We thank the referee for insightful comments and suggestions. We have responded to all of them, as detailed below. The modified text is shown in blue. > The authors present a new method to study helicity in galaxies which may > be important to compare predictions of dynamo models with observations. To > be understandable for non-experts, additional explanations are recommended > (see comments below). We have used this opportunity to improve the paper in various places to make it clearer. We have therefore expanded the introduction and have moved the text around Eqs. (1) and (2) to a new section. The following section numbers are therefore shifted by one. > General comment: > - Requesting future polarization observations in the submm range > is not the best strategy because there is little dust in galaxy > halos that could trace magnetic fields at large heights. Radio > synchrotron observations are much better suited. VLA and Effelsberg > polarization observations are available for NGC891 and other galaxies (see > https://ui.adsabs.harvard.edu/#abs/2019Galax...7...54K for a review). New > radio polarization observations were performed with the EVLA for a sample > of edge-on galaxies (project CHANG-ES, paper submitted to A&A). Galaxies > with large-scale patterns in Faraday rotation measures provide the best > chances to measure large-scale patterns of magnetic helicity. These > points should be addressed in the revised version of the paper. It is true that dust grains are mostly spread over the disk. However, dust polarization and cm-wavelength synchrotron observations are dealing with different types of gas. The former traces the cold ISM, while the latter the ionized hot gas. Measurements of cold ISM has two advantages. One is that we can assess how the magnetic field regulates the evolution of the ISM which is natal gas of stars. This cannot be achieved by cm radio imaging. Another is that 850 um seems to be optically thin. We have now alluded to this at the ends of the first and penultimate paragraphs of the introduction. > Detailed comments: > - Section 1, para 2: Please explain whether polarization observations > (suffering from the 180° ambiguity) are sufficient to measure field > helicity, or whether additional Faraday rotation data (resolving the > ambiguity) would be helpful or are even needed. We emphasize that we are here working with a helicity proxy and not with the actual magnetic helicity itself. Not much is known about this proxy, which is why we do this study. That Faraday rotation is not invoked is now stated explicitly in paragraph 6 of the introduction, starting with "Let us emphasize at this point..." > - Para 3: Please explain whether the definition of "parity" with respect > to E and B modes is identical to that used to describe the large-scale > pattern of dynamo-generated fields (i.e. odd parity = dipolar, even > parity = quadrupolar). It is not the same. This is now explained this in paragraph 4 of the introduction, starting with "In dynamo theory, one ..." > - Para 4: Please mention that "inward spiraling" and "outward spiraling" > has nothing to do with spiral arms in galaxies. This is now stated explicitly at the end of paragraph 3 of the newly arranged Section 2, starting with "The E polarization corresponds ..." > - Please explain what "inhomogeneous system" means with respect to > galaxies. This is now explained in paragraph 5 of the introduction, starting with "Before we define ..." > - Para 6: "positive E and B in the north" - hence negative E and B in > the south? Does this mean "even parity"? No; this is now stated explicitly at the end of the last paragraph of the newly arranged Section 2. > - Section 2, para 1: As only the (x, z) plane is considered, this > model is 2-D, assuming axisymmetry (constant b_y). Please mention. No, it is 3-D. This is now stated explicitly in the second sentence of Section 3. We also now show a 3-D visualization in Figure 1. > The presented dynamo model is a rather simple one. Other models (e.g. by > Moss, Shukurov, or Chamandy) should be mentioned. For the aims of this > paper, it should also be mentioned that disk and halo of galaxies > may drive different dynamos (see section 8.2 in Beck et al. 2019, > https://ui.adsabs.harvard.edu/#abs/2019Galax...8....4B). These and other models are now mentioned in the last paragraph of the introduction. > - Para 2: Please specify the assumed properties of the galactic > wind. A wind can support but also suppress dynamo action (e.g. Chamandy > et al. 2015, https://ui.adsabs.harvard.edu/#abs/2015MNRAS.446L...6C). This is now also stated in the last paragraph of the introduction, where we write "The wind can both ...". > - Line after Eq.(5): Please express \Omega_0 in units of Gyr^-1, > as in Table 2. We have now omitted the numerical value here, because it is given in Table 2 and is different for different models. > - A sharp transition from rigid to constant rotation is unrealistic. Is > this assumption needed for numerical reasons? We now state "This is not very important and, in hindsight, could have been avoided." > - Fig. 1+2: Please add a figure where the predicted polarized emission > is shown in the usual way, as intensity + angle. We have done this now for Figure 2. > - Fig.3: The statement that the vertical field in the model is "almost > completely negligible" is in sharp contrast to the observations of > edge-on galaxies. Please clarify. As shown in earlier work, this is most likely because we have not included a galactic wind in our models; see the discussion in the last paragraph of the introduction. > - Para after Eq.(6): An equipartition strength of the total (mostly > turbulent) field of 5 muG is too small compared to observations (see > Table 3 of Beck et al. 2019). We now refer to Eq.(1) of Beck+19. There could be a misunderstanding with respect to another equipartition value having to do with the cosmic ray energy density. We have now mentioned the range of observed field strengths in galaxies in paragraph 5 of Sect.3 starting with "We also define ...". > - Table 2: Please mention why NGC891 was chosen for model B. At the end of the 7th paragraph of the introduction, we now mention NGC 891 and NGC 4631, and say that "The former may soon become accessible to observations of dust polarization with the James Clerk Maxwell Telescope ..." > - Section 3, para 3: Most galaxies indeed have even-parity symmetry > (see Table 2 in Beck et al. 2019), but the observed large-scale field > pattern is not that of a simple quadrupole. Please replace "quadrupolar > symmetry" by "even symmetry" to avoid misunderstanding. We have now replaced these occurrences by "even symmetry" in two cases. > - To my understanding, both the azimuthal and the vertical components > of an even-parity ("quadrupolar") field are symmetric with respect to > the midplane. No, the vertical field vanishes for a quadrupolar field in the midplane, so it is antisymmetric. > - Section 4, para 1: Please clarify (again) that "spiraling patterns" > have nothing to do with spiral arms. We have added now a corresponding sentence at the end of the first paragraph of what is now Section 5. > - Para 2: Dust polarization cannot measure field direction, only > orientation. Right, we have now replaced "direction of the magnetic field" by orientation of the magnetic field. > - Para 2+3: Please mention HAWC+, a new far-infrared > polarimetric instrument on board of SOFIA (e.g. Jones et al. 2019, > https://ui.adsabs.harvard.edu/#abs/2019ApJ...870L...9J). On the other > hand, radio synchrotron observations are more promising to measure > field patterns in the halos of edge-on galaxies (Stein et al. 2019, > https://ui.adsabs.harvard.edu/#abs/2019A&A...623A..33S > and Mora-Partiarroyo et al. 2019, > https://ui.adsabs.harvard.edu/#abs/2019A&A...632A..11M). We have now added an extended discussion on this in paragraph 3 of Section 5.