Collaborative Research: MRA: Constraining the continental-scale terrestrial carbon cycle using NEON data

美国森林中的植物消耗大量的二氧化碳，在调节大气中这种重要气体的浓度方面发挥着重要作用。科学家们迫切需要减少未来大气二氧化碳浓度预测中的不确定性，并更自信地评估森林中的植物是否会继续抵消化石燃料燃烧向大气排放的二氧化碳。美国国家科学基金会资助的国家生态观测网（NEON）是一个大陆尺度的生态观测设施，收集并提供来自美国81个野外站点的质量控制数据，这些数据描述和量化了我们国家生态系统的变化情况。NEON对植物光合作用和呼吸作用的测量为计算机预测大陆尺度模式、年度变化和全球碳循环趋势提供了所需的数据。该项目将采用一种新颖的方法来改进复杂的计算机模型，并提供广泛寻求的NEON数据应用，从而开发更精细的大陆尺度碳剖面。研究人员将通过向早期职业科学家提供培训、扩大多样性、与研究界共享产品以及加强研究和教育基础设施来扩大这个项目的影响。研究结果预计将对全球变化的综合评估具有很高的价值，并将对联邦、州和地方机构以及正在制定管理美国自然资源决策的土地管理者有用。研究人员将通过整合NEON观测、卫星数据、数据驱动方法和数据模型集成技术来限制大陆尺度的陆地碳循环。目标是：(1)利用NEON数据开发大陆尺度通量产品，以帮助实现NSF的NEON目标；(2)评估NEON数据集的信息含量，降低模型的不确定性；(3)将多个NEON数据集同化成复杂的土地模型，量化美国土地碳汇及其不确定性；(4)了解大陆尺度碳动态及其调控机制。首先，研究人员将利用NEON数据开发基于卫星观测和气象数据的不确定性估计的每小时网格化初级生产总值（GPP）产品。其次，他们将使用NEON数据和卫星上的太阳诱导叶绿素荧光（SIF）数据来开发网格化的免费GPP产品，并进行不确定性估计。第三，他们将利用NEON的多个异构数据集和网格化GPP产品来量化美国陆地碳汇潜力，并评估NEON数据在约束参数估计和模型预测方面的有效性。第四，他们将开发一个多模型集合，以了解模型预测中基于通量和基于池的数据对结构不确定性的约束。第五，他们将应用数据同化技术来量化参数不确定性，并评估多个与单个NEON数据集对三种不同模型参数估计的约束。最后，他们将利用基于NEON数据的网格通量产品、NEON数据训练模型和可追溯性框架评估美国陆地碳汇、其不确定性和监管机制。该项目最终将通过不确定性分析为改进模型和NEON观测提供反馈。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Plant life in U.S. forests consume large volumes of carbon dioxide, playing an important role in regulating the concentrations of this important gas in the atmosphere. There is an urgent need for scientists to reduce uncertainties in future projections of atmospheric carbon dioxide concentrations and to more confidently assess whether plants in forests will continue to offset carbon dioxide emissions to the atmosphere from the burning of fossil fuels. The NSF-funded National Ecological Observatory Network (NEON) is a continental-scale ecological observation facility that collects and provides quality-controlled data from 81 field sites across the United States that characterize and quantify how our nation's ecosystems are changing. NEON's measurements of plant photosynthesis and respiration are providing the data needed for computer projections of continental scale patterns, year-to-year variations, and trends in the global carbon cycle. This project will develop more refined continental-scale carbon profiles by using a novel approach to improve complex computer models and delivering a widely-sought application of NEON data. The investigators will broaden impacts of this project by providing training to early-career scientists, broadening diversity, sharing products with the research community, and enhancing research and education infrastructure. The results are anticipated to be of high value for integrated assessments of global change and will be useful for federal, state and local agencies, and land managers who are making decisions on managing U.S. natural resources.The investigators will constrain the continental-scale terrestrial carbon cycle by integrating observations from NEON, satellite data, data-driven methods, and data-model integration techniques. The objectives are to: (1) use NEON data to develop continental-scale flux products that are need to help realize NSF's goals for NEON; (2) evaluate information content of NEON data sets for reducing model uncertainty; (3) assimilate multiple NEON data sets into complex land models to quantify the U.S. land carbon sink and its uncertainty; (4) understand the continental-scale carbon dynamics and underlying regulatory mechanisms. First, the investigators will utilize NEON data to develop an hourly, gridded Gross Primary Production(GPP) product with uncertainty estimates based on satellite observations and meteorological data. Second, they will use NEON data along with solar-induced chlorophyll fluorescence (SIF) data from satellites to develop gridded complimentary GPP products with uncertainty estimates. Third, they will utilize multiple heterogeneous datasets from NEON and the gridded GPP products for quantifying the US land carbon sink potential and assess the effectiveness of NEON data to constrain parameter estimation and model prediction. Fourth, they will develop a multiple model ensemble to understand constraints of flux- vs. pool-based data on structure uncertainty in model prediction. Fifth, they will apply data assimilation techniques to quantify parameter uncertainty and assess constraints of multiple vs. single NEON data sets on parameter estimation with three different models. Finally, they will assess the US land carbon sink, its uncertainty, and regulatory mechanisms with NEON data-based gridded flux products, NEON data-trained models, and the traceability framework. This project will ultimately provide feedback towards improvement of models and NEON observations via uncertainty analysis.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.