MRA: Macroscale Resilience: Assessing the recovery of western U.S. forests following compound disturbance by linking observations from trees to ecoregions

据估计，目前美国西部森林中有63亿棵死树，这是在地区变暖的背景下，甲虫侵袭、野火和干旱的频率或严重性不断增加造成的后果。一个关键的问题仍然存在：森林在过去36年里从这些干扰中恢复了吗？利用南落基山脉、北落基山脉和太平洋西北部的三个核心森林站点作为国家生态观测网(NEON)的一部分，这项工作将探索：1)什么干扰组合和大小会造成突然和持续的森林变化？2)森林弹性如何随森林类型、干旱和区域变暖而变化？这项工作将通过整合从个别树木到整个生态区的数据来探索这些问题，以促进对西部森林恢复的理解。一个新的核心科学社区--霓虹灯复原力网络将能够使用一个包含数据、代码、方法和三个软件包的开源霓虹灯工具包，对复原力进行数据密集型探索。此外，将与地区和联邦土地管理者制定如何抵制、适应或促进生态转型的最佳实践，以实现持续的接触点。这一联系提供了一条关键的联合生产途径，直接将霓虹灯启动的科学发现和公共土地管理决策联系起来。该项目将从理论上加深对复原力如何随着复合的大干扰、生态区域内和生态区域之间的变化以及与十年尺度区域变暖的函数而变化的理解。这些都是美国西部森林需要解决的基本大系统生态问题。这项研究将应用深度学习技术从霓虹灯空中收集的高光谱、激光探测和测距(LiDAR)和红绿蓝图像中识别单个植物物种，有助于推进研究前沿。结合通过无人驾驶航空系统(UAS)进行的扩展采样，这项研究将开发植被类别光谱特征，从Landsat卫星记录(1984-现在)中提取每年可分辨的植物功能类型图(针叶林、落叶林、木本灌木、草/草本植物和裸地)，提供前所未有的美国西部森林动态的时间重建。霓虹灯复原力网络还将在美国大陆范围内扩展核心复原力问题，询问：是否存在可预测的干扰组合、序列或大小，从而导致不同植被类型之间更持久的状态变化？这项研究将帮助释放陆地卫星记录的力量，探索美国西部的森林复原力，在霓虹灯开始监测未来30年的大陆规模的森林生态时，设定一个重要的历史基线。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

There are an estimated 6.3 billion dead trees currently across western U.S. forests, a legacy from increasing frequency or severity of beetle infestations, wildfires, and droughts against a backdrop of regional warming. A critical question remains: have forests recovered from these disturbances over the last 36 years? Leveraging three core forest sites in the Southern Rockies, Northern Rockies, and Pacific Northwest that are part of the National Ecological Observatory Network (NEON), this work will explore: 1) What combinations and sizes of disturbances create abrupt and persistent forest changes?, and 2) How does forest resilience vary as a function of forest type, drought, and regional warming? This effort will explore these questions by integrating data from individual trees to entire ecoregions to advance understanding of western forest recovery. A novel, core scientific community, the NEON Resilience Network, will be enabled for data-intensive exploration of resilience using an open-source NEON toolkit that contains data, code, methods, and three software packages. In addition, sustained touchpoints will be made with regional and federal land managers developing best-practices for how to resist, adapt, or facilitate ecological transformation. This connection offers a critical co-production pathway, directly connecting NEON-enabled scientific discovery and public lands management decisions.This project will advance theoretical understanding of how resilience varies as a function of compound, large disturbances, within and among ecoregions, and with decadal-scale regional warming. These are essential macrosystems ecological questions to address for western U.S. forests. This research will apply deep-learning techniques to identify individual plant species from NEON’s airborne data collections of hyperspectral, LiDAR (light detection and ranging), and red-green-blue imagery, helping to advance a research frontier. Coupled with extended sampling via unmanned aerial systems (UAS), the research will develop vegetation class spectral signatures to derive annually-resolved plant functional type maps (coniferous forest, deciduous forest, woody shrub, grass/herb, and bare ground) from the Landsat satellite record (1984-present), providing an unprecedented temporal reconstruction of western U.S. forest dynamics. The NEON Resilience Network will also scale core resilience questions across the continental U.S., asking: are there predictable combinations, sequences, or sizes of disturbances that lead to more persistent state changes across vegetation types? This research will help unlock the power of the Landsat record to explore forest resilience in the western U.S., setting an important historical baseline as NEON embarks on monitoring continental-scale forest ecology over the next 30 years.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.