Numerical Modeling study of hurricane boundary layer wind flow over mountainous terrain

Hurricanes are one of nature's most devastating natural disasters that can cause extensive damage along coastal communities and areas directly in the storm path. While it's usually high winds and storm surge that most often affects the shoreline areas, heavy rainfall can cause widespread flooding further inland. This problem can be made much worse if there are hills or mountains nearby that help to lift the air and subsequently, cause more rain to be produced on windward facing slopes. This makes the track of an impinging storm very important to where heavy rainfall can occur on the elevated topography, and this is where a simple hurricane model could prove to be very useful. The small time frame needed to produce a three-hour forecast (under 1 hour) and the high resolution (~5 km) of this model make it ideal for producing many forecast outputs for different land areas. This model could be used in conjunction with hurricane forecast models to cover an expanded area of land which would prove useful for storm track deviations. Many different features of the storm can be looked at with this model, but in this study we specifically concentrate on the hurricane boundary layer winds.

The model being run is a dry, high-resolution numerical model with specific focus on boundary layer processes. The model consists of fifteen levels from zero to sixteen thousand meters, seven of which are in the boundary layer (£ 1500 m). The low-level windflow and its interaction with the mountain topography are the main feature being looked at. The differences between computer generated terrain and actual land data are looked at, and several model runs over the island of Taiwan are also looked at in detail. Preliminary findings show that the greatest low-level wind speeds are over the highest mountain peaks, and height of the terrain is a direct factor to the top speed of the wind. Also found is that vertical velocities are greatest over the windward side of the mountain ridges and areas of sinking motion are found on the leeward side. Some evidence of increased winds due to downsloping effects can be seen, however the frictional effects of the surface tend to combat strong downsloping winds. Further studies on model accuracy are being looked at although the findings to date show good results and can prove to be very useful in mountainous regions.