Required Corrections

• Add to chapter 1 a short discussion, possibly with figure, of solar wind speed as a function of radial distance. Some of the references in this review may be a useful starting point: https://oxfordre.com/physics/view/10.1093/acrefore/9780190871994.001.0001/acrefore-9780190871994-e-19

• Add to chapter 1 a very short summary of fast and slow wind properties

• Chapters 3-5: When discussing correlation, be explicit about what variables are being correlated, particular in figure captions and tables.

• When discussing significance of correlation coefficients, acknowledge the issue of auto-correlation and that the “independent” N will be less than the actual N. (Thus quoted significance levels are likely upper limits.).

• Chapter 4: Describe what you were responsible for in this study and what was done by your collaborators?

• The appendices are related to the study and should be put into the main body of the text, where they are relevant.

Optional, Complementary Corrections

• Title: You probably don’t need a full stop at the end of the thesis title.

• Chapter 1: Consider slightly trimming some of the less directly relevant material in chapter 1. (The litmus test: ”Does the reader need to know this in order to understand the work in the subsequent chapters?”)

• Chapter 1:State reference frame for 27-day rotation rate (i.e. in frame of reference of Earth).

• 635: Make sure you define all of your letters both here and throughout the thesis.

• 655: Define T.

• 680: Why are you citing a 2007 paper for Ohm’s law?

• 710: Explain why have you cited Zyla+?

• 755: There are a number of astrophysical plasma environments that are collisional.  What about the Sun?  What about the magnetosphere?  Change the statement.

• 800: Define rho_q.  Also, explain how the last two terms are linked to the Lorentz force.

• 865: You could describe the special point in the solar wind related to the Alfven speed.

• 895: 1 AU is not 200 solar radii, this needs correcting.  You also have km, Mm and R_s in the same paragraph?

• 940: There are better figures that display the solar atmosphere which show the electron number density and/or more of the coronal structure.

• 1020: How small is the salt and pepper pattern.

• 1040: Explain what you mean by generally 3d, as the universe is completely 3d in space.

• 1045: Interchange reconnection is not simple.  Better to remove this comment.

• 1110: CMEs are not always dependent on flares.  Corrent

• 1170: You could include the derivation for the angle of the magnetic field at a given radius.

• 1195: Define the speed of the fast and slow solar wind.

• Chapter 2: optical thickness is not only limit to measuring coronal magnetic fields. Thermal broadening of spectral lines inhibits the use of Zeeman effect.

• 1535: Mention what is the cadence of HMI.

• 1550: It would be useful to have all of the curves on a single plot for comparison.  Additionally, the significance of the different response functions could be explained in the text.

• 1655: Equation 2.1 does not take into account transport effects.  Therefore there is a huge uncertainty in the estimation of delta_t.  In particular the acceleration of the solar wind will cause a large variation.  This should refer back to the new introduction section on solar wind speed.

• 1670: Describe how successful has this backmapping been in the past.

• 1685: Explain how the minimum and maximum points are found.

• 1700: Explain what defines this pre-determined value.

• 1730: Explain how this is done.

• 1770: Which imagers are discussed.

• 1790: These simple empirical models are, however, more reliant on output verification to provide realistic information, as they are governed by over-simplified kinematics.” What do you mean by this? Why would simpler techniques need more verification and why would more verification provide more realistic results? [In fact, empirical methods often out-perform MHD. E.g. WSA versus MAS.]

• Line 1910: When generating the synthetic data, things like autocorrelation in each signal will have a big effect on the correlation between two variables. How was this handled? E.g. is the use of white noise a good proxy for the observational data sets that you’re using? Solar wind is not white noise for any of the MHD parameters

• 1805: Quantify what is meant by significant.

• 1825: Explain what is causing these bright point eruptions and what you mean by bright point?

• 1935: Clarify what you mean by duration.  Is this the standard deviation of the Gaussian?

• 2050: Is the Parker spiral the biggest mapping effect between low corona and 1 AU? What about coronal field topology (e.g. through PFSS)?

• 2095: The 193 A signal is not density information, let alone related to the other quantities.  Describe how the EUV signal should be related to the measured physical quantities.

• Figure 3.10: Given IMF 6 is very long time period, presumably significance of a particular pearson R value is lower than for, e.g. IMF 1, where there are more independent data points?

• 2130: Do you mean 250x250 arcsecs or 50x50 arcsecs?

• 2120: Figure 3.7 has no axis labels.

• 2195: Mention what the speed of the jet is

• 2260: The suggested solar wind travel time. Does the Neugebauer 4/3 factor imply a gentle or impulsive solar wind acceleration? How does the best-fit travel time compare with various proposed acceleration profiles? Allan Macneil had a nice paper on this and your work seems really complementary (should you ever be looking for another space physics project to take on…)

• 2265: What is the jet travel time.  Also, consider analysing the signal before the event.

• 2420: Define how the expanding closed coronal loops can join the solar wind.

• 2440: Why did you choose 2.5 solar radii?

• 2482: Tracking the differential rotation of the solar corona? What is this and how/why is it done?

• Figure 4.3: What are two lines in RH panels?

• 2540: “The horizontal for all panels” seems a misprint.

• 2558: Why is solar wind mass density a proxy for volume of coronal plasma release?

• 2605: If PFSS does not capture open flux from active regions, why should it be trusted for coronal holes? i.e. wouldn’t the fact that a major source of open flux is missed mean all mapping is suspect? Would NLFF really improve things? Is that practical?

• 2730: Explain that the response function can detect the low temperatures of the solar holes.

• 2830: What limits the volume of data that has been considered? I.e. why only very short case studies?

• 2960: Why is longitudinal alignment prioritised over latitude? Small changes in latitude can mean completely different solar wind streams are encountered, whereas longitudinal offset means a corotation delay to see the same solar wind stream, though not identical plasma parcels.

• 2990: A travel time of 7 days over 0.8 AU? That doesn’t seem plausible?

• 3030: Are BR and nP data z-score normalised? Why not scale by r^2, which is more physically meaningful?

• 3080: Discuss what the cross-correlation of the variables provides as, other than the radial field,  they show quite a connection and even the radial field looks similar.

• Figure 4.10: Colour bars are actually correlation?

• Figure 5.1: Would be useful to say where spacecraft are at this time.

• Figure 5.2: What lat plane is this in? Same for Figure 5.5?

• 3095: Why switched from BR to BT? If parameters have been selected to find best matches, significance has been reduced.  (Also B_T was previously used for total B, so it’s a bit confusing)

• 3175: If the B_R reversals are crossings of the HCS, are they really switchbacks? One is a spatial structure, the other temporal?

• 3195: Mention why the STEREO-A not observe the switchbacks?

• 3365: Yes, a good extension could be to see what one might predict from the successful results and to develop the algorithm.