THe manuscript is almost ready for publication, however it would benefit from a minor revision, including  improving the english.

Authors should take advantage of the figure captions by fully explaining what is in the figure and what is the reader supose to look for. In the current form most of figure captions are to short to fully follow the authors reasoning. 

Some minor coments are also included in the text attached. 

This paper presents a result of seismic studies of iron oxide-apatite mineralisation in Kiruna mine area of Northern Sweden. The paper is well structured and describes 2D and 3D surface seismic studies, downhole seismic data recorded with DAS and key rock physics properties defining the host and ore formations.

Limitations in seismic acquisition and poor seismic data quality, that is typical for crystalline environment, resulted in fragmented subsurface picture. Nevertheless, the integration and interpretation section of the studied geophysical data presents convincing analysis of iron mineralisation. Although, there is a great room for alternative interpretations.

Overall, I would recommend this paper for publication but would like to see more detailed comparison between the three types of seismic data at the pre-image level. One of the important decisions to make before any further seismic exploration here or in similar geological environments is the trade-off between the value and the cost of the presented seismic methods.

Assessment of the 2D, 3D and VSP gathers from close by locations, their respective frequency spectra and signal-to-noise ratio at raw and pre-processed stages will be helpful in evaluation of the significance of each dataset.

Some minor comments and suggestions:

Line 51: typo: accelerated wight-drop – should be - accelerated weight-drop

Line 120: typo: source have been used - source has been used

Line 141: However, as both holes were waterfilled the coupling should have been adequate.

DAS is sensitive to axial strain, so coupling via water is not adequate for shear contact of the cable with the media.

Figure 4: add arrows pointing at the reflections discussed in the text

Also, this figure is a good place to add illustrations of the 2D, 3D and VSP gathers from close by locations, their respective frequency spectra and signal-to-noise ratio at raw and pre-processed stages.

Line 181: more moderately dipping or sub-horizontal along Profiles 1 and 3 (Fig. 5) - there is no profile 3 in Fig 5

Figure 5: it will be more convenient for readers to see West on the left and East on the right side for profile 2 - to be consistent with the map.

Line 186: Note that the fold is varying much along Profile 2, which is clearly noticeable in the stacked time section (Fig. 5). - no CMP fold presented in the figure or elsewhere – need to add if refereeing to it in the text.

Also, as Profie2 is bent, the image deficiency might be due to the geometry and, hence, it is good to see CDP location at the bending point or schematic plot of the profile XY geometry on top pf the section to have a reference.

198: The reflection at about distance 4 km along the profile, and at a depth 1.5 km,

coincides with the Per Geijer Deep mineralization – indicate by arrows in the figure.

Line 217: All surface seismic data have been depth-converted using a constant velocity of 6000 m/s.

Why velocity used for surface seismic is 6000m/s here, but it was 5000m/s for the downhole processing (Line 243)?

Line 243: A velocity of 5000 m/s was chosen for the migration, and the inherent depth conversion.

Figure 7: East-West indications are missing; similar to profile 2 in Fig 5, the section should be flipped to be consistent with the map orientation: West - left, East – right.

Figure 8: West-East orientation is good here, should be the same in Fig 5 and Fig 7.

It would be good to have an example of inline from migrated 3D depth volume coinciding (or located as close as possible) with profile 2 for one-to-one comparison of the 2D and 3D products.

Figure 13A: indicate reflection from the ‘2.40-1.96 Ga basalt-andesite units’

Show amplitude spectra for all datasets 2D, 3D, VSP for raw and pre-processed, just before stack (or before migration)

Appendix:

Bandpass filter: 10-20-150-220 Hz

Bandpass filter: 6-12-130-200 Hz

Bandpass filter: 7-14-50-80 H - why frequency content is so different for 3D processing post-stack in comparison to the other two?