Scalability of Saharan Dust Outbreak Modelling with the Advanced Weather Research and Forecasting Model Coupled with Chemistry (WRF-Chem)
摘要
A fully coupled meteorology and chemistry model, for example, the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem), analyzes air pollution diffusion, including the dust aerosol, and its radiative and microphysical effects. The Mediterranean area is frequently influenced by the Saharan Dust coming from North Africa, which is a relevant source of Particulate Matter (PM) impacting climate, air quality and human health. Moreover, the climate changes ongoing in the Mediterranean region suggest an enhancing frequency and intensity of the Saharan Dust events. In this study, we evaluated the forecasting skills of WRF-Chem simulations describing a Saharan Dust Outbreak (SDO) applied to a large domain, including Northern Africa and Central Europe. In particular, we analyzed the performances obtained using two different compilers, i.e., GNU (gfortran/gcc) and INTEL (ifort/icc). The results showed an improvement in time of the Saharan Dust simulation for INTEL over the GNU compiler, which ranges from 46% to 12% using respectively from 1 to 24 threads. Moreover, the strong scalability was assessed by analyzing the execution time as a function of the number of threads, finding that the best performance is achieved by 12 and 24 threads for INTEL and GNU, respectively.