In order to make the sudden rainstorm nowcasting and early warning more timely and accurately， using the detection data including conventional synoptic analysis, Beijing radar regional mosaic and VDRAS (variational Doppler radar analysis system) retrieved wind field, AWS (auto weather stations), wind profile data, and satellite images, we have conducted the comparative analysis on synoptic scale and mesoscale systems of two severe convective rainstorm events, which happened in Beijing area on August 14, 2008 and July 13, 2009, and put the forecast focus in the rainstorm nowcasting and early warning in Beijing area. It shows that the two rainstorm events have many similarities, e.g. in the good synoptic scale conditions, such as moisture condition, the dynamic condition at the middle and low levels, and the intrusion of the cold air, the clear areas between the prestorm southern and northern mesoscale systems of satellite images, the occurrence and development of mesoscale systems being associated with mesoscale convergence below 1 km ASL. And on the other hand, they have many differences in the direction and speed of the movement of the radar echoes, the structures of VDRAS wind fields in the lower troposphere, the structures of vertical wind profiles. It is proved that by the study and analysis, the nowcasting ability of severe convective rainstorm can be promoted using high space resolution detection data and there is important significance to the forecasting service in the main flooding season.