We thank the referee for a thorough and timely report on our manuscript entitled "Recollimation Boundary Layers as X-ray Sources in Young Stellar Jets". We have taken all the referee's suggestions into account when we revised our work. We have significantly expanded the article and reorganized several sections. In particular we expanded the introduction to add a more thorough review of the observational properties of jets from young stellar objects and to discuss previous models in the literature in more detail to highlight where this work contributes new results. We have also added a numerical fit of our model to the X-ray spectrum of DG Tau and show fit results. We hope that this makes it clear that our model is really "compatible" with the observed data. While the numerical values we discuss are geared specifically to DG Tau, we believe that our model might have a much broader relevance, but for stars other than DG Tau this is a purely theoretical investigation because we are lacking the required observations. In our opinion, this is still a worthwhile proposition to discuss, and we have tried to separate this more clearly from the discussion of DG Tau, which is based on direct observational data. Last, we have changed numerous wordings and grammatical structures in an attempt to make the article better understandable as suggested by the referee. Unfortunately, that makes is not very practical to highlight all passages of the article that have changed in bold font. If required, we can supply such a version, but essentially the entire sections 1,4, 5, and 6 and large parts of section 3 would appear bold. Specific response to points raised by the referee ------------------------------------------------- (in different order then in the original report) >>> -I first suggest to specify what is new with respect to previous models and >>> the reason why it is needed to use a different approach with respect to >>> previous models to improve the impact of the paper [References to previous >>> models have been added only briefly in the second version of the >>> manuscript uploaded] >>> "confirm or rule-out the scenario we suggest in this article": again, a >>> comparison with previous models should be added to stress what is new in >>> the present manuscript and why it is necessary to propose a different >>> approach to handle the same topic We have reorganized the introduction of our manuscript and added a new subsection where we review previous models in the literature and we point out where our work expands on other models or supplements a previous analysis. We hope that this clarifies the purpose of our calculations: Our goal in this article is to see how well a recollimation shock of the stellar wind can explain the X-ray observations from CTTs, specifically DG Tau. Other models in the literature explain the X-rays by other mechanisms, but they may not be the only possible explanation. It is important to study every promising scenario in some detail to decide if it is compatible with observations or not. In this work, we want to decide if a recollimation shock is a viable model. Additional references or remarks connecting our work to previously published scenarios have been added at different points in the manuscript. >>> - The authors do not take into account the role of the magnetic field in >>> their model. Cabrit et al. 2007 show that the magnetic field is necessary >>> to collimate jets as pressure of the ambient is not enough; Kohler et al. >>> 2012 discuss that even if their model does not take into account the >>> magnetic field, it should be considered and explain how their model can be >>> used to improve MHD simulations or complete future models with the magnetic >>> field as also made in their subsequent paper Kohler et >>> al. 2012b. Therefore, the authors should properly explain why the magnetic >>> field has been neglected in this case and discuss if it is actually >>> negligible >>> - The model presented by the authors is an application of Kohler et al. >>> 2012 model. Note that Kohler et al. 2012b also extend their results for the >>> MHD case (see also previous comment) >>> "due to the shape of boundary": collimation not due to ambient but a >>> magnetic field is required (Cabrit et al. 2007) >>> "the influence of the magnetic field probably dominates": so why it is >>> justified the approach without the magnetic field taken into account? It >>> should be explained if it is negligible or why it is neglected (see also >>> Kohler et al. 2012b and Cabrit et al. 2007) >>> "can confine...": Cabrit et al. 2007 shows that the magnetic field is >>> needed to collimate jets, no collimation is possible just with pressure There are two very different regions where the magnetic field might play a role in our model: As an outer boundary condition and in the stellar wind. As an outer boundary for the stellar wind, we prescribe some external pressure profile P(z). For our work it is unimportant which force provides this outer boundary condition. We realize that the last version of the manuscript may have caused the impression that this P(z) should be a thermodynamic pressure but in fact any pressure is fine for our model. The external pressure might just as well be supplied by an external magnetic field or by some combination of a magnetic field and a thermodynamic pressure. We have added this statement to the article. The article by Cabrit et al (2007) mentioned by the referee looks at some gas flow and estimates an external pressure. The authors conclude that in their case the external gas pressure is too low to explain the jet