Running Large and Dynamic Climate Ensembles in the Cloud

Climate and meteorological oriented software and simulations usage in Cloud Systems is growing and has proven to be a feasible and solid solution for these scientific applications.

In previous works, we exposed how Cloud systems can be the base to run climate simulations for researchers that can not or do not want to manage the needed computational resources.

Here we present our work, results, and findings, including insights about performance and high availability, about running large ensembles with WRF over BOINC (Berkeley Open Infrastructure for Network Computing) on Cloud Computing environments.