Abstract: In an effort to assess the impact of the individual components of meteorological observations (ground ased GPS precipitable water vapor, automatic and conventional meteorological observations) on the torrential rain event of 4-5 July 2000 in Beijing (with 24 h accumulated precipitation reaching 240 mm), 24 hour observation system experiments are conducted numerically by using the MM5/WRF 3DVAR system and the non-hydrostatic MM5 model. Results indicate that, because the nonconventional GPS observations are directly assimilated into the initial analyses by 3DVAR system, better initial fields and 24h-simulation for the severe precipitation event are achieved than those under the MM5/Litter objective analysis scheme. Further analysis also shows that each of the individual components of meteorological observation network plays its special positive role on the improvement of initial ield analysis and forecasting skills. With or without radiosonde and pilot observation in 3DVAR scheme has the most significant influence on numerical simulation, automatic and conventional surface meteorological observations ranks into the second place. After ingested the supplement information from the other meteorological observations, the ground ased GPS precipitable water vapor data can play more obvious roles on initial ield assimilation and precipitation forecast. By incorporating the ground ased GPS precipitable water vapor data into the 3DVAR analyses at the initial time, the threat scores (TS) with thresholds of 1, 5, 10 and 20 mm are increased around 1%-8% for 6 and 24 hours accumulated precipitation observations. This work gives one helpful example that assesses the impact of individual components of the existing Meteorological Observation Network on high influence weather event using 3DVAR numerical system.