General comments

This paper provides a meticulous analysis of selected ASMA-related trace gas in situ measurements taken by the HALO research aircraft. It is very well written and of excellent scientific quality.

Complementing previous campaigns, PHILEAS probed within the same season the western ASMA flank and also its eastern outflow. Three case studies are used to work out the differences between the respective measurements in detail. But are the selected cases really comparable? Measurements in the east probed a detached filament, i.e. ASMA outflow. The two western cases probed the edge and partly the interior of the ASMA, which may not be representative of the eastern ASMA outflow. Outflow via eddy shedding, for example, is likely to mix air masses across the transport barrier of the anticyclone.

The paper highlights the previously recognized interaction with tropical cyclones at the eastern ASMA flank (Vogel et al., 2014), which could lead to an asymmetry between eastern and western ASMA outflow. Gaining a general insight into the differences between the two types of outflow would require analyses that go beyond individual case studies. For instance, the extremely valuable ground truth provided by the case studies could be linked to satellite data or simulations, which provide longer temporal and wider spatial coverage, allowing for more general insights.

Correlations between methane and dichloromethane are shown to be sensitive to different source regions, providing a valuable tool for studying ASMA-related transport. On the other hand, it would be worth discussing the different chemical compositions in the context of emission inventories for these trace gases, too.

The above aspects (eastern vs. western outflow, quantification of ASMA-cyclone interactions, CH2Cl2) are examples that could provide novel insights into ASMA-related processes or inventories. However, each would require analyses/quantification beyond the interpretation of the measurements. In the current draft I am having difficulty identifying the scientific focus. What do we learn about the ASMA that is not already known from previous studies, some of which are discussed in the introduction?

Nevertheless, the current draft provides a valuable, in-depth analysis. It is certainly worth publishing. Rather than revising the paper as suggested above: Why not making it a measurement report? Then it would likely require only minor revisions. Given the interesting dataset and the impressive, elaborate analyses, it even has the potential to be a highlight paper.

Specific comments

L59: eastward -> westward?

L145: Is there an additional parameterisation for convection available (Konopka et al.), but was not used here? Or did Konopka et al. evaluate a setup without additional convection parameterisation? GENERAL: The representation of convection seems to be central for this study. If an alternative setup is available, a comparison would be warranted.

L147: Is this necessarily a lower limit? Assume that a similar amount of instabilities is occurring in any resolution and that instabilities are removed by vertical motion: Small scale convection in high-res, larger scale in low-res. The locations and trajectories of the vertical exchange might differ, but not the amount of exchange. This means that the surface tracers are not necessarily at a lower limit. Instead, the low resolution imposes an additional uncertainty on the origin of the tracers. Comparisons to observed convection (clouds, lightning) might provide an idea of model performance in this context.

Since this aspect of vertical transport seems to be crucial for the study, a detailed discussion or even a dedicated evaluation of this aspect in the model seems to be warranted.

L151: Are the gradients right for the right reasons? How sensitive are those evaluations to the representation of convective transport? See previous comment: An evaluation specifically for convection / the campaign period would increase confidence in the simulations.

L160: Back trajectories essentially serve the same purpose. Consider to add a few sentences explaining the purpose, the pros and cons of the two approaches. Forward simulations with source regions provide an impression of the air surrounding the measurements and they account for mixing. Do the back-trajectories also account for mixing?

L168: ASMA location and extent is very variable. Fixed surface regions neglect that. Why don’t use some dynamic criterion (e.g. back-trajectories from the actual ASMA) to flag tracers as "processed by ASMA"?

L225: How is that convection represented in ERA5?

L233: Consider to evaluate/discuss the representation of cyclones in ERA5 for cases or regions relevant in this study.

L255: “Air masses … monsoon region.” -> Consider to either provide some values (and references), or to omit the sentence.

L261-L269: The collection of facts about CH2Cl2 is somewhat inconsistent: Sources vs emission changes vs mixing ratios. How about showing or describing a snapshot of an emission inventory (if available)?

L296: Branch 3 has lowest CH2Cl2. Is this consistent with the main source in South Asia (as indicated in L262)?

L315: Consider to colour code the flight altitude in terms of potential temperature to indicate how it relates to the chosen isentropic level. A panel zoomed on the flight path might help visibility. In such detailed panel for the flight track you might also indicate the intervals used in figs 8, 11, 15.

Fig. 10: Consider to refer to fig. 9 for recurring descriptions of the annotation etc.

L409: Why outflow? South Asian air in F02 seems to be mainly within the ASMA (in interval 2).

L486: Case 1 also indicates that Pacific air might be transported around the outer edge of the ASMA. Then it is just a question of how this Pacific air is mixed with South Asian air during eddy shedding to the west. An outflow event at the western edge has not been analysed here?