Parallel Computation for Inversion Algorithm of 2D ZTEM
Round 1
Reviewer 1 Report
There is a tremendous computational burden in the computation of 2D Z axis tipper electromagnetic method. The computation cost in the coming big data and artificial intelligence era is decreasing fast. It is worthwhile to trade computation costs for higher computation efficiency. This study introduces MPI, one of the parallel computation techniques, to speed up the computation. The manuscript is well written, logically organized, and thus are appropriate for publication in Applied Sciences. I found that the code is not deposited in a public repository. The authors should upload their code and tell the readers how to download it. By doing so the influence of the study can be improved.
Below are minor suggestions that the authors should address appropriately. I hope the authors found them useful.
Line 7: Even if the abbreviation “ZTEM” has been defined in the main text, i.e., in Line 20, it should be defined separately in the abstract.
Line 21: Define “MT”
Line 22: “in the method”?
Lines 46-47: Please define the variables in the equations.
Line 108: 18 times
Line 148: The flow chart of parallel computing of 2D inversion is shown in Figure 2.
Line 206: “improves”: Do you mean “proves”?
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
In this research, the author use MPI parallel programming to realize the inversion calculation of two-dimensional ZTEM, which effectively reduces the inversion time.The model test proves that the method is effective.It is of great significance for practical application. The overall research idea is clear, but there are the following suggestions.
(1) For example, the text format in fig.1 and fig.2 needs to be adjusted. Some variable descriptions, such as "λ is positive", are not accurate. The author should check them again carefully.
(2) The author should add domestic and foreign research progress related to the research problem of this paper in the introduction to highlight the significance of innovation.
(3) For openMP and mpi mixed programming, whether the calculation speed can be further accelerated, whether the author will consider later.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
