Abstract:Because large quantities of automatic weather station (AWS) data are increasing dramatically as well as the weather forecast system and data assimilation system are developing continually, the current processing efficiency of AWS data is unable to meet the requirement of data users. Aimed at efficiency problem of AWS data, a dynamic multi process data processing model is proposed. In this model， configuration files are used to define handling scope of each process， and a process with parameter is designed to deal with the data of particular region which needs to be got efficiently. The model has been implemented in three cases such as disastrous debris flow at Zhouqu, Gansu Province on 7 August 2010. The result shows that data processing strategy can be adjusted easily in terms of important weather phenomena and data processing timeliness can be guaranteed, increasing data processing efficiency by 80%.