Grid Computing for Mesoscale Climate Prediction
Abstract
The RECLIRS project is focused to establish an operational environment for seasonal climate prediction for the Rio Grande do Sul state, Brazil. The system uses four computational systems geographically distributed, where these systems are connected by the internet: a grid computing for climate prediction. The software platform includes two middlewares (OurGrid, OAR-Cigri: only one is necessary, but we are testing these two approaches), a portal for the internet, based on the gridsphere package, and the mesoscale meteorological model G-BRAMS. The model is the version for Grid Computing for the (Grid-) BRAMS (Brazilian Regional Atmospheric Modeling System) [1]. RAMS is a sophisticated computer model, and it can be used for research or for weather prediction. This model was developed by the Atmospheric Science Department from the Colorado State University. The Center for Weather Prediction and Climate Studies (CPTEC) from the INPE will be providing 50 years data for analysis, from the integration of CPTEC global meteorological model. The climatology can be obtained using an average over ensemble, where the ensemble is the long range integration with variation on initial and/or boundary conditions. The DERF/ECMWF strategy was employed to obtain the climatology. First preliminary predictions are described and discussed. References: [1] BRAMS, 2004: http://www.cptec.inpe.br/brams
- Publication:
-
AGU Spring Meeting Abstracts
- Pub Date:
- May 2008
- Bibcode:
- 2008AGUSMGC53A..07C
- Keywords:
-
- 3309 Climatology (1616;
- 1620;
- 3305;
- 4215;
- 8408)