Prestack elastic reverse time migration( RTM) requires multicomponent seismic data. But for multicomponent elastic Kirchhoff migration,there is a limitation that ray theory no longer applies if thegeology becomes complicated. In this paper,the authors have created a new 2D migration context for isotropic,elastic RTM,which included decomposition of the elastic source and receiver wavefields into P and S wave vectors by decoupled elastodynamic extrapolation,which retained the same stress and particle velocity components as the input data. Then we appliedsource-normalized crosscorrelation imaging condition in elastic reverse time migration to compensate the energy of deep strata. We found that the resulting images were nearly identical to the velocity model,and the resolution has been improved. Our method is a wavefielddecomposition based on vector,and we can alsoavoid the problem of polarity reversal of converted shear wave imaging. It proved the applicability of the method proposed in our paper. But for multicomponent elastic Kirchhoff migration, there is a limitation that ray theory no longer applies if the theory becomes complicated. In this paper, the authors have created a new 2D migration context for isotropic , elasticized RTM, which included decomposition of the elastic source and receiver wavefields into P and S wave vectors by decoupled elastodynamic extrapolation, which retained the same stress and particle velocity components as the input data. Then we appliedsource-normalized crosscorrelation imaging condition in elastic reverse time migration to compensate the energy of deep strata. We found that the resulting images were nearly identical to the velocity model, and the resolution has been improved. Our method is a wavefield decomposition based on vector, and we can also avoid the problem of polarity reversal of converted shear wave imaging. It proved the applicability of the method proposed in o ur paper.