Peat-MOSS (Peatland Mapping and Observation with Satellite Sensors)

泥炭地是最安全的温室气体汇之一，储存了全球土壤碳的25%，但它们的退化占所有人为排放的2%-5%。认识到储存和排放的巨大潜力，旨在减少碳排放和加强泥炭地碳储存的恢复项目呈指数级增加。然而，目前还没有准确和可扩展的方法来测量、报告和验证泥炭地恢复项目(MRV)。目前，泥炭地恢复对温室气体减排(GGR)的影响只是从泥炭地分类的变化(例如，从“排干”到“恢复”)来假定的。此外，由于对全球泥炭范围的了解有限，为进一步部署GGR项目寻找合适的地点受到阻碍。例如，在热带地区，未测绘的泥炭地总面积可能比目前测绘的面积大三倍，体积大四倍[Q1ref-2]。基于卫星对地观测(EO)数据的泥炭地数字测绘和监测具有扩大规模的潜力，以满足全球对准确的MRV和GGR项目测绘的需求，但在三个关键领域受到限制：*对基本统计技术的严重依赖。*缺乏对不确定性的量化。*在分辨率和数据输入方面缺乏一致性。这些限制限制了监测泥炭地项目的有效性和为未来的泥炭地GGR项目确定/优先考虑地点的能力。Sylvera将通过将其在机器学习(ML)和EO数据方面的专业知识与英国生态和水文中心(UKCEH)的研究相结合，克服数字泥炭地测绘和监测的局限性。英国生态和水文中心是泥炭地GGR示范的领先者。Sylvera将领导下一代数字泥炭地测量，提供可扩展的MRV解决方案，这是为GGR项目带来完整性和信心所急需的。作为回应，Peat-Moss将开发一种ML驱动的方法，将其基于EO数据的行业领先的深度学习模型与UKCEH的广泛研究相结合。泥炭-苔藓将绘制泥炭范围的地图和泥炭退化的额外指标，这些指标可以用来确定泥炭地是排放还是封存二氧化碳。泥炭-苔藓地图的使用案例是1)泥炭GGR项目的MRV和2)泥炭GGR项目地点的识别和优先顺序。Sylvera将利用EO数据和ML技术提供数字GGR MRV。这一MRV解决方案将通过提供一致且可扩展的泥炭地监控，确保GGR项目的有效性、安全性和持久性。

Peatlands are among the most secure sinks for greenhouse gases, storing 25% of global soil carbon, yet their degradation is responsible for 2-5% of all anthropogenic emissions\[Q1ref-1\]. Understanding the large potential for both storage and emissions, there has been an exponential increase in restoration projects aimed at reducing emissions and enhancing carbon storage in peatlands. However, there is currently no accurate and scalable method for measurement, reporting, and verification (MRV) of peatland restoration projects. At present, the impact of peatland restoration for greenhouse gas reduction (GGR) is simply assumed from a change in peatland classification (e.g., from "drained" to "restored"). Additionally, finding proper sites for further deployment of GGR projects is hindered by the limited knowledge of global peat extent. For example, in tropical regions the total un-mapped peatland area may be three times larger, and four times more voluminous than currently mapped\[Q1ref-2\]. Digital mapping and monitoring of peatlands based on satellite earth observation (EO) data has the potential to scale to meet the global need for accurate MRV and mapping of GGR projects, yet has been limited in three key areas:* Heavy reliance on basic statistical techniques.* Lacking quantification of uncertainty.* Lacking consistency in resolution and data inputs.These limitations have restricted the ability to monitor the effectiveness of peatland projects and identify/prioritise locations for future GGR projects.Sylvera will conquer the limitations of digital peatland mapping and monitoring by combining its expertise in machine learning (ML) and EO data with the research of the UK Centre for Ecology and Hydrology (UKCEH) - leaders in peatland GGR demonstration. Sylvera will lead the next-generation of digital peatland measurements to provide the scalable MRV solution that is desperately needed to bring integrity and confidence to GGR projects. Peat-MOSS will respond by developing an ML-driven method, combining its industry-leading deep-learning models based on EO data with the extensive research from UKCEH. Peat-MOSS will develop maps of peat extent and additional indicators of peat degradation that can be used to determine whether peatlands are emitting or sequestering CO2\. The use cases of Peat-MOSS maps are 1) MRV of peat GGR projects and 2) identification and prioritisation of peatland GGR project sites. Sylvera will leverage EO data and ML technology to provide digital GGR MRV. This MRV solution will ensure the effectiveness, security, and permanence of GGR projects by providing consistent and scalable monitoring of peatlands.