Building the next generation of database systems for the aeolian sciences
Louis Scuderi, Gary Weissmann
University of New Mexico, Albuquerque, New Mexico, USA
There is a great need within aeolian science for data management systems capable of coping with highly complex multimodal queries, intelligent retrievals and knowledge discovery. This includes tools for recognition of features and phenomena in data embedded in complex environments (e.g., large multidimensional data sets, fuzzy/poorly defined boundaries) and dealing with the logical complexity inherent in physical phenomena (e.g., difficulty of detecting and tracking patterns evolving in a multidimensional space). Analyzing these complex environments, coupled with extraordinarily large field, sensor and satellite data sets, imposes staggering computational demands.
Aeolian systems are highly intertwined with other local, regional and global scale systems, and have significant impact on humans. Their study contributes to atmospheric, biologic, hydrologic, geomorphic and sedimentary science. The coupling of aeolian processes and features at multiple spatial and temporal scales, and with other disciplines, further highlights the need for these interrelated processes to be studied from an integrated perspective that transcends traditional disciplinary boundaries. However, there is a pronounced lack of integration of data into a scientifically credible, globally assembled, information platform to guide decision-making within aeolian science. It is not that relevant data does not exist. Rather, this data has not been (1) assembled from the many diverse sources available into a single data repository (2) used in an informed, consistent, and scientifically robust way, and (3) made easily accessible and available in the public domain.
Moving the field into a data rich future requires a coordinated and well-organized approach. However, creating this integrated framework remains difficult for a number of reasons. Least of all is the daunting size of a task that extends over continents and planets with time scales from milliseconds to millennia.
In this presentation we outline an organizational approach for making a global aeolian database a reality in the near future. We include examples from our approach showing how database technology can be used to (1) conduct innovative research over a range of spatial and temporal scales, (2) share data sets and innovative analytic tools, (3) collaborate, and (4) ask questions once considered computationally impossible. Considerable challenges stand in the way of building such a database, however the returns are very high in terms of expanding the range of research questions that can be tackled and in improving understanding and prediction within aeolian science.