collimation. This is an interesting result. However, Cabrit et al look at a different regime than we discuss here. First, their PdBI observations deal with the jet about 500-1000 AU from the central source, which is at least an order of magnitude larger than the size of the recollimation shock that we discuss in our manuscript; second, they observe in molecular lines. If the jet still has an onion like structure with several layers at those large scales (which might not be the case) then the molecular gas should be an outer layer. We describe a much smaller component, that is collimated by an outer layer of optically emitting gas, which in turn is collimated by an outer layer of molecular gas which (according to Cabrit at al) is collimated by a magnetic field. Nevertheless, the results of Cabrit et al. serve as an interesting example how the collimating force can actually be constrained. The second region where we need to discuss magnetic fields is the stellar wind that is INSIDE of the shock front. Here indeed we would have to treat magnetic fields explicitly if they are important for the dynamics. Unfortunately, we cannot follow the example of Koehler et al (2012b) in this case. In that article the authors rely on asymptotic magnetic dominance and look for self-similar solutions at large radii. In contrast, we analyze a shock front that bends back towards the axis of symmetry within a few tens of AU. A simple numerical estimate (added to the paper) indicates that the ram pressure in this region is higher than the magnetic pressure and thus it dominates the dynamics, analogous to the situation in the solar wind, where the pressure of the solar wind opens some of the field lines. While we agree that it would be very interesting to study the influence of a magnetic field, the complex topologies generally observed in CTTS (see papers by Donati et al and reconstruction by Jardine et al) are beyond what can be done in the semi-analytical treatment that we present here. >>>- The work is focused on properties derived only for one object (DG Tau) and >>> using the Kompaneets approximation (relaxed in the analogous paper Kohler >>> et al. 2012 which show that this approximation affects the morphology). >>> Therefore, the strength of the results could be improved if a more general >>> discussion valid for other objects and in other approximations is added >>> "best studied case": in the manuscript the fact that this object should be >>> the best case is just stated and the reason why it should be considered the >>> best case is never explained. It seems like the reason should be the fact >>> that there are multiple observations with Chandra, but this is true also >>> for other jets for which multi-epoch and multi-lambda observations are >>> available. >>> Please explain why DG Tau should be considered the best case or remove this >>> sentence and in the rest of the paper DG Tau is the best case for such a study for two reasons: 1) There are more X-ray observations than for any other X-ray jet from a young stellar object. DG Tau was observed in 2004, 2005, 2006 and 2010 by Chandra and in 2004 and 2012 by XMM-Newton. The next best case is probably HH 154 with three Chandra and two XMM-Newton observations, but that source of HH 154 is a binary, making the interpretation more complicated. All other YSO jet are observed only once or twice in X-rays. 2) In DG Tau we have the fortunate situation that the central star is absorbed, such that only the hard X-rays escape. Thus, we know for sure that the soft X-rays seen close the central star must come from the inner jet. In all other cases the central region is either absorbed so that star AND jet are hidden (e.g. HH 2) or BOTH are visible (e.g. HD 163296) and since the star is much brighter than the inner jet, we cannot separate stellar soft component and inner jet emission. In our extended and reorganized introduction we added a few sentences to explain this. We believe that DG Tau really is a proto-type for X-ray emission from jets and that all that we deduce in the article is equally applicable to jets from other CTTS and HAeBes and most likely class I sources, too. It is just these fortunate observational circumstances that make DG Tau special. Indeed, there are indications for an equally stationary component in HH 154 and HD 163296; it is not necessarily the jet physics, just the geometry of the absorption, that makes is harder to analyze a stationary X-ray component in those sources, too. Thus, while the observational evidence for stationary components in sources other than DG Tau is weak, there also is no observational evidence to the contrary. We see the X-ray spectrum at 30 AU from DG Tau. Should we not expect that other CTTS have similar properties, even if we cannot observe them? We realize that this is speculative, but if DG Tau is powered by a stellar wind shock, there is no reason to believe that stellar wind shock might not occur in other sources, too. We have extended the discussion on this point significantly and we hope that the revised version now addresses the referee's concerns. >>> It seems that the proposed scenario fits better for C IV than for X-rays. >>> Therefore, maybe the structure and also the title of the manuscript could >>> be changed properly. If even the stellar wind is too slow to explain the X-ray observations, then the disk wind is certainly also too slow. In other word, if a recollimation shock cannot explain the X-rays, then we need to invoke some other non-local (scattering) or non-thermal (magnetic heating) model. It is the goal of this article to see how far we can get with