Collaborative Research: CIBR: Cyberinfrastructure Enabling End-to-End Workflows for Aquatic Ecosystem Forecasting

由于人类活动，美国和全球的水生生态系统正在经历越来越多的变化。在环境条件迅速变化的情况下提供饮用水，促使人们需要发展对未来水质的预测。近期水质预报可以在一天到一周的时间尺度上指导管理行动，以减轻饮用水和其他基本淡水生态系统服务的潜在中断。为了最大限度地利用管理者和决策者的水质预测，预测必须是近乎实时的，可靠的，并不断更新的环境传感器数据。然而，开发迭代的短期生态预测需要广泛分布的复杂网络基础设施，从在淡水湖和水库收集信息的传感器和计算机到执行预测模型的云计算服务。因此，对于环境科学家来说，要想轻松有效地部署预测工作流程，仍然面临着重大的软件挑战。该项目将通过设计、实施和部署开源软件——FLARE：预测湖泊和水库生态系统——来满足这一需求，该软件将能够创建灵活、可扩展、健壮和近实时的迭代生态预测。该软件将进行测试，并广泛分发给供水公司、饮用水管理人员和许多其他决策者。FLARE将极大地提高生态研究界进行近实时水生预报的能力。FLARE预测系统在其架构上是新颖的，因为它集成了一个软件定义的虚拟分布式基础设施，涵盖了从网络边缘的传感器网关设备到云计算和存储的资源。FLARE将支持软件在湖泊和水库水质传感器附近的灵活部署，以及在云资源中进行端到端数据采集和处理。FLARE通过虚拟专用网络将其分布式资源互连起来，以确保通信中的数据完整性和隐私性，并支持适用于各种湖泊和水库的灵活模型。FLARE采用了最好的技术，它建立并集成了几个当代的、广泛使用的开源软件框架，从而降低了生态学家部署和管理生态预测工作流程的障碍。重要的是，该项目开发的可扩展和开源网络基础设施工具和端到端工作流程用于创建迭代水生预测，将为推进生态预测研究界提供关键资源，并为其他生态系统的预测提供模板。该项目将建立并扩展现有的跨学科教学工具和本科生和研究生研究交流计划，以提供计算机科学，淡水科学和生态系统建模交叉的培训。最终，该项目将开发可扩展的、健壮的、安全的工作流程，这将提高全球生态预测的能力、实践和培训机会。该项目的结果可以在http://flare-forecast.orgThis上找到，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。

Aquatic ecosystems in the United States and around the globe are experiencing increasing variability due to human activities. Provisioning drinking water in the face of rapid change in environmental conditions motivates the need to develop forecasts of future water quality. Near-term water quality forecasts can guide management actions over day to week time scales to mitigate potential disruptions in drinking water and other essential freshwater ecosystem services. To maximize the utility of water quality forecasts for managers and decision-makers, the forecasts must be accessible in near-real time, reliable, and continuously updated with environmental sensor data. However, developing iterative, near-term ecological forecasts requires complex cyber-infrastructure that is widely distributed, from sensors and computers collecting information at freshwater lakes and reservoirs to cloud computing services where forecast models are executed. Consequently, significant software challenges still remain for environmental scientists to easily and effectively deploy forecasting workflows. This project will address this need by designing, implementing, and deploying open-source software — FLARE: Forecasting Lake And Reservoir Ecosystems — that will enable the creation of flexible, scalable, robust, and near-real time iterative ecological forecasts. This software will be tested and widely disseminated to water utilities, drinking water managers, and many other decision-makers. FLARE will greatly advance the capability of the ecological research community to perform near-real time aquatic forecasts.The FLARE forecasting system is novel in its architecture, as it integrates a software-defined virtual distributed infrastructure spanning resources from sensor gateway devices at the edge of the network to cloud computing and storage. FLARE will support the flexible deployment of software in close proximity to water quality sensors in lakes and reservoirs, and in cloud resources for end-to-end data acquisition and processing. FLARE interconnects its distributed resources through a virtual private network to ensure data integrity and privacy in communications, and supports a flexible model applicable across a variety of lakes and reservoirs. Reusing best-of-breed technologies, FLARE builds upon and integrates several contemporary, widely-used open-source software frameworks in a manner that lowers the barrier to the deployment and management of ecological forecasting workflows by ecologists. Importantly, this project’s development of scalable and open-source cyberinfrastructure tools and end-to-end workflows for creating iterative aquatic forecasts will provide a critical resource for advancing the ecological forecasting research community, as well as provide a template for forecasting in other ecosystems. This project will build on and expand an existing program for cross-disciplinary teaching tools and research exchanges of undergraduate and graduate students to provide training at the intersection of computer science, freshwater science, and ecosystem modeling. Ultimately, this project will develop scalable, robust, secure workflows that will advance the capacity, practice, and training opportunities for ecological forecasting worldwide. Results from this project can be found at http://flare-forecast.orgThis 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.