WRF-SFIRE: Análisis de rendimiento y optimización de recursos en ambientes HPC

Abstract

The WRF-SFIRE model has become a valuable tool for wildfire prediction worldwide. However, its performance in parallel environments can be affected by the complexity of the model and resource limitations. This study evaluates the model’s performance in parallel environments and proposes optimization strategies to improve its efficiency and resource management/utilization. Representative cases will be conducted to assess the model’s performance with different software and hardware configurations, including CPU/GPU. The results of this study
can be useful for enhancing wildfire prediction capabilities and resource management efficiency in emergency situations.