CAREER: CAS-Climate: Multiscale Data and Model Synthesis Informed Approach for Assessing Climate Resilience of Crop Production Systems

今天的生产者面临着一系列不断扩大的挑战，水资源的可持续性是主要问题之一。适应这些问题，特别是在气候变化和日益极端的天气条件下，需要改变耕作做法。人们正在以不同的时间和空间分辨率收集大量的相关数据。这些数据涵盖土壤的性质和条件、气候数据、作物管理数据、作物健康、不同类型的压力和压力源，以及水的可获得性和消耗量。现在，越来越多的食品生产管理决策被委托给机器学习模型和伴随的传感器网络，这些网络提供和生成各种规模的不同数据。但是，这些模型和方法本身并不能全面解决气候适应性适应和作物生产系统可持续集约化所需的广泛的规模、环境变量和地方/区域差异。有必要整合科学和工程专业知识，评估一系列作物管理方案，并制定复原力指标，以延长不可再生和有限水资源的生存能力。该项目将建立研究能力，以开发和完善跨尺度的建模能力，范围从特定的点到区域，从一天到一个世纪。这些信息对于指导适应战略并评估其在气候变化背景下对粮食生产和水可持续性的影响至关重要。项目组将通过将实地和遥感数据与新的建模框架联系起来应对这一挑战，新的建模框架能够在不同的未来气候情景下生成多种作物生产情景，以确保为水资源的可持续性和复原力制定最佳战略。该项目将使用田间试验、新颖的分析以及人、农场和自然系统之间的联系来帮助改变大田作物的种植方式。其目标是创建一个能够更好地维持水资源并管理养分和土壤的一体化系统。该项目的方法基于一个战略性的五年计划，以实现PI的总体职业目标，即通过对粮食生产系统进行系统调查，将她的研究和教学与环境问题相结合，研究时空尺度和物理条件之间的联系，这些条件阻碍了对气候智能水管理实践的理解和有效应用于作物生产。这项工作将需要融合多尺度、异质、多源、时变的数据，包括来自地下传感器的数据、地面数据、天气预报、作物生长和土壤养分等；对气候-水-作物生产循环的了解；以及复原力衡量标准。将通过以下综合目标执行该战略：(1)对作物生产系统对作物生长和水文的时空变化反应进行机器学习的多尺度模拟；(2)调查气候(变化和极端事件)和作物管理情景(灌溉、养分使用、作物选择、土地过渡)及其对粮食生产的影响；(3)量化作物生产系统可持续性的复原力指标，以指导在未来气候下确定管理措施的优先次序。该项目的教育目标是通过与现有的本科和研究生教学和学习课程相结合，让学生参与并装备受农业生态系统启发的基础培训，从而加强他们加入与STEM相关的工作队伍的准备，这些工作涉及数据科学、自然资源管理以及环境决策支持和咨询。为该项目设计的研究活动将吸引早期职业教师和学生在他们的职业生涯中前进，并指导学生在不同的教育阶段(研究生、本科、高中和中学)，并通过实地考察和教育推广活动与农村社区接触。该项目由CBET/ENG环境可持续发展计划和既定的激励竞争研究计划(EPSCoR)共同资助。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。

Today’s producers encounter a continuously expanding array of challenges with the sustainability of water resources being one of the major issues. Adapting to these issues, especially under a changing climate and increasingly extreme weather conditions, necessitates a shift in farming practices. A large amount of related data is being collected at different resolutions in time and space. These data range across properties and condition of soils, climate data, crop management data, crop health, different types of stresses and stressors, and water availability and consumption. More and more food production management decisions are now delegated to machine learning models and accompanying sensor networks that provide and generate diverse data across various scales. But these models and methods alone fall short in comprehensively addressing the wide range of scales, environmental variables, and local/regional variations necessary for climate-resilient adaptation and sustainable intensification of crop production systems. There is a need to integrate scientific and engineering expertise, assess a range of crop management scenarios, and develop resilience metrics to prolong the viability of non-renewable and finite water resources. This project will build research capacity for developing and refining modeling capabilities across scales, ranging from specific points to regions and from one day to a century. This information is crucial to steer adaptation strategies and assess their effects on both food production and water sustainability in the context of climate change. The project team will address this challenge by linking on- ground and remotely sensed data with a new modeling framework that is capable of generating multiple scenarios for crop production under different future climate scenarios to ensure the best set of strategies for sustainability and resilience of water resources. The project will use field trials, novel analytics, and links between people, farms, and natural systems to help change how field crops are grown for the better. The goal is to create an all-in-one system that can better sustain water resources and manage nutrients and soils. The project approach is based on a strategic 5-year plan for achieving the PI’s overall career goal of integrating her research and teaching through systematic investigations of food production systems with environmental concerns by studying the connections between spatial-temporal scales and physical conditions that have impeded understanding and effective application of climate smart water management practices for crop production. This effort will require a fusion of multiscale, heterogeneous, multi-sourced, time-varying data including data from sub-surface sensors, surface data, weather forecasts, crop growth, and soil nutrients, etc.; understanding of the climate-water-crop production loop; and resilience metrics. The strategy will be pursued through the following integrated objectives (1) conduct machine learning-informed multiscale modeling of crop production systems’ spatio-temporally varying responses of crop growth and hydrology; (2) investigate climate (change and extreme events) and crop management scenarios (irrigation, nutrient use, crop choice, land transition) and their impact on food production; and (3) quantify resilience metrics for the sustainability of crop production systems to guide prioritization of management measures under future climate. The education goal of this project is to engage and equip students with agroecosystem-inspired fundamental training through integration with the existing curriculum of undergraduate and graduate teaching and learning, thus strengthening their readiness to join a STEM- related workforce in data science, natural resource management, and environmental decision support and consulting. The research activities designed for the project will engage an early career faculty member and students in advancing through their careers and guiding students at different stages of education (graduate, undergraduate, and high and middle school) and engaging with rural communities through field days and educational outreach activities.This project is jointly funded by the CBET/ENG Environmental sustainability program and the Established Program to Stimulate Competitive Research (EPSCoR).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.