Abstract: The cloud-resolving scale detection and analysis of mesoscale dynamics and hydrometeors based on radar data assimilation are key to simulating and predicting severe thunderstorms and analyzing their formation and evolution mechanisms. In this study, based on the rapid updating and assimilation of S-band weather radar data, we reproduced a squall line accompanied by severe convective winds over the mountainous regions of wouthern Zhejiang Province and discussed the mechanisms for the mesoscale dynamical and microphysical processes involved. This in-depth study shows that: (1) rapid updating and assimilation of radar radial winds were crucial for reproducing the organizational progress of the squall line. The evolution and adjustment of mesoscale dynamics determined the development of the severe wind-producing convection system. (2) Cloud-resolved radar radial wind assimilation strengthened the mid-layer rear inflow and convection outflows and their convergence in the lower troposphere, which facilitated the organization of the squall line. (3) Assimilation of radar radial winds and reflectivity significantly changed the hydrometeor characteristics and enhanced the cooling of graupel melting below the 0°C layer in the stratiform region, as well as the evaporation of low-level raindrops. Consequently, the rear inflow descended along the slope toward the front of the convection while extending into the stratiform cloud region. This phenomenon played a key role in momentum transmission from mid- and high-level to the ground. (4) The rapid development of mesoscale dynamics and hydrometeors resulted in adiabatic warming generated by rear inflow downdrafts, which facilitated the evaporation of low-level raindrops and the strengthening of the cold pool. The contributions of the downward transport of high-level large momentum transmission and strong horizontal pressure gradient force near the surface greatly enhanced the formation of severe winds.