Supplementary disturbance figures for quantifying impacts of natural gas development on forests.

As energy demands continue to rise, energy production from sources including natural gas is expected to rapidly accelerate in the coming decades, potentially leading to substantial land-use changes. In the Appalachian region of the United States, natural gas development often occurs in forested areas, which can cause high levels of forest disturbance and loss. Here, we use nationwide forest inventory and remotely sensed data in a Bayesian model to quantify the impacts of natural gas development at fine spatial resolutions between 2008-2021. Based on well permit locations in the states of Ohio, Pennsylvania, and West Virginia, the analysis quantifies disturbance area, forest carbon loss, and opportunity cost with associated levels of uncertainty at the pixel-level. Overall, we estimate 10,854 ha of forest land were disturbed, resulting in 542,675 Mg ($\pm$ 4,275) of forest carbon loss. The opportunity cost associated with these disturbances is estimated to be 575,246 Mg ($\pm$ 30,774). Pixel-level estimates are generated for individual well sites, which can be aggregated to the county-level to highlight regional patterns. Specifically, we observe greater levels of disturbance in Northern West Virginia, while opportunity costs are greatest for large forested counties in Northeastern Pennsylvania. This study demonstrates the importance of quantifying balances and tradeoffs between energy production and forest ecosystem services, and provides important insights into the impacts of energy development on forest carbon.随着能源需求持续增长，未来数十年内，包括天然气在内的各类能源生产预计将快速扩张，可能引发大规模土地利用变化。美国阿巴拉契亚地区的天然气开发活动常发生于林区，极易造成严重的森林干扰与林地流失。本研究借助全国森林清查数据（nationwide forest inventory）与遥感数据（remotely sensed data），构建贝叶斯模型（Bayesian model），对2008至2021年间天然气开发活动的影响开展精细空间分辨率下的量化分析。研究基于俄亥俄州、宾夕法尼亚州与西弗吉尼亚州的钻井许可点位（well permit locations），在像元级（pixel-level）量化了干扰面积、森林碳流失量以及机会成本（opportunity cost），并给出对应的不确定性水平。整体而言，研究估算共有10854 ha林地受到干扰，导致森林碳流失量达542675 Mg（±4275 Mg）。此次森林干扰对应的机会成本估算值为575246 Mg（±30774 Mg）。研究生成了单个钻井点位的像元级估算结果，可通过聚合至县级尺度以凸显区域分布特征。具体而言，西弗吉尼亚州北部的森林干扰程度更高，而宾夕法尼亚州东北部大型林区的机会成本则最为显著。本研究证实了量化能源生产与森林生态系统服务之间平衡与权衡关系的重要性，并为理解能源开发对森林碳循环的影响提供了关键见解。