Study and development of grid computing technology with potential application in numerical weather forecasts

• Grant number: 10/05823-7
• Support Opportunities: Scholarships abroad - Research
• Start date: September 01, 2010
• End date: August 31, 2011
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Eugênio Sper de Almeida
• Grantee: Eugênio Sper de Almeida
• Host Investigator: Michael Anthony Bauer
• Host Institution: Instituto Nacional de Pesquisas Espaciais (INPE). Ministério da Ciência, Tecnologia e Inovação (Brasil). São José dos Campos , SP, Brazil
• Institution abroad: Western University , Canada

Abstract

Nowadays, one of the great computational challenges in meteorology is to research techniques to ensure atmospheric numerical models execution and to reduce publishing time of their results to the whole society. This challenge arises from the incorporation of additional physical-chemical processes, for example, in chemical weather forecast in atmospheric models and from its spatial resolution increase. The Center for Weather Forecasting and Climate Studies (CPTEC), belonging to the National Institute for Space Research (INPE), execute atmospheric models daily on a global and regional (South America) scale. Results from these models execution are disseminated for several Brazilian institutions (IAG / USP, FUNCEME-EC, SIMEPAR-PR, etc.), that use them as input for model execution at local scale (high spatial resolution and limited geographic coverage). The execution of atmospheric models at local scale is dependent on the conclusion of model execution at regional scale by INPE/CPTEC. This research project aims to study operational or production computational grids, to evaluate its effectiveness in shortening dissemination time of atmospheric models information at local scale. For this purpose, possible techniques for synchronized execution of local scale model with regional model will be investigated. Computational grid SHARCNET (Shared Hierarchical Academic Research Computing Network) will be used in order to simulate this environment. As application and infrastructure can fail, fault tolerance mechanisms employed at SHARCNET will be studied. Expect results from this research project is knowledge increase on computational grids and a solution to ensure that results of local scale models can be disseminated to the entire society in a shorter time. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

DE ALMEIDA, EUGENIO SPER; BAUER, MICHAEL ANTHONY; FAZENDA, ALVARO LUIZ. Numerical weather model BRAMS evaluation on many-core architectures: a micro and macro vision. International Journal of Computational Science and Engineering, v. 12, n. 4, p. 330-340, 2016. (10/05823-7)