NSF Convergence Accelerator Track J: Building a digital twin for national-scale field-level crop monitoring, prediction, and decision support

NSF 融合加速器轨道 J：为国家规模的田间作物监测、预测和决策支持构建数字孪生

• 批准号: 2236137
• 负责人: Liping Di
• 金额: $ 75万
• 依托单位: George Mason University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-12-15 至 2023-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2236137&HistoricalAwards=false
• 关键词: NSF; Convergence; Accelerator; Track; Building

Crop production in the U.S. feeds not only the U.S. but also the world. During the 2020/2021 fiscal year, U.S. exports accounted for over 25% of total grain traded globally. A healthy crop cropping systems (CCS) is vital for achieving food and nutrient security of the U.S. and the world and enhancing the competitiveness of U.S. agriculture in the world market. Yet, crop production creates large environmental footprint. The USDA Agricultural Innovation Agenda calls for increasing U.S. agricultural production by 40% while cutting its environmental footprint in half by 2050. Sound crop management decision-making is a key in reaching this ambitious goal. Traditionally, crop management decisions are made by individuals based on their empirical judgment, which is often subjective and far from optimal. On the other hand, the science-based, data-driven approach for crop management decision-making relies on timely and accurate information on current and predicted future conditions of crop, soil, weather, and market to make optimal decisions. Studies demonstrated that the data-driven approach can overcome the inherent deficiencies in the empirical approach and bring significant economic and environmental benefits. However, it remains a challenge for stakeholders to utilize the data-driven approach because they don’t have full and effective access to the timely and accurate information and lack facilities or knowledge to process the information. This project will provide such timely information and decision support to stakeholders for enabling the data-driven optimal decision-making nationwide at field scales by developing the CropSmart Digital Twin (CSDT). CSDT will not only accurately represents the current conditions, but also predict, with acceptable confidence levels, future conditions of CCS with hypothetical “what if” scenarios, resulting in actionable predictions. The project will provide significant help in reaching the USDA Innovation goal and greatly enhance food and nutrition security of the U.S. and the world. Crop production is the foundation for food and nutrition security in the U.S. and the world. However, it also creates large environmental footprint. The USDA Agricultural Innovation Agenda calls for increasing U.S. agricultural production by 40% while cutting its environmental footprint in half by 2050. The data-driven approach for crop management decision-making, which relies on timely and accurate information on current and predicted future conditions of crop, soil, weather, and market to make optimal management decisions, has demonstrated its great potential to help USDA reach its ambitious goal. However, it remains a challenge for stakeholders to adopt the approach because they don’t have effective access to the decision-ready information (DRI) and lack facilities or knowledge to process the information. This project proposes to build the CropSmart Digital Twin (CSDT) with innovative Earth system DT technologies to facilitate the data-driven approach. The overarching goal is to ensure food and nutrition security by enhancing crop productivity and reducing environmental footprint in the U.S. through wide adoption of the data-driven approach enabled by CSDT. The major project activities include: (1) understanding stakeholders’ requirements on DRI and decision support; (2) identifying existing data, technologies, and gaps for CSDT; (3) quickly prototyping CSDT by integrating existing technologies and developing gap-filling technologies; 4) broadening participation and impact by training agricultural workforce through comprehensive extension; and (5) establishing a community-based CSDT network for long-term sustainability. This project explores the convergent approach for quickly constructing an operational DT through integration of multi-disciplinary components and services with interoperability technology. It demonstrates the advantage of multi-disciplinary collaboration and feasibility, usability, and value of DT as a multi-disciplinary integration platform for enabling the data-driven approach. The project will help USDA reach its Innovation Agenda goal and enhance food and nutrition security of the U.S. and the world.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.