the recollimation model and to explain the highest energies observed. Although we did not give X-ray luminosities explicitly in the last version of our manuscript we knew from a comparison with our work in Guenther et al (2009) that the velocities and the mass flux of our fiducial model will produce a significant amount of X-rays. The figures of the mass-fraction-per-bin vs. T_postshock can be misleading in this regard - for mass losses of the order 10^-8 M_sun/yr only a very small fraction of the gas needs to be heated to 1.2 or 2 MK to emit enough X-rays (see Guenther et al 2009). We have added a new section 4.3 that deals with this point. >>> "luminosity and temperature": add values or range of values The luminosity changes by about 1.6 (Schneider et al., in prep), the 25% we quoted are for the temperature and we have now added the range of observed temperatures to the sentence. >>> "stationary nature...remained unexplained": different explanations have >>> been proposed in the literature; please check and discuss previous models >>> on the same topic We wrote "remains unexplained in this scenario" with the intended meaning "unexplained in the scenario of paper~I". We have changed the wording of the paragraph to make it more obvious that we do not claim that stationarity is unexplained in the literature (there are several scenarios by different authors, some of which were pointed out by the referee), but just that it is not explained in paper I. >>> "we explain": a discussion on previous explanations should be added as well >>> as an explanation of the reason why in this case (proper just for DG Tau? >>> Valid also for other jets?) it is necessary to consider a different approach In our restructured introduction we have added an entire subsection dedicated to discuss previous models in the literature, their advantages and disadvantages and how our analysis supplements and expands on those previous models. >>> "strong shock": only stated here, but not explained why this approximation >>> is valid in this case We have added a new section that justifies the main assumption made in the model (magnetic fields, sound speed, initial wind temperature). >>> "Matsakos...in apparent contrast to Kompaneet's approximation": also >>> Kohler et al. 2012 relax this approximation; explain why this is not made >>> in this case or explore also this scenario Kompaneet's approximation is clearly violated at large distances from the jet axis since the pressure beyond the outer edge of the disk falls to the much lower values of the star forming cloud. In our model, we deal with much smaller scales - scales so small that they all fall within the innermost resolution element of the Matsakos et al simulations. We have use the word "apparent contrast" here to highlight that this contrast is only "apparent" - the simulations do not tell us if the approximation is violated at the relevant scales. We have rewritten this paragraph to address the referee's concerns. >>> "so narrow...resolution elements": how much narrow with respect to the >>> resolution? We replaced this with "smaller than ... for the fiducial parameters". If we vary the parameters we can also obtain solutions with a much larger radius, but we know that those are excluded by observations anyway since we do not see an hole filled with stellar wind within the disk wind. >>> "typical densities...1.e6 cm-3": Podio et al. 2006 find lower values for >>> the jet densities, these are very high Our model deals with the innermost component of the jet up to 50 AU from the central star. In contrast, Podio et al. measure the density several hundred AUs from the central source. Even for well collimated jets with small opening angles we expect lower densities at such large distances. We have added a discussion of these points in section 1.1. >>> "60 AU": should it be 40 AU as in Fig. 3? The figure is zoomed in to show the shape of shock front for smaller mass loss rates. The maximal extend of the high mass-loss model is at 60 AU, although this happens outside the figure (we mention this number in the text, because the reader cannot see that him/herself in the figure). >>> "Fortunately": is this result expected? Please explain if it is expected >>> or not For small radii the ram pressure is high, so we expect a large dw/dz and it is not surprising that we find this in the result. What we wanted to express using the word "Fortunately" is that the choice of omega_0 has only a small influence on the overall shape of the solution, which removes the need to do a detailed parameter study for different omega_0. We have deleted the word "Fortunately" from the sentence. >>> "fiducial model": it should be described better the reason why the >>> "fiducial" model is the best case In this context, we use the term "fiducial" in the sense that this is our "reference" model and we vary one parameter at the time from the values of this model. The parameter values are chosen form what we expect to find given the discussion in section 3 (e.g. the speed is taken to be just a little higher than the highest observed speeds in the optical component where we can measure them by Doppler shift). The "fiducial" model is not necessarily the model that provides the best fit (and we show later that a model with e.g. a higher velocity describes the X-ray spectrum from DG Tau better). >>> "temperatures...sufficient...power the observed X-ray...": in Fig. 6 T is >>> lower than 2 MK, a value lower than observed in X-ray emitting jets >>> "> 1 MK": but always < 2 MK which is lower than values typically derived >>> from the observations; just saying "< 1MK" is misleading and this is not >>> consistent with most X-ray emitting jets We think there are three arguments why it is still worthwhile to explore this temperature range: 1) The fiducial values are chosen as a "best-guess" of parameters based on measurements of the optical velocity etc. We should explore what X-ray gas we expect in this case. 2) Soft X-rays are absorbed more than hard X-rays. If the astrophysical source has a distribution of temperatures and some absorption, a fit with one temperature component (which is the preferred method in the literature given the low count numbers for stellar jets) tends to pick out a high temperature. So, it is not surprising that the average temperature in our model is lower than the fitted temperatures in the literature. 