RII Track-2 FEC: Computational Methods and Autonomous Robotics Systems for Modeling and Predicting Harmful Cyanobacterial Blooms

世界上的淡水湖是人类饮用水、灌溉、降温、娱乐和食品生产的重要水源。然而，这些基本湖泊服务的提供受到了全世界湖泊中有害蓝藻水华发生率增加的威胁。有害的水华会降低湖泊的水质、透明度和美观性，对财产价值产生负面影响，并可能通过产生可能损害多个器官系统的强力毒素来威胁人类和动物的健康。该项目旨在通过将机器人和大数据技术与传统的水样相结合，揭示蓝藻水华在哪里、何时以及如何发展和传播的驱动因素。该项目将提高评估和预测蓝藻水华的能力，有可能使娱乐湖泊和供应饮用水的湖泊更早进行公共卫生干预。干预措施可以改善水的处理和分配。该项目的劳动力发展活动将培训下一代专业人员，通过使用大数据工具和技术，跨学科和社区工作和交流，以解决具有重大社会影响的复杂科学问题。该项目将大数据工具与机器人、传感器网络和湖泊采样相结合，开发了实时、自适应、自主的环境数据收集和处理策略，以增强对早期水华湖泊中有害蓝藻水华发展的预测能力。具体地说，配备了一套测量物理、化学和生物参数的传感器的自主水面飞行器和配备了高光谱、多光谱和可见光相机的无人驾驶飞行器将在蓝藻水华的开始和演替期间产生大量关于湖泊的数据。采集后处理和模型开发将近乎实时地检查对水华的发生和传播的控制。该项目汇集了一个跨学科的调查小组，他们来自四个EPSCoR司法管辖区：缅因州、新罕布夏州、罗德岛州和南卡罗来纳州，他们在大数据、环境科学、生态学、人类人口学、仪器和机器人方面具有专业知识。与参与机构、当地湖泊协会、市政自来水供应商和州政府机构合作，将产生影响所有四个州湖泊水质的物理、化学和生物因素的大规模数据集，然后将用于创建新的模型来预测有害的蓝藻水华。十几名早期职业科学家将接受跨学科研究方面的培训。社区合作伙伴将参与数据收集、数据解释以及监测和管理战略的实施。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。

The world's freshwater lakes are a crucial source of water for human use, for drinking, irrigation, cooling, recreation, and food production. However, provision of these essential lake services is threatened by the increased incidence of harmful cyanobacterial blooms in lakes worldwide. Harmful blooms decrease lake water quality, clarity, and aesthetics, negatively impact property values, and can threaten human and animal health through the production of potent toxins that can damage multiple organ systems. This project aims to unravel the drivers of where, when, and how cyanobacterial blooms develop and spread, by combining robotics and big data technologies with traditional water sampling. The project will advance the ability to evaluate and predict cyanobacterial blooms, potentially allowing earlier public health interventions in recreational lakes and in lakes that supply drinking water. Interventions can enable improved water treatment and distribution. The project's workforce development activities will train next generation professionals to work and communicate across disciplines and communities in order to address complex scientific problems that have major societal implications, though use of big data tools and technology. Using the tools of big data jointly with robotics, sensor networks, and limnological sampling, the project develops strategies for real-time, adaptive, autonomous environmental data collection and processing to enhance the ability to predict the development of harmful cyanobacterial blooms in lakes with incipient blooms. Specifically, autonomous surface vehicles equipped with a suite of sensors measuring physical, chemical, and biological parameters and unmanned aerial vehicles equipped with hyper-spectral, multispectral, and visible-light cameras will generate large volumes of data on lakes during the onset and succession of cyanobacterial blooms. Post-acquisition processing and model development will examine controls on the genesis and spread of blooms in near-real time. The project brings together an interdisciplinary group of investigators with expertise in big data, environmental science, ecology, human demography, instrumentation, and robotics from four EPSCoR jurisdictions: Maine, New Hampshire, Rhode Island, and South Carolina. Partnerships with the participating institutions, local lake associations, municipal water providers, and state agencies will produce large-scale datasets of physical, chemical, and biological factors influencing water quality from lakes in all four states, which will then be used to create new models to predict harmful cyanobacterial blooms. Over a dozen early career scientists will be trained in interdisciplinary research. Community partners will be engaged in data collection, data interpretation, and implementation of monitoring and management strategies.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.