Ionospheric Scintillation Prediction on S4 and ROTI Parameters Using Artificial Neural Network and Genetic Algorithm

Round 1

Reviewer 1 Report

General comment:

The subject of this paper is obviously actual and the method of scintillation prediction provided by the authors is novel. However, the manuscript quality needs of being improved significantly before accepting. My recommendation is major revision. Hope that my comments below will help the authors to improve the manuscript quality.

 

Major comments:

• Abstract: “…..about 81% for S4 and about 80% for 29 ROTI”. S4 and ROTI indices relates one to another but not in such a close dependence (see Bhattacharrya, 2000). Please provide your reasons is it typical relations between the S4 and ROTI prediction or it depends on conditions and may vary significantly?
• Introduction section is too long and it needs to be cut off by replacing waste information with appropriate references and more details about the motivation of the paper. In my opinion, motivation of the paper starts from the Line 88 only. Lines 44-62 are mostly waste information. It is enough to leave 1-2 sentences about “plasma bubbles” as a source of the amplitude and phase scintillations within the magnetic equator region and add following references. I recommend to include following references: Ma and Maruyama, 2006; Anderson and Straus, 2005; Demyanov at al, 2012. Lines 63-86 – are the same. There is a lot of waste information about well known things. I would recommend to the authors cut off the text till 2-3 sentences and just insert the references.
• Lines 147-148: “….The parameter S4 is defined by the standard deviation of the S1 and S2 values….”. CNR at L1 and L2 are not the same as signal intensity (look Tiwari and Strangeways, 2015). The authors need to provide details about the procedure of transformation between CNR (i.e. S1, S2) – signal intensity (I) and S4. The integration time for S4 computation is also crucial factor to detect scintillations with high confidence (see Demyanov et al., 2019a). Hence the authors needs to provide details about this integration time (the explanations like “…from a few seconds to a few hours…(Line 149)” are not enough).
• Section 2.3 contains the list of ionosphere affecting parameters, but there is no an analytical, or empirical model which tide the above mentioned parameters and scintillations (linear or non-linear dependence). In my opinion, the authors should enhance the section with such a model and explanations how the model is trained with artificial network and genetic algorithm.
• Section 3: this section contains mostly general description of the ANN, GA and its combination but there is luck of explanations about the very methodology of scintillations prediction based on the ionospheric parameters and the scintillations forecasting model (see comments above).
• Line 345: replace “The” with “the”
• Lines 350-351: “….The RINEX data received from this station with the observation rate of 30seconds….”. Such a time span mostly does not allow observing pure scintillations since the scintillations “period” is mostly less than 1 sec (see Demyanov et al, 2019 b, Bhattacharrya, A., et al. ,2000, McCaffrey and Jayachandran, 2017). Please explain why such a long time span was involved for your research? If frankly, I doubt about the results of this paper, because most of ROTI or S4 event the authors observed probably does not relate to the pure scintillations.
• Line 353: “This section describes the modeling process of an ANN integrated with a GA.” First, I did not see modelling process description neither in this section nor in the manuscript et all (see my comments above). Please explain, what you mean here?
• Fig 2 and 3 need of being explained about the accordance with the known appearance of scintillations. It is well known that scintillations peak mostly happen within sunset and post sunset local time. Most of the pictures of Fig.2 does not match to this theory. Why?

 

 

Minor comments

• Line 44: “15-20 degrees from the equator…” add “magnetic” equator
• 1 is wrong, check it
• Line 197: “…? denotes the rate of ion reconstruction…”: replace “reconstruction” with “recombination”
• Please enhance References with following

