A comparison of two tropical cyclone bogussing schemes

The impact of two bogussing schemes on tropical cyclone (TC) forecasts is compared. One scheme for bogussing TCs into the initial conditions of the nonhydrostatic version of the fifth-generation Pennsylvania State University-National Center for Atmospheric Research (PSU-NCAR) Mesoscale Model (MM5) is proposed by NCAR and the Air Force Weather Agency (AFWA), and four-dimensional variational data assimilation technology is employed for the other bogus data assimilation (BDA) scheme. The initial vortex structure adjusted by the NCAR-AFWA (N-A) scheme is more physically realistic, while the BDA scheme produces an initial vortex structure that is more consistent with the model. The results from 41 forecasts of TCs occurring over the western North Pacific (WNP) in 2002 suggest that the adjustment of the initial structure in the BDA scheme produces a greater benefit to the subsequent track and intensity forecasts, and the improvements in the track and intensity forecasts are significant using the BDA scheme. It seems that when using a model with 45-km grid length, the N-A scheme has a negative impact on the track forecasts for the recurving TCs and on the intensity predictions after 24 h.