Towards Improving Explicitly Resolved and Sub-grid-scale Clouds in Hurricane WRF

Representing clouds on spatial scales from relatively small and isolated tropical cumulus to large sheets of stratus/stratocumulus as well as tropical cyclone-scale convective bands and cirrus anvils is extremely challenging. Considering NWP models like Hurricane WRF (HWRF) often contain grid spacings in the "grey zone" of convection and cloud formation, a strategy to couple microphysics, planetary boundary layer, radiation, and convective parameterizations is essential. The Developmental Testbed Center (DTC) has started this coupling of physics parameterization and tested a few alternative radiation and microphysics parameterizations.

A recent area of testing involves newer and more complex microphysics parameterizations by Thompson et al (2008) and Ferrier and Aligo (2015). These schemes show significant promise in terms of capturing various cloud types and formations but mostly at the grid-resolvable scales. Furthermore, other researchers (Bu et al, 2014) have shown the importance of properly accounting for cloud radiative feedbacks in tropical cyclone forecasting and rather clearly illustrated that the more advanced Rapid Radiative Transfer Model-Global (RRTMG) scheme has more physical realism than HWRF's predecessor Geophysical Fluid Dynamics Laboratory (GFDL) radiation scheme. Unfortunately, however, preliminary testing of the RRTMG scheme in HWRF initially produced worse tropical cyclone track/intensity forecasts than the GFDL scheme.

With this motivation, the DTC investigated the cause of the forecast degradation and created and evaluated a new method to treat sub-grid-scale clouds by the RRTMG scheme even in the absence of explicitly resolved clouds by the microphysics parameterization. This effort led to switching the radiation scheme from the 2014 to 2015 HWRF operational implementation. This talk will discuss the implementation and testing of the new fractional cloudiness scheme and its designed intent to be "scale-aware" since HWRF typically uses relatively coarse resolution parent grid and two higher resolution nests. Initial focus will be the improvements to cloud and radiation results, but we will also discuss the overall impacts to cyclone track/intensity.