Data Management Challenges in Coastal Applications
Sai Pranay Tummala, Tevfik Kosar · 2008
The goal of this paper is to identify the data management challenges in coastal applications such as hurricane track prediction, storm surge modeling and coastal erosion modeling. The problems in managing the data due to different paradigms such as increase in data and computational requirements, conceptual changes in the computational models involved, and changes due to the evolution of objectives of the models are explained. Potential problems in a complete processing cycle that can be solved using automation are enumerated right from the selection of input data to the archival of output data or feeding the output data into a visualization system. Challenges to the user like having to complete the data management operations manually, to learn the underlying complexity of the resources, and to intervene during data placement failures are explained. Specialized tools for data placement and automation of specific tasks and the workflow mechanisms that are being used currently to automate the entire end-to-end cycle of scientific computation are mentioned as a solution to these problems.