Cloud Computing and Spatial Cyberinfrastructure

In this perspectives paper, we review the current state of Cyberinfrastructure and illustrate opportunities that we see if Cloud Computing strategies are adopted. In summary, Cloud Computing provides elastically provisioned computing, software, and service infrastructure, typically implemented on a foundation of virtual machine and virtual data storage technologies. This elasticity allows users to outsource their computing infrastructure, growing or shrinking it as necessary. Commercial investments in Cloud infrastructure make it likely that these systems will dominate large-scale computing hardware and software in the next decade. Furthermore, open source Cloud software makes it possible for universities and research laboratories to investigate and build open-architecture clouds for scientific computing and other uses. Given these general advantages, we consider the applicability of the approach to scientific computing generally and Spatial Cyberinfrastructure specifically through two case studies (flood modeling and radar image processing). We map these projects’ requirements to both infrastructure and runtime capabilities typically provided by Clouds. Based on these case studies, we discuss gaps and research opportunities in Cloud Computing from the geospatial point of view. Our preliminary conclusion from this review is that Spatial Cyberinfrastructure’s requirements are a good match for many common capabilities of Clouds, warranting a larger scale investigation and research by the community.