1. V. Demyanov, Yu. V. Yasyukevich, A. B. Ishin and E. I. Astafyeva Ionospheric super-bubble effects on the GPS positioning relative to the orientation of signal path and geomagnetic field direction (2012). GPS Solut (2012) 16:181–189. DOI 10.1007/s10291-011-0217-9.
2. Anderson PC, Straus PR (2005) Magnetic field orientation control of GPS occultation observations of equatorial scintillation. Geophys Res Lett 32:L21107
3. Ma G, Maruyama T (2006) A super bubble detected by dense GPS network at east Asian longitudes. Geophys Res Lett 33:L21103
4. Tiwari, R., & Strangeways, H. J. (2015). Regionally based alarm index to mitigate ionospheric scintillation effects for GNSS receivers. Space Weather, 13, 72–85. https://doi.org/10.1002/2014SW001115.
5. Demyanov V.V., Sergeeva M.A. and Yasyukevich A.S. (2019a) GNSS High-Rate Data and Efficiency of Ionospheric Scintillation Indices. Ionospheric and Atmospheric Threats for GNSS and Satellite Telecommunications., Prof. Vladislav Demyanov (Ed.), ISBN: ISBN 978-1-78985-996-6, InTech. http://dx.doi.org/10.5772/intechopen.90078
6. Demyanov V.V., Yasyukevich Yu.V., Jin S., and Sergeeva M.A. (2019b) The Second-Order Derivative of GPS Carrier Phase as a Promising Means for Ionospheric Scintillation Research // Pure Appl. Geophys. Springer Nature Switzerland AG. https://doi.org/10.1007/s00024-019-02281-6
7. Bhattacharrya, A., et al. (2000). Nighttime equatorial ionosphere: GPS scintillations and differential carrier phase fluctuations. Radio Science, 35(1), 209–224.
8. McCaffrey, A. M., & Jayachandran, P. T. (2017). Spectral characteristics of auroral region scintillation using 100 Hz sampling. GPS Solutions, 21, 1883–1894. https://doi.org/10.1007/s10291-017-0664-z.

Comments for author File: Comments.docx

Author Response

Thank you for the valuable comments.

"Please see the attachment."

Author Response File: Author Response.docx

Reviewer 2 Report

See attached file 

Comments for author File: Comments.pdf

Author Response

Thank you for the valuable comments.

"Please see the attachment."

Author Response File: Author Response.docx

Reviewer 3 Report

This study proposes a method to predict the ionospheric scintillations combining an Artificial Neural Network (ANN) with a Genetic Algorithm (GA). Its main contributions consist in using a method able to predict daily ionospheric scintillations based on the prediction of Signal to Noise Ratio and Rate of TEC Index.

The document is easy to read and follow.

The English is in general clear

The document is well supported with references.

The subject of the paper is not new with an average potential of application.

Authors should consider reviewing section 3. The description of theory behind ANN and GA is very detailed but lacks detailed description about the modeling of the ANN and GA essential for the reproducibility of the presented work. Authors should also include schematics (block diagram for example) supporting the text explanation.

One of the weaknesses of this study is that it presents low accuracy when strong ionospheric scintillations occurred which in fact is when the method is expected to perform more robustly for the work to be more relevant. This might be due to not ideal parameter choice or not well enough parameter representation of the phenomenon or ultimately due to instability of the proposed method.

Another limitation of this study is the lack of comparison with other methods from literature.

Before line 60 of the Introduction section authors use the “present” but in the phrase in line 60 use “past tense”. Please correct this.

In line 107 please correct “The model was proposed by”.

In line 138 please correct “using for the detection…”

In line 139 did you meant “.” In “The S4 value” ??

In line 150 and 151 and also 168 and 169 authors say that “the average values over a period of five-minutes were used”. It is not clear from the text why it is used the five-minute period and not other. Please explain.

From line 175 to 179 authors should explain how the “parameters affecting the ionosphere” were obtained.

In line 184 and 185 authors say they “tried to select a month from each of the seasons… predicting for a thirty-day period”. Please explain the need for data from each season and why thirty-day period.

In line 242 please correct “that using for performance”.

In line 272 please correct “will replace by their parents”.

In line 345 “This The data” ??

The explanation starting in line 374 should be included earlier when was first approached In the text (line 150 and 151).

Tables are inserted in the text before referred. Please correct.

Author Response

Thank you for the valuable comments.

"Please see the attachment."

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

No additional comments

Reviewer 2 Report

All comments have been modified.