Impacts of physics, resolution and nesting on hurricane simulation with regional NMMB

The Nonhydrostatic Multiscale Model on the B-grid (NMMB), which is the new generation of nonhydrostatic models developed at NCEP, is being built towards having the capability to simulate and forecast extreme weather events and has the potential to continue to advance our hurricane forecasting capability. This work reports progresses in developing and improving the skill of regional NMMB in the application of hurricane simulations. In the first series of experiments evaluating the impact of physics schemes on hurricane simulations, the regional NMMB is configured with 3 domains, a parent domain (18 km resolution) and two telescopic, movable and interactive nested grids (6 km and 2 km resolutions), to simulate several landfalling hurricanes. The grid setup is similar to that in the NCEP operational Hurricane Weather and Forecast model (HWRF). Two sets of runs were made, both with the same boundary and initial conditions. One uses the operational NAM physics suite, while the other adopts the newly available physics suite from HWRF. Results show that simulations using the HWRF physics package produced a better track, more accurate landfall time, and better intensification than those using the NAM physics suite. In addition, simulations from NMMB with the HWRF physics package are similar to those from the operational HWRF runs in terms of track, intensity, and structure. The impacts of individual physics schemes such as scale-aware deep convection, microphysics, and radiation schemes were also investigated. In the second series of experiments, regional NMMB was run to simulate hurricanes with different weights of nesting feedback and nested grid sizes. Besides, a single domain run using a full cloud resolving mode of regional NMMB over the entire parent domain was conducted as a benchmark to evaluate the impact of nesting feedback and nested domain size on hurricane simulations. Preliminary results suggest that the feedback of parent-nest interaction is important to simulate the impacts of the multi-scale dynamic and physical processes on the track, intensity, and structure of simulated hurricanes. Results from our experiments suggest that there is potential for further improvement in simulating hurricanes using regional NMMB when other HWRF components such as data assimilation, ocean coupling, and moving nest algorithm are integrated into regional NMMB. With proper physics schemes and grid setup, regional NMMB can provide more accurate high-resolution forecasts for extreme weather events.