美国的农作物生产不仅养活了美国，也养活了全世界。在2020/2021财年，美国出口占全球粮食贸易总量的25%以上。健康的作物种植系统（CCS）对于实现美国和世界的粮食和营养安全以及提高美国农业在世界市场上的竞争力至关重要。然而，农作物生产造成了巨大的环境足迹。美国农业部农业创新议程要求到2050年将美国农业产量提高40%，同时将其环境足迹减少一半。健全的作物管理决策是实现这一宏伟目标的关键。传统上，作物管理决策是由个人根据他们的经验判断做出的，这往往是主观的，远非最佳的。另一方面，以科学为基础、以数据为驱动的作物管理决策方法依赖于有关作物、土壤、天气和市场当前和预测未来状况的及时准确信息，以做出最佳决策。研究表明，数据驱动方法可以克服实证方法的固有缺陷，带来显著的经济效益和环境效益。然而，对于利益相关者来说，利用数据驱动的方法仍然是一个挑战，因为他们没有充分有效地获得及时准确的信息，也缺乏处理信息的设施或知识。该项目将通过开发CropSmart数字孪生（CSDT），为利益相关者提供及时的信息和决策支持，从而在全国范围内实现数据驱动的最佳田间决策。CSDT不仅能准确地反映当前状况，还能以可接受的置信度，通过假设的“如果”情景预测CCS的未来状况，从而得出可操作的预测。该项目将为实现美国农业部的创新目标提供重大帮助，并大大加强美国和世界的粮食和营养安全。农作物生产是美国和世界粮食和营养安全的基础。然而，它也造成了巨大的环境足迹。美国农业部农业创新议程要求到2050年将美国农业产量提高40%，同时将其环境足迹减少一半。数据驱动的作物管理决策方法依赖于当前和预测的作物、土壤、天气和市场状况的及时准确信息，以做出最佳的管理决策，已经证明了其帮助美国农业部实现其雄心勃勃的目标的巨大潜力。然而，对于利益相关者来说，采用这种方法仍然是一个挑战，因为他们无法有效地获取决策就绪信息（DRI），并且缺乏处理信息的设施或知识。该项目建议利用创新的地球系统DT技术构建CropSmart数字孪生（CSDT），以促进数据驱动的方法。总体目标是通过广泛采用CSDT支持的数据驱动方法，通过提高作物生产力和减少美国的环境足迹，确保粮食和营养安全。主要的项目活动包括：(1)了解利益相关者对DRI和决策支持的需求；(2)确定CSDT的现有数据、技术和差距；(3)整合现有技术，开发空白填补技术，快速实现CSDT原型化；4)通过全面推广，培训农业劳动力，扩大参与和影响；(5)建立以社区为基础的可持续发展CSDT网络。该项目探索了通过集成多学科组件和服务与互操作性技术来快速构建可操作的DT的融合方法。它展示了多学科协作的优势，以及DT作为实现数据驱动方法的多学科集成平台的可行性、可用性和价值。该项目将帮助美国农业部实现其创新议程目标，并加强美国和世界的粮食和营养安全。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。

项目成果

Automated In-Season Crop-Type Data Layer Mapping Without Ground Truth for the Conterminous United States Based on Multisource Satellite Imagery

• DOI: 10.1109/tgrs.2024.3361895
• 发表时间: 2024
• 期刊: IEEE Transactions on Geoscience and Remote Sensing
• 影响因子: 8.2
• 作者: Hui Li;Liping Di;Chen Zhang;Li Lin;Liying Guo;E. Yu;Zhengwei Yang

Cyberinformatics tool for in-season crop-specific land cover monitoring: Design, implementation, and applications of iCrop

• DOI: 10.1016/j.compag.2023.108199
• 发表时间: 2023-10
• 期刊: Comput. Electron. Agric.
• 作者: Chen Zhang;Liping Di;Li Lin;Haoteng Zhao;Hui Li;Anna Yang;Liying Guo;Zhengwei Yang

Hyperspectral Image Classification of Agricultural Tillage Practices Using Spatial-aware Collaborative Representation

• DOI: 10.1109/agro-geoinformatics59224.2023.10233552
• 发表时间: 2023-07
• 期刊: 2023 11th International Conference on Agro-Geoinformatics (Agro-Geoinformatics)
• 作者: Chiranjibi Shah;Q. Du