CyberSEES: Type 1: Meghdoot: A Multi-Cloud Infrastructure for Enhancing Sustainability via Effective Monitoring of Inland Waters and Coastal Wetlands

该项目的总体目标是多云计算框架的设计和原型实现，该框架无缝集成了社区观测、遥感测量和先进的多媒体分析，以实现有效的环境监测。这个被称为Meghdoot的基础设施将用于早期检测沿海地区的盐沼压力和内陆水域的蓝藻有害藻华。该框架将利用多个云，包括社区云、传感器云和计算云，来开发有效和及时的事件检测策略。此外，将调查激励和使社区成员为加强环境监测提供高质量内容的有效机制。拟议项目的第一个研究重点是发明有效利用社区云监测湖泊和盐沼的机制。在这种社区即传感器范式中，在线社交媒体平台（Facebook、Twitter、Flickr等）上社区个人产生的数据（即博客、图像、推文）将用于生成可信的、可操作的信息。反过来，这将作为启动传统传感基础设施的初始触发器。由高分辨率相机和高光谱辐射计组成的传感器云的数据将通过计算云使用专业技术进行处理，如分割、特征提取、注册、索引和光谱模型，以产生来自研究地点的蓝藻浓度和沼泽生物物理特征的估计。本项目最终的研究重点是基于社区云的信息和计算云的结果，优化传感器云的部署和运行。该项目解决了对美国东南部沿海州重要的两个环境问题，即有害藻华和沼泽褐变事件。在过去十年中，由于干旱和海平面上升，这些对环境有害的事件发生的频率和程度都有所上升。因此，对这些事件进行准确、具有成本效益和针对性的监测对于环境的可持续管理是必不可少的。拟议的基于网络基础设施的预警系统将使早期发现和及时实施先发制人的措施，以减少未来事件的频率和严重程度，同时确保环境保护和可持续性。该项目计划通过培训、研讨会和社交媒体，让学生、社区领袖、资源管理者和公众参与研究的各个方面，从众包到环境传感器部署、数据采集、处理和解释。拟议项目的成功将为全州范围内的自动早期检测、预警和快速反应系统铺平道路，该系统可被州级环境机构采用，以提醒修复官员和普通公民与这些事件相关的即将发生的风险。

The overall goal of the project is the design and prototype implementation of a multi-cloud computing framework that seamlessly integrates community observations, remote sensing measurements, and advanced multimedia analytics for effective environmental monitoring. This infrastructure, called Meghdoot, will be employed for early detection of salt marsh stress in coastal areas and cyanobacterial harmful algal blooms in inland waters. This framework will harness multiple clouds including community clouds, sensor clouds, and computational clouds to develop efficient and timely event detection strategies. In addition, efficacious mechanisms for motivating and enabling community members to contribute high-quality content for enhanced environmental monitoring will be investigated. The first research thrust of the proposed project is to invent mechanisms to effectively leverage community clouds for monitoring lakes and salt marshes. In this community-as-sensors paradigm, data produced by individuals of the community (i.e., blogs, images, tweets) in online social media platforms (Facebook, Twitter, Flickr, etc.) will be used to generate trustworthy, actionable information. This, in turn, will act as the initial trigger for activating the traditional sensing infrastructure. The data from sensor clouds comprised of high resolution cameras and hyperspectral radiometers will be processed through the computational cloud using specialized techniques such as segmentation, feature extraction, registration, indexing, and spectral models for producing estimates of the cyanobacteria concentration and marsh biophysical characteristics from the study sites. The final research thrust of this project will focus on the optimization of the deployment and operation of the sensor cloud based on the information from the community cloud and the results from the computational cloud.This project addresses two environmental issues important to coastal states in southeastern United States, namely, harmful algal blooms and marsh browning events. Over the last decade the frequency and magnitude of these environmentally detrimental events have gone up primarily because of drought as well as sea level rise. Therefore, accurate, cost-effective, and targeted monitoring of these events is indispensable to sustainable management of the environment. The proposed cyber-infrastructure-based warning system will enable early detection and timely implementation of preemptive measures to reduce the frequency and severity of future events while ensuring environmental conservation and sustainability. The project plans to engage students, community leaders, resource managers, and the general public via training, workshops, and social media in various aspects of the research starting from crowd sourcing to environmental sensor deployments, data acquisition, processing, and interpretation. The success of the proposed project will pave the way for a state-wide automated early detection, warning and rapid response system that can be adopted by state-level environmental agencies to alert restoration officials and lay citizens of impending risks associated with these events.