Scientific applications have been at the forefront of driving computer and information technology since the early days: from the development of early computers for numerical computing, to the introduction in the USA of the NSFNET (which helped launch the Internet), and the subsequent invention of the World Wide Web. The geosciences, in particular, have been a long-standing user of such technologies, given the importance of applications related to weather, natural resources, natural hazards, and environmental monitoring. Scientific computing was focused initially on the need for fast computers to perform larger numbers of complex numerical calculations. The concerns more recently have turned towards the ability to manage the very large amounts of data that are being generated by a wide range of sensors and instruments, sophisticated observing systems, and large-scale simulations on large computer systems. Data rates of terabytes per day and petabytes per year are not uncommon (1 petabyte = terabytes) (Hey et al., 2009, p. 9). Yet, computer science and information technology solutions must deal not only with the size and scale of data, but also the inherent richness and complexity of scientific data – especially when data are combined across multiple projects, institutions, and even multiple science disciplines and subdisciplines. The need to understand complex, interdependent, natural as well as anthropogenic phenomena has made science a team sport, requiring collaborations among multidisciplinary teams of scientists to process, analyze, and integrate extremely heterogeneous data.