Liu, C., Wang, D., Hu, B. and Wang, T., 2016. Seismic deconvolution with shearlet sparsity constrained inversion. Journal of Seismic Exploration, 25: 433-445. The application of conventional deconvolution methods must be under some assumptions, meanwhile the processing procedure is through single trace cycle, which may destroy the continuity of seismic events. Besides, these methods are serious interfered by noise. For these reasons, we proposed seismic deconvolution based on multiscale and multidirectional shearlet transform sparsity constrained inversion. Shearlet has the ability to represent multidimensional signals with optimal sparse representation. We expressed the reflected signals sparse characteristic by the sparse shearlet coefficients. Deconvolution based on multi-dimensional space transform maintains the continuity along reflectors theoretically compared to the traditional single channel method. We expressed the deconvolution problem as a 1-norm optimization problem and combined with a fast iterative thresholding algorithm. Experiments on synthetic and field seismic data show our method could improve the resolution of seismic data effectively, attenuate random noise and make the events more smoothly.