afernandezody
Member
Hello,
I've used a limited-area version of MPAS-A to try forecasting Hurricane Erin. In order to be able to make an assessment, I ran the model twice, initially as a forecast once it began to form (about 3 weeks ago) and the 2nd run earlier this week with the updated GFS data. Both simulations look good but the latter is obviously closer to the actual path. Another minor difference is that the hurricane seems to lose cohesiveness earlier in the forecast vs the reanalysis.
Said that, my question is in any potential restriction on how to use the model for hurricane forecasting. The paper discussing the Typhoon Morakot is using microphysics scheme (WSM6), the Tiedtke convective cumulus scheme, Yonsei University (YSU) planetary boundary layer parameterization, Monin–Obukhov surface layer parameterization, the Noah land surface model, and the Rapid Radiative Transfer Model for GCM applications (RRTMG) longwave and shortwave radiation. The used 'mesoscale_reference' model seems to use the revised Monin-Obukhov and new Tiedtke models. My concern is about the latter as the User's Guide states "This suite has been tested for mesoscale resolutions (> 10 km cell spacing), and is not appropriate for convective-scale simulations because the Tiedtke scheme will remove convective instability before resolved-scale motions (convective cells) can respond to it". I was wondering if I should expect to model to break down at finer resolutions or if there is any correlation predicting which would be the finest recommended resolution. Any other recommendation for hurricane forecasting would (of course) be welcomed,.
Thanks,
Arturo
I've used a limited-area version of MPAS-A to try forecasting Hurricane Erin. In order to be able to make an assessment, I ran the model twice, initially as a forecast once it began to form (about 3 weeks ago) and the 2nd run earlier this week with the updated GFS data. Both simulations look good but the latter is obviously closer to the actual path. Another minor difference is that the hurricane seems to lose cohesiveness earlier in the forecast vs the reanalysis.
Said that, my question is in any potential restriction on how to use the model for hurricane forecasting. The paper discussing the Typhoon Morakot is using microphysics scheme (WSM6), the Tiedtke convective cumulus scheme, Yonsei University (YSU) planetary boundary layer parameterization, Monin–Obukhov surface layer parameterization, the Noah land surface model, and the Rapid Radiative Transfer Model for GCM applications (RRTMG) longwave and shortwave radiation. The used 'mesoscale_reference' model seems to use the revised Monin-Obukhov and new Tiedtke models. My concern is about the latter as the User's Guide states "This suite has been tested for mesoscale resolutions (> 10 km cell spacing), and is not appropriate for convective-scale simulations because the Tiedtke scheme will remove convective instability before resolved-scale motions (convective cells) can respond to it". I was wondering if I should expect to model to break down at finer resolutions or if there is any correlation predicting which would be the finest recommended resolution. Any other recommendation for hurricane forecasting would (of course) be welcomed,.
Thanks,
Arturo
