| 27 Jul 2026
Quantifying the difference between methodologies for locating the Asian Summer Monsoon Anticyclone boundary and position
Abstract. Locating the boundary of the Asian summer monsoon anticyclone (ASMA) in the upper troposphere-lower stratosphere (UTLS) (350 K–410 K potential temperature) is important for analyzing the transport of pollutants by the ASMA into the stratosphere. The definition of the ASMA boundary highly impacts the information about the anticyclone's behavior and affects the results when studying its spatio-temporal variability in particular regarding inferred trends over the last decades. In this work, we quantify the differences between methodologies based on potential vorticity (PV) and the Montgomery streamfunction (MSF) used to determine the location of the ASMA boundary. In addition, we apply the unsupervised machine learning spectral clustering method which combines both the PV and MSF fields to derive the location of the boundary. By analyzing the centroid position and boundary geometry for July and August 2023, we demonstrate that at 380 K, the PV-based, MSF-based, and spectral clustering methods exhibit good agreement during the Asian summer monsoon season. We find that the MSF-based method tends to incorporate the eastward-shed eddy over the Pacific around 140° E into the main anticyclone, whereas the PV-based method tends to separate it from the main body. Spectral clustering, however, can exhibit either behavior during the season. Finally, we compare the derived boundaries with satellite measurements of carbon monoxide, showing that the PV-based and MSF-based methodologies successfully enclose the core of elevated carbon monoxide levels as effectively as the spectral clustering method.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Chemistry and Physics.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.
The manuscript presents a solid, well-structured study relevant for the dynamics of the upper troposphere and the lower stratosphere (UTLS). The paper is concise, written with clarity and has well-built figures. I also believe that the new methodology proposed can significantly enrich the set of analysis tools for the Asian summer anticyclone (ASMA).
I hence appreciate the general quality of the work and strongly encourage its publication, with some minor revisions meant to strengthen the discussion of the physics and methodological steps.
General Comments
Section 3: Although the methodological comparison is rigorous, the discussion lacks an assessment of its practical consequences. I feel that the section misses to address the significance of these boundary discrepancies: could the magnitude of these differences meaningfully impact quantitative estimates in transport, chemistry, and climatology studies, or do they turn out to be minor when in practice? This perspective would provide some context for the choice of methodology.
Section 4: The authors use satellite data of CO (MLS) averaged on a bimonthly basis (July-August) to validate the ASMA boundaