Oral Presentation AMOS Annual Meeting and International Conference on Tropical Meteorology and Oceanography

Boundary Layer Structure and Occurrence of Elevated Ducts associated with the Typhoon Lupit (#26)

Juli Ding 1 , Xiaogang Huang 1 , Xiaoping Cheng 1 , Zeming Sun 1 , Zhichao Liang 1 , Wei Zhong 1
  1. College of Meteorology and Oceanography, National University of Defense Technology, Nanjing, JIANGSU, China

On the basis of global positioning system dropsonde data from Dropsonde Observations for Typhoon Surveillance near the Taiwan Region (DOTSTAR) and the WRF model, the boundary layer structure of typhoon Lupit (0920) and the associated elevated ducts are investigated. In the inner core of Lupit with strong updrafts, the marine boundary layer is characterized by a shallow mixing layer (usually lower than 500m) overlaid by the above free atmosphere, with the specific humidity almost linearly declining with height. Due to the absent of typical inversion layer (also as entrainment zone), elevated ducts can hardly be observed. Otherwise, in the weak subsidence area outside Lupit and especially on the front side of its track, a typical marine convective boundary layer is detected as our expectation, where a much deeper mixing layer exceeding 1000m is capped by an entrainment zone about 200~400m thickness. Because a sharp decrease of moisture and a temperature inversion are both satisfied in the entrainment zone, in which elevated ducts are favorable to be formed. If the entrainment effect is combined with strong upper dry air advection from the north such as in this case, the observed and simulated elevated ducts can be much stronger since decreases of moisture with height are strengthened.

Our preliminary simulation shows a good performance of the WRF model with DI and YSU PBL scheme on simulating the typhoon Lupit and its associated ducts. But uncertainties of some turbulent scaling parameters in YSU PBL scheme are one of the main inaccurate sources. Among them, the exponent pfac (default as 2.0) in turbulent diffusivity K calculation is the most identifiable parameter, suggesting an effective way to reduce model errors by proper estimation of its value.