The paper deals with a generally interesting topic, namely nonlinear effects on both the translational ground motion components

(which are the domain of traditional seismology), and on rotational components. Rotational seismology is currently a rapidly developing field, 

hence the subject is highly topical. The authors investigate the numerical solution of the equation of motion with a strain tensor that includes some 

nonlinear terms in addition to the linear ones and compare this with the 'classical' solution considering linear strain tensor that is commonly used 

because of the small-deformation assumption. For numerical simulations they use a staggered grid finite difference scheme. The authors then 

confront the simulations with  the records of two selected Taiwan earthquakes. 

Unfortunately the paper is not well written and will require significant refinement in a number of respects. The level of English is not good and often

makes it very difficult for the reader to understand the meaning. The quantities and equations should be better described in accordance to 

mathematical conventions. The numerical models do not appear to be properly tested, and there is a lack of numerical error estimates.

Comparisons with real data from two selected Taiwan earthquakes are not convincing. The references show deficiencies or errors. 

In my opinion, the paper requires MAJOR REVISION.

See the list of comments in the attached file.

Please find attached my comments.

Best regards,

The reviewer