The quest for better computing infrastructure for geospatial applications has been a constant endeavor for geoscientists. With the recent proliferation of cloud offerings, a range of new opportunities have become available. The challenge, however, is to make the best use of cloud platforms. Two directions are particularly important for addressing this challenge: a) developing new design approaches that are suitable for geoscience applications destined for the clouds, and b) accurately assessing

more »

... the level of performance that can be expected when a given application is hosted on a given cloud platform with a specific configuration. This would enable scientists to better choose cloud solutions. In this paper, we focus on the latter direction. We use a typical data-and compute-intensive geoscience application, namely spatial interpolation, as a case study to assess the benefits of cloud computing for geoscience applications. We study the performance of the application on several types of cloud instances and provide a cost/benefit analysis that gives useful insights to geospatial and Earth scientists when they consider cloud options.