Bi-Weekly to Monthly Typhoon Forecast Using a Global Model with Variable Mesoscale Resolutions

MPAS (Model for Prediction Across Scales) is a new-generation globalnonhydrostatic model which allows a convenient higher mesoscale resolution for aspecific domain of interest. We will attempt to fully explore and extend thefeasibility of MPAS to at least bi-weekly typhoon prediction as most regionalmodels (like WRF) will be significantly impacted by lateral boundary conditions atlater stages. In the three years, we will develop a MPAS framework of global60-15 km variable resolutions to conduct bi-weekly prediction for TyphoonMorakot (2009), Typhoon Megi (2010), Typhoon Haiyan (2013) and/or others,which will commence from an earlier spin-up stage to the final dissipation stage.This task will allow a long-term track prediction with a possible projection ofheavy rainfall over local regions like Taiwan as the steep topography can bereasonably resolved by the global model. In the first year (2015 phase), we haveinvestiagted the impacts of various cloud microphysical schemes, cumulusparameterizations and PBL parameterizations on prediction of typhoon tracks andlocal rainfall for Typhoon Haiyan. In the second year (2016 phase), with a 30-6km MPAS (single zoom), we will develop suitable dynamic typhoon initializationand investigate various scenarios for synoptic moisture, topography and seasurface temperature to highlight the roles of terrain and oceanic fluxes inprediction of typhoons impinging Taiwan, and currently, the research work isongoing very well. In the last year (2017 phase), this study will employ low-passfilter to construct various scenarios for large-scale environmental flow and focuson understanding the influence of environmental flow (including MJO) and itsvariations on northward turning of typhoons. Two relevant events, TyphoonsMorakot (2009) and Megi (2010) have been completed. We will also conduct60-15-3-km (double zooms) MPAS simulations compared with regional HWRF3-km simulations to assess the relative advantage of a global model at predictingconvective rainfall. This project will establish the skill of long-term (bi-weekly andbeyond) forecasts for track and rainfall associated with typhoons impingingTaiwan using the advanced global MPAS at high resolution.