3) Very different temperatures are observed in different jets. In DG Tau (arguably the best comparison for the model we discuss here) the temperature is a little higher: Gudel 05,08 say 3.4 MK, but Guenther et al 09 show that there is a strong degeneracy in the parameters space with the absorption and values between 0.2-0.4 keV = 2-4 MK should be considered. In other objects, lower temperatures are routinely found: - HH2: kT = 0.06-0.11 keV (around 1 MK) - Pravdo et al 2001 or Schneider et al 2012 - HH 80/81: 0.13+-0.05 keV (=1.5 MK) - Pravdo et al 2004 - Z CMa: 0.2 keV = 2 MK - Stelzer et al 2009 - RY Tau: 0.1-0.2 keV - see discussion in Skinner et al. 2011 Notice that HH 154 is significantly hotter close to the launching source: 0.6 keV = 6 MK (Schneider et al 2011 analyze the temperature vs. the distance from the source). We have argued above that most likely the emission seen in these other sources is due to a different process, e.g. a termination shock or faster material in the jet catching up with slower one downstream as in the simulations of Bonito. Yet, some fraction of the cool material could also be due to a wind shock because the wind shock and the first Herbig-Haro knot might blend together at the spatial resolution of Chandra. The fact that we can reproduce those temperatures means that this scenario is at least plausible. >>> "cannot be resolved": HST can resolve shocks in Halpha and [SII] with >>> resolution comparable to numerical models for near YSO jets; does the >>>authors mean in X-rays? Please clarify as this is not the case in general HST is a spectacular instrument for resolving stellar jets and current IR instruments with AO on the VLT or Gemini also reach fantastic resolutions of the order 0.1". Yet, that is still insufficient to resolve the recollimation shocks we predict here, although new observations are getting very close to this limit and we are excited that this may provide a new test to verify or falsify our model in the future. The distance to the closest stellar jets (incl DG Tau) is about 140 pc. At that distance 0.1" provides a resolution of 14 AU. The maximal radius of our fiducial model is about 5 AU, the diameter is therefore 10 AU, comparable to the resolution limit. However, the contrast between a slightline that passes through the jet axis (this slightline passes: disk wind - stellar wind - disk wind) and a slightline that passes at omega = 5 AU (this sightline passes only through disk wind) is expected to be small, because the layer of the disk wind (which is common to both slightlines) is much thicker than the inner few AU of stellar wind. Together with the fact that the current resolution limit is still larger than the expected stellar wind zone in the center of the jet this means that we cannot expect to resolve the inner hole in the disk wind that is created by a small stellar wind zone such as predicted in our model. We have added a footnote to give a short version of this argument in the paper. >>> "stationary...consistent": consistent if the same time scale, temperature, >>> and luminosity are derived, but LX is not derived and T is lower than >>> observed >>> "shocks are major heating agents": [ ...] 3) if the comparison >>> with the observations is the aim, the authors should derive parameters to >>> be compared with the observations (T, L, size, location,...) >>> Second sentence: T are lower than observed, the position and the radial >>> extension of the shock front are in good agreement with DG Tau but are not >>> the proper for all the other X-ray emitting jets. A detailed comparison >>> with previous models and scenarios proposed to state the differences and >>> strength of the results presented should be added Paper I explores the mass flux that would be required to explain the X-ray emission with a shock. In the first version of our manuscript we tuned our fiducial parameters to match those mass fluxes. ("Thus, it is fully consistent that our model predicts a mass loss fraction larger than 10^−3 of the stellar wind at X-ray emitting temperatures."), so we know from paper I that this reproduces the correct L_X. However, L_X is easy to estimate from our solutions to the ODE, so we now give this number explicitly. We have gone one step further and added a new section 4.5 where we perform a numerical fit, that reproduces the observed Chandra spectrum from DG Tau and the observed distance between the star and the soft emission at the same time. The resulting fit is statistically acceptable and we added best-fit values to table 1 and show a figure with the fitted spectrum. Given the limitations of our model, especially the uncertainty in the external pressure profile, other parameter values might also provide a good description of the data (if a different pressure profile is chosen). In that sense the fit results do