A Short Note on Data-Intensive Geospatial Computing

Abstract

For over 1,000 years, human knowledge has been recorded in printed formats such as books, journals, and reports archived and stored in libraries or museums. In the past decades, ever since the invention of computers, human knowledge can also be hidden in digital data. Although part of the data has been researched and reported, the pace of the data explosion has been dramatic and exceeds what the printed medium can deal with. This is particularly true for the twenty-first century, enormous data volume of scientific data, including geospatial data (NRC 2003), collected by all kinds of instruments in a variety of disciplines mostly on a 24/7 basis. The lack of related cyberinfrastructure for the data capture, curation and analysis, and for communication and publication alike led Jim Gray to lay out his vision of the fourth research paradigm – data-intensive science (Bell et al. 2009). This fourth paradigm differs fundamentally, in terms of the techniques and technologies involved, from the third paradigm of computational science (ca. 50 years ago) on simulating complex phenomena or processes. Before the computer age, there was only empirical science (ca. 1,000 years ago), and then theoretical science (ca. 500 years ago) like Newton’s Laws of Motion and Maxwell’s Equations. Nowadays, scientists do not look through telescopes but instead mine massive data for research and scientific discoveries. Every discipline x has evolved, over the past decades, into two computer- or information-oriented branches, namely x-informatics and computational x (Gray 2007, cited from Hey et al. 2009). Geography is no exception. Both geoinformatics, which deals with the data collection and analysis of geoinformation, and computational geography or geocomputation (Gahegan 1999), which has to do with simulating geographic phenomena and processes, are concurrent research issues under the banner of geographic information science.