A Data Serving Platform for HYCOM Ocean Prediction System Outputs in Support of Earth System Prediction Capabilities (ESPC)
Abstract
ABSTRACT Wide distribution of the global HYCOM ocean prediction system outputs is essential in order for the ONR scientific community to evaluate the usefulness of the outputs in (a) providing accurate boundary conditions for coastal and regional models, (b) supporting ONR DRIs, and (c) initializing Earth System Prediction Models. We have a comprehensive data management strategy to make the ocean prediction system outputs available to general users in near real time (i.e., within 24 hours). In order to achieve the above goal of efficient dissemination of the ocean model outputs to the community, it is not only necessary to have a computational platform that allows efficient storage, management, analysis, and distribution of the data generated by the HYCOM ocean prediction systems, but also the personnel to maintain such a platform and to ensure that users can easily access the model outputs. The Nexsan large scale Storage Area Network (SAN) currently provides up to one Petabyte (PB) of usable storage capacity. This storage is tightly coupled with our existing data processing clusters and builds on the existing infrastructure of the FSU Research Computing Center (RCC/HPC). The aforementioned components are presently utilized for the processing and storage of the 1/12¡ HYCOM global model outputs and is large enough to host outputs from the upcoming 1/25¡ global HYCOM with tides. However, the data serving services (virtualized FTP/THREDDS/OPeNDAP servers) have outgrown the processing, network, and memory of the HYCOM infrastructure initially purchased in 2007 and this legacy hardware lacks the necessary reliability and speed to serve efficiently outputs of the upcoming 1/25¡ global HYCOM with tides. In addition to making the large multi-terabyte datasets easily accessible by partners and the oceanographic community at large, the updated data serving platform will allow us to perform highly data intensive diagnostic tests that are needed to evaluate the performance of the ocean prediction system. The resulting technological infrastructure will also provide a state-of-the-art system for the training of students in ocean modeling and prediction as well as data management.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- May 22, 2016
- Source ID
- N000141512605
Entities
People
- Eric Chassignet
Organizations
- Florida State University
- Office of Naval Research
- United States Navy