Cloud Hosted Real‐time Data Services for the Geosciences (CHORDS)
Cloud Hosted Real‐time Data Services for the Geosciences (CHORDS)
复制标题
云托管地球科学实时数据服务 (CHORDS)
DOI：
10.1002/gdj3.36
复制
发表时间：
2016
影响因子：
3.2
通讯作者：
F. Vernon
中科院分区：
文献类型：
--
作者：
B. Kerkez;M. Daniels;S. Graves;V. Chandrasekar;K. Keiser;Charlie Martin;M. Dye;M. Maskey;F. Vernon
虽然现代传感和通信技术能够以前所未有的时空分辨率观测地球物理过程，但这些技术的发展速度大大超过了它们在地球科学中的实际应用。对于实时数据系统来说尤其如此，它现在允许在测量的瞬间对数据进行流化和分析。虽然实时科学数据的使用是有限的，但它们的重要性正在不断增加，特别是在必须迅速做出明智决策的关键任务场景中。除了与抗灾能力（地震预测、洪水预测等）相关的应用之外，现在比以往任何时候都更有可能利用实时数据从根本上改变科学实验的进行方式。例如，在许多地球科学实验中，故障传感器往往发现得太晚，迫使实验重复进行。在使用移动传感器节点的情况下，或者需要调整采样频率以捕获感兴趣的事件的情况下，很少有工具可以自适应地指导实验过程。这通常会导致错过观测和浪费实验投资，但可以通过启用分析和响应流数据的手段来快速补救。虽然实时数据能够实现地球科学实验的范式转变，但它们很少（如果有的话）成为地球科学工作流程的第一步。绝大多数现有数据平台天生就适合非实时应用程序，其中数据通常存储在大型数据库中，以便进行回顾性分析和可视化。然而，现有的少数实时数据平台要么是专有的，要么是为特定任务工具提供的，要么是地球科学领域更广泛的利益相关者无法使用的。虽然这些平台的复杂性对实时数据系统的广泛采用构成了主要障碍，但在实时数据在地球科学领域变得普遍之前，还有许多技术挑战必须解决。
While modern sensing and communication technologies are enabling the observations of geophysical processes at unprecedented spatiotemporal resolutions, the development of these technologies is significantly outpacing their actual use across the geosciences. This is particularly true of real-time data systems, which are now permitting the streaming and analysis of data at the instant of their measurement. Though the use of real-time scientific data is limited, their importance is ever increasing, particularly in mission critical scenarios where informed decisions must be made rapidly. Beyond applications tied to disaster resilience (earthquake prediction, flood forecasting, etc.), now more than ever there is potential to leverage real-time data to fundamentally change how scientific experiments are conducted. For example, in many geoscientific experiments, faulty sensors are often only detected too late, forcing experiments to be repeated. In settings where mobile sensor nodes are used, or where sampling frequencies need to be adjusted to capture events of interest, few tools are available to adaptively guide the experimental process. This often results in missed observations and wasted experimental investments, but can be remedied rapidly by enabling means to analyse and respond to streaming data. While real-time data stand to enable a paradigmshift in geoscientific experimentation, they rarely, if ever, form the first step in a geoscientific workflow. The vast majority of existing data platforms are inherently tuned to nonreal-time applications, where data are often stored in large databases for retrospective analysis and visualization. The few existing real-time data platforms, however, are either proprietary, feed into mission-specific tools, or are otherwise not available to broader stakeholders within the geosciences. While the complexity of these platforms presents a major barrier to the broader adoption of real-time data systems, there are also a number of technical challenges that must be addressed before the use of real-time data becomes commonplace across the geosciences.