Quantifying the carbon balance of forested ecosystems has been the subject of intense study involving the development of numerous methodological approaches. Forest inventories, processes-based biogeochemical models, and inversion methods have all been used to estimate the contribution of U.S. forests to the global terrestrial carbon sink. However, estimates have ranged widely, largely based on the approach used, and no single system is appropriate for operational carbon quantification and forecasting. We present estimates obtained using a new spatially explicit modeling framework utilizing a “gain–loss” approach, by linking the LUCAS model of land-use and land-cover change with the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3). We estimated forest ecosystems in the conterminous United States stored 52.0 Pg C across all pools. Between 2001 and 2020, carbon storage increased by 2.4 Pg C at an annualized rate of 126 Tg C year−1. Our results broadly agree with other studies using a variety of other methods to estimate the forest carbon sink. Climate variability and change was the primary driver of annual variability in the size of the net carbon sink, while land-use and land-cover change and disturbance were the primary drivers of the magnitude, reducing annual sink strength by 39%. Projections of carbon change under climate scenarios for the western U.S. find diverging estimates of carbon balance depending on the scenario. Under a moderate emissions scenario we estimated a 38% increase in the net sink of carbon, while under a high emissions scenario we estimated a reversal from a net sink to net source. The new approach provides a fully coupled modeling framework capable of producing spatially explicit estimates of carbon stocks and fluxes under a range of historical and/or future socioeconomic, climate, and land management futures.

中文翻译：

---

美国森林碳动态的业务评估工具：连接 LUCAS 和 CBM-CFS3 模型的新空间显式方法

量化森林生态系统的碳平衡一直是深入研究的主题，涉及开发多种方法。森林清单、基于过程的生物地球化学模型和反演方法都已用于估计美国森林对全球陆地碳汇的贡献。然而，估算的范围很广，主要基于所使用的方法，并且没有一个系统适用于业务碳量化和预测。我们通过将土地利用和土地覆盖变化的 LUCAS 模型与加拿大森林部门的碳预算模型 (CBM-CFS3) 联系起来，展示了使用新的空间显式建模框架获得的估计值，该框架利用“收益-损失”方法。我们估计美国本土的森林生态系统在所有水池中储存了 52.0 Pg C。从 2001 年到 2020 年，碳储存量增加了 2.4 Pg C，年化率为 126 Tg C year-1。我们的结果与使用各种其他方法估算森林碳汇的其他研究大体一致。气候变率和变化是净碳汇规模年度变率的主要驱动因素，而土地利用和土地覆盖变化和干扰是该幅度的主要驱动因素，使年汇强度降低了 39%。美国西部气候情景下的碳变化预测发现不同情景下对碳平衡的估计存在差异。在中等排放情景下，我们估计碳的净汇增加了 38%，而在高排放情景下，我们估计从净汇逆转为净源。