The HIRLAM (high resolution limited area modelling) limited-area atmospheric model was originally developed and optimized for shared memory vector-based computers, and has been used for operational weather forecasting on such machines for several years. This paper describes the algorithms applied to
Parallelization and performance of a meteorological limited area model
โ Scribed by M Marrocu; R Scardovelli; P Malguzzi
- Publisher
- Elsevier Science
- Year
- 1998
- Tongue
- English
- Weight
- 195 KB
- Volume
- 24
- Category
- Article
- ISSN
- 0167-8191
No coin nor oath required. For personal study only.
โฆ Synopsis
We have developed a parallel implementation of a meteorological limited area model using a standard domain decomposition technique. Several communication strategies and message-passing libraries are compared. Results are presented for low, medium and high spatial resolutions. A good performance is obtained in the high resolution test case with an efficiency higher than 85% of the theoretical maximum.
๐ SIMILAR VOLUMES
## Abstract In this paper four different postโprocessing techniques: the Bayesian model averaging (BMA), the ensemble model output statistics (EMOS) with a variant known as EMOS+ and a new dressing kernel (DRESS) are applied and compared, in a preโoperational context, to calibrate a mesoscale multi
## Abstract A limited area, sixโlevel, primitive equations numerical weather prediction model has been developed to provide improved 24โ to 36โhour forecasts for the Australian region. Important features of the model include semiโimplicit time differencing, and a nesting procedure which enables bou