs-modern software (DOI 10.5281/zenodo.19802672).本数据集基于经过现代化改造、贝叶斯校准的森林植被模拟器（Forest Vegetation Simulator, FVS）生成至2125年的美国大陆（Continental United States, CONUS，49个州）森林碳储量预测结果，模拟器初始化数据源为完整的美国农业部森林调查与分析数据库（USDA Forest Inventory and Analysis Database, FIADB）。研究设置三类生长/死亡率模型模块：默认原生死亡率模型、贝叶斯校准死亡率模型以及全国尺度密度依赖型冈珀茨死亡率模型，并在四类管理情景下运行这些模块，分别为：保护禁采情景、广泛择伐情景、仅在人工林开展高强度皆伐的现实经营情景、全高强度经营情景。采伐与干扰制度源自CONUS-HCS决策面，并在FIA样地位置进行采样。各州的地上活体碳储量与生物量均以FIADB碳储量数据为基准锚定。本数据集量化了贝叶斯后验抽样参数不确定性与自举（bootstrap）抽样不确定性，同时包含FIADB与TreeMap的多尺度面积扩展对比分析，以及土地所有者、生态区、州级尺度的趋势分解结果。这些数据为PERSEUS森林智能仪表盘的FVS运行引擎提供数据支撑。 本数据集包含各州时间序列数据（CSV格式）、仪表盘合并序列数据（JSON格式）、各管理情景输出结果、人工林与采伐率输入数据、完整可复现的分析流程代码，以及包含移交说明在内的研究成果文档。本数据集基于fvs-modern软件构建（DOI：10.5281/zenodo.19802672）。