Abstract
In this study, a realtime wildfire smoke prediction model is developed as an application of the PUFF model, which has been developed for the purpose of the realtime volcanic ash prediction. As a case study, the PUFF model is applied to the wildfire occurred in May 2003 near Lake Baykal and near Lake Khanka in Siberia. As a result of the PUFF model simulation, it is found that the fire smoke from Lake Khanka can reach northern Japan on 22 May 2003 if the smoke is released on 18 May 2003. The result explains the observed aerosol increase at around 3000 m height in northern Japan and the red sun observed in Sapporo Japan. There was another huge wildfire near Lake Baykal reported simultaneously by the NOAA satellite. However, it is found that the smoke is unlikely to reach Japan by the upper air wind condition. It is inferred by the PUFF model simulations that the abnormal increase of aerosols around 22 May 2003 in northern Japan was originated from the wildfire near Lake Khanka released around 18 May 2004.
