Open AccessFinal Report for "Analyzing and visualizing next generation climate data"
Author(s) -
A. Pletzer
Publication year - 2012
Language(s) - English
Resource type - Reports
DOI - 10.2172/1054621
Subject(s) - spec# , computer science , block (permutation group theory) , metadata , grid , process (computing) , interpolation (computer graphics) , mosaic , database , data mining , operating system , programming language , computer graphics (images) , geography , animation , geometry , mathematics , geodesy , archaeology
The project "Analyzing and visualizing next generation climate data" adds block-structured (mosaic) grid support, parallel processing, and 2D/3D curvilinear interpolation to the open-source UV-CDAT climate data analysis tool. Block structured grid support complies to the Gridspec extension submitted to the Climate and Forecast metadata conventions. It contains two parts: aggregation of data spread over multiple mosaic tiles (M-SPEC) and aggregation of temporal data stored in different files (F-SPEC). Together, M-SPEC and F-SPEC allow users to interact with data stored in multiple files as if the data were in a single file. For computational expensive tasks, a flexible, multi-dimensional, multi-type distributed array class allows users to process data in parallel using remote memory access. Both nodal and cell based interpolation is supported; users can choose between different interpolation libraries including ESMF and LibCF depending on the their particular needs