When Spatial Analytics Meets Cyberinfrastructure: an Interoperable and Replicable Platform for Online Spatial-Statistical-Visual Analytics
When Spatial Analytics Meets Cyberinfrastructure: an Interoperable and Replicable Platform for Online Spatial-Statistical-Visual Analytics
复制标题
DOI：
10.1007/s41651-020-00056-5
复制
发表时间：
2020-06
影响因子：
4
通讯作者：
H. Shao;Wenwen Li;Wei Kang;Sergio J. Rey
中科院分区：
文献类型：
--
作者：
H. Shao;Wenwen Li;Wei Kang;Sergio J. Rey
开发空间分析方法作为开源库是实现开放和可复制科学的重要奋进。然而，尽管大型地理空间数据和地理空间网络基础设施（GeoCI）资源正在变得可用，但许多库和工具包仅在桌面环境中初始化和设计用于分析。将空间分析功能与大数据和高性能计算相结合，将在解决具有挑战性的社会经济和环境问题方面为多学科研究带来直接利益，并支持物理分布研究小组参与者之间的远程协作，并协助知情决策。在这篇文章中，我们提出了一个通用的工作流程的设计和实现，以集成国家的最先进的开源库与GeoCI资源。我们还解决了在实施过程中出现的各种互操作性和可复制性问题。流行的开源Python空间分析库（PySAL）被选择来构建可互操作的Web服务WebPySAL，然后成功地集成到GeoCI中。通过空间分析和网络基础设施之间的这种集成，新的GeoCI平台为公众用户提供了易于使用，高效和交互式的探索性空间分析功能。通过两个区域经济案例研究证明了GeoCI的能力：（1）评估全球空间自相关性，并确定美国各县家庭收入中位数空间模式中的局部聚类（使用全球和局部Moran'sI统计）;（2）在州一级对人均收入的时空动态进行建模（使用空间马尔可夫统计）。
Developing spatial analytical methods as open source libraries is an important endeavor to enable open and replicable science. However, despite the fact that large geospatial data and geospatial cyberinfrastructure (GeoCI) resources are becoming available, many libraries and toolkits are only initialized and designed for analytics in a desktop environment. Coupling spatial analytical functionality with big data and high-performance computing will result in immediate benefits for multidisciplinary research in terms of addressing challenging socioeconomic and environmental problems, as well as supporting remote collaboration between participants from physically distributed research groups, and assisting informed decision-making. In this article, we present the design and implementation of a general workflow to integrate state-of-the-art open source libraries with GeoCI resources. We also solve various interoperability and replicability issues that arise during the implementation process. The popular open source Python Spatial Analysis Library (PySAL) was selected to build the interoperable Web service, WebPySAL, which was then successfully integrated in GeoCI. With this integration between spatial analytics and cyberinfrastructure, the new GeoCI platform provides easy-to-use, efficient, and interactive exploratory spatial analysis functions to public users. The GeoCI capability is demonstrated through two regional economic case studies of (1) evaluating global spatial autocorrelation and identifying local clusters in the spatial pattern of median household incomes for US counties (with global and local Moran’sIstatistics) and (2) modeling the space-time dynamics of per capita incomes at the state level (with spatial Markov statistics).