Big Data Supported Strategic Advancement in Model-Based Environmental Assessment and Management

包括气候变化和新出现的污染在内的环境问题正变得越来越复杂。特别是，这些环境问题正在对包括公共健康和生态系统在内的区域环境媒体产生实时不利影响。以前，环境多媒体命运、运输和暴露评估模型在控制风险和清理场地方面做出了重大贡献。例如，申请人提出了以下主要环境介质的环境模型，包括空气、土壤、地下水和地表水，据报道在过去十年中得到了现场应用：石油污染土壤和地下水的多相多组分化学去向和迁移模型，用于模拟污染物在加拿大近海水域中扩散的近场和远场耦合数值模型，以及应用于加州的地理信息系统支持的空间多源空气质量模型。*然而，以前的环境多媒体模拟研究存在挑战，包括模型精度不一致，与不同介质相关的多方面不确定性，以及数据问题。这些挑战与进一步改进模型和解决数据差距的需要有关。例如，从地面站和遥感器收集的数据量呈爆炸性增长，数据处理花费了上述模型开发和应用的一半或更多的时间和精力。因此，本研究计划提出以大数据技术为支撑的新的环境建模和评价概念和体系，从战略上扩展最近的研究进展。这一研究计划将(1)以高效的数值求解算法和非结构化计算网格改进先前开发的单一和多媒体化学命运和传输模型；(2)进一步开发高分辨率卫星数据处理方法，与现场智能传感器一起有效地支持环境建模和评估；(3)开发大数据支持的环境数据分析，包括基于网络的数据处理和云计算算法，用于新的基于分布式模型的决策工具；(4)开发针对复杂环境的系统不确定性量化模块；(5)开发用户友好的基于模型的决策支持系统。拟议的研究计划将有助于建立区域级别的大数据支持的新一代环境风险测绘和决策支持工具，以管理动态建筑和自然环境中出现的污染物和主要环境问题。*

Environmental problems including climate changes and emerging contamination are getting exponentially more complex. Particularly, these environmental issues are having real-time adverse effects on the regional environmental media including public health and ecosystem. Previously, environmental multimedia fate, transport, and exposure assessment models have contributed significantly to control the risks and to site cleanups. For example, the applicant has put forward the following environmental models for major environmental media including air, soil, groundwater, and surface water with reported field applications in the past decade: a multiphase multicomponent chemical fate and transport model for the petroleum-contaminated soil and groundwater, a coupled near- and far- field numerical model for modeling contaminant's dispersion in Canadian offshore water, and a GIS-supported spatial multi-source air quality model with application to California.**** However, there have been challenges of the previous environmental multimedia modeling studies including inconsistent model accuracy, multifaceted uncertainties associated with different media, and data issues. These challenges are related to the need of further model advancement and resolving data gaps. For instance, the volume of data collected from both ground station and remote sensors is explosively growing, data processing costs half or more of the time and efforts for the above-mentioned model developments and applications.**** Therefore, this research program proposes to put forward new environmental modeling and assessment concept and system supported by the big data technology to strategically extend the recent research developments. This research program will (1) advance the previously developed single- and multi- media chemical fate and transport models with efficient numerical solving algorithms and unstructured computational mesh, (2) further develop high-resolution satellite data processing method to effectively support environmental modeling and assessment together with field smart sensors, (3) develop big data supported environmental data analytics involving web-based data processing and cloud computing algorithm for new distributed model based decision tools, (4) develop systematic uncertainty quantification module for complex environment, and (5) develop user-friendly model-based decision support system. The proposed research program will help building regional level big data supported new-generation environmental risk mapping and decision support tools to manage emerging contaminants and major environmental issues within the dynamic built and natural environment.***