not provide a unique description of DG Tau. However, they certainly show that the model we suggest is compatible with the observed data. >>> "0.5 MK": maybe this approach fits better for lower T than X-rays >>> "CIV...": it seems that this scenario fits better CIV than X-rays (page 17) >>> , why is here just CIV discussed and before just X-rays? Maybe the >>> structure and also the title of the manuscript should be changed properly. The fitted model reproduces the X-ray spectra. As the referee suspected, the best-fit value for the wind velocity is quite a bit higher than our fiducial value (which is estimated based on the highest velocities observed at in proper motion and red-shift in the optical and FUV), but they are compatible with the speed of the fast solar wind. As stated above, our goal in this article is to explain the hottest emission. Of course, we hope our model predicts the C IV emission at the same time, but the simulation of a full position-velocity diagram is beyond the scope of this semi-analytical work. >>> "shocks are major heating agents": 1) if the main conclusion is that the >>> X-ray emission originates from shocks, there are several previous models >>> on this topic based on quantitative methods; 2) if the authors want to >>> explain the stationary shock in an alternative way than previous models, >>> these should be properly discussed [but they were totally missing in the >>> first version of the manuscript and only a brief sentence has been added >>> in the second version of the manuscript uploaded]) >>> "Our work is complementary to the simulations... several hundred AU from >>> the central star - about an order of magnitude further along the jet than >>> the scenario we discuss here": actually this is not the case of HH 154 >>> which shows X-ray emission located at the base of the central source as >>> well as DG Tau and in the literature mechanisms generating X-ray emission >>> located at the base of the jet has been investigated in details. The method >>> presented in this manuscript is valid only for one object as the other >>> jets show X-ray emission located also at greater distances. >>> "very close": actually the approach developed in this manuscript is >>> strongly tuned on only one object, as all the other X-ray emitting jets >>> discovered so far show X-ray emission located at the base of the jet up >>> to greater distances with respect to the central engine or very distant >>> from the YSO from which they originate. Also in the radial direction, >>> other jets have been resolved with higher radial extension. We have significantly expanded the introduction and the discussion section to address these points. We discuss previous publications in more details now, including those papers pointed out by the referee. We set out to test how well a recollimation shock can explain the observed features of jets in general. We pick the example of DG Tau, which we consider the best case for such a study (see above), but we have added significant discussion about the applicability to other sources. In short, we can resolve the X-ray spectrum of the component we need only in DG Tau. This does neither prove nor disprove the existence of recollimation shocks as X-ray sources in other jets, it is only a testimony to our limited observational capabilities. We regard it as likely, that DG Tau is not a special case and that other source might show similar phenomena, which (for different reasons) are unobservable so far. However, the current data also allows the interpretation that DG Tau is fundamentally different in some way from all other known young stars. We believe that our work is complementary to other proposed models in the sense that in this article we just explore one possibility to explain the X-ray generation. The fact that our model is compatible with the observations does not prove that it is right; other scenarios could also be compatible with the observations. Some alternative scenarios have been discussed in detail (e.g. a time-variable ejection speed in several papers by Raga et al and Bonito et al.), while for others only a first-order estimate of the plausibility has been published (e.g. Bally et al 03 estimate rough numbers for a scenario where the resolved stellar X-ray emission is in fact scattered stellar emission, but those estimates could be refined with numerical modeling). In this article, we simply explore one possible scenario that has not received much attention in the literature before and we hope that in future articles (by us or other groups) other scenarios will be studied to the same level of detail. >>> This figure is the same as in Kohler et al. 2012, therefore I suggest to >>> remove it in the manuscript and just refer to the original one. Furthermore, >>> the analogous figure in Kohler et al. 2012 is described in more details >>> and the authors should refer to that description We have added a statements that refer the reader to the excellent work of Koehler et al for more details both in the text and in the figure caption. However, we think that a figure like this is critical for the reader to follow the definition of the coordinate system and the angles. Unfortunately, very few readers from the star formation community will be familiar with the work of Koehler and al. and will remember how the coordinate system and alpha, theta etc. are defined there. Thus, we have a preference to keep this figure with a prominent statement that more details can be found in the Koehler et al. article. However, if the referee strongly disagrees then we will follow his/her advice and remove the figure.