Temporal Analysis of Ground Movement at a Metal Mine in China
Round 1
Reviewer 1 Report
Dear authors,
It is an interesting ms for a hot problem in an area of mining. However, the experiment for gathering the data does not appear at all; i.e. do you have permanent monuments in this nickel mine? Are the monuments installed from the early beginning of the mining activity etc? Errors in the measurements do not appear in the paper. You mention about 500 monument measures, but you present only 6 monuments showing changes. Also, horizontal movement is not saying anything to me, please convert all changes to N-S or E-W. Although subsidence is well understood, uplift is not explained at all. Overall, although I am glad to see this analysis, the manuscript is not convincing someone that the phenomenon is well analysed and the conclusions are solid. In several parts, the manuscript is a review paper about the fractal properties of changes in the earth-surface in a mining area. Finally, it is not a paper suitable for publication, but trying to save this work, I propose its rejection and resubmission.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors of the manuscript “Temporal analysis of ground movement at a metal mine in China” analysis the GPS time-series of some points in the area of the Jinchuan Nickel Mine (China). This study is focused on the stochastic characterization of the time-series using Power spectrum and autocorrelation analysis.
In my opinion there are some critical points in the manuscript, that I list below.
The experiment should be described with more details because I-‘m not sure to really understand the type of time series used. The authors write that “Data sets consists of 500 random selected GPS monitoring points, spanning the April 2001–October 2015 time period” and “The monitoring frequency was 6 months, and it had been going on for 14.5 years”. So do the author refer to campaign data instead of continuous GPS stations? In this case, as the campaign have been performed? I mean session length, monumentation and so on.
In the de-trending of the time series do the author take into account of seasonal signals?
The authors affirm that the spectral exponents of horizontal and vertical displacements range from 0.47 to 3.58 and 0.43 to 3.37, so they are in the stationary range and non stationary range, but these different behaviors with respect to the stationarity condition not taken into account in the discussion of the results.
The autocorrelation plot of the 6001 point reported in Figure 4 shows a very short correlation length, of about 2 months, so a long-range persistence is not aspected.
Moreover, concerning the autocorrelation analysis the authors just report the analysis for point 6001, although they affirm to have analyzed 500 monitoring points. So, in my opinion, the authors should report more information, for instance a summary of the statistical proprieties of these autocorrelation function.
Minor remarks:
What do the authors mean with the sentence “recorded for October 2015, using May 2005 as a baseline” in the Figure Caption 2(Line 155)?
In Fig. 5 the indices (a), (b),… are missed, although they are quoted in the figure caption.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
The paper offers original supposition of the structure of the of rock movements as a consequence of the underground mining. That will be of interest to the readers. At the same time, I found few ambiguities and inaccuracies that I would like to express below.
1. “The ground movement process can be regarded as a stochastic process”. That is not always true. At least, on the plots in Fig there clear deterministic term – oscillations of the movements level with period of about 12 months. Also, authors mention some trends in the data values and propose to remove the trends.
2. Line 133: “The nominal accuracy is 3 mm ± 0.5 ppm”. What does the accuracy ± 0.5 ppm mean? Is that parts per million? Why?
3. Line 135: “The monitoring frequency was 6 months, and it had been going on for 14.5 years.” Is it true, that the GPS measurements were taken twice a year during 14.5 years and each time series consists of 28-29 data samples? At least, in Fig 3 one can see only 28 points in each time series.
If so, then taking in mind the fact the GPS samples are real numbers, the highest observable frequency will be 1/ (12months) . Also, for the 180 months observation interval the frequency resolution will be 1/(180months).
4. Why the eqs (4) and (5) are identical? The latter one should be Discrete Fourier transform of the autocorrelation function.
5. The equation 4 looks strange. In fact, there is no need to require N to go to infinity. The definition of the spectrum of discrete Fourier transform is true in the case of any N. Authors themselves calculate spectra of the time series for finite values of N (probably, N=28).
6. Fig 5 is very unclear. The plots should represent spectra calculated using Fourier transform. Horizontal axis there is frequency axis. As I mentioned earlier, the maximal unaliased frequency equals to the half of sampling frequency, which is 1/(6 months). The frequency samples spacing in the Fourier spectrum will be about 1/(180months). Very unclear, how authors got the frequency of the spectral peak as 0.9558/month. Also, according to the plots of time series in Fig. 3, there is strong 12 months period of oscillations, and the peak value should be located at about 1/(12months) frequency, on the right side of the plots in Fig. 5. And why the peak is unseen on the log-log plots? So far, the scale of the frequency axis seems to be wrong, and the spectrums of the time series are unclear.
7. Line 342: “The sinusoidal shape also appears in autocorrelation plots”. The statement is not supported by the illustrations. The manifestation of the oscillation with ~ 1 month period (according to authors) on the plot of autocorrelation function from Fig. 4 with 10 months spacing of samples is not obvious.
8. The idea of self-affinity and long-range persistence of the mining induced rocks movements is of high interest. In Discussion section authors mention different factors controlling the ground movement magnitude. I think the results obtained in the paper will constitute a good “basis to study the underlying patterns of mining induced ground movement in more detail” in future.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
No more changes are needed
Author Response
Thank you for your valuable comments which have greatly improved this paper. The manuscript has been revised for another round according to the comments from the academic editor.
