Doppler radar is capable of providing wind field information of precipitation echoes

and this makes it a powerful tool for the study of mesoscale weather processes. Assuming the change of radar reflectivity in a short time comply with the Lagrangian conservation

this paper proposes a new wind field retrieval method by which both radar reflectivity echoes and radial velocity data are used in the retrieval. Avoiding to assume the distribution of radial velocity in the retrieval

this method is theoretically applicable to all kinds of wind fields. This method is applied in two wind field simulation experiments

which use two temporally consecutive reflectivity echoes motions

and errors are generated. Error analysis shows that the retrieval results are more reliable when random error of the echo motion is less than 40% . Additionally

this method is used to analyze two mesoscale weather processes. The agreement between the wind retrieval results and the actual wind field structure proves the feasibility of this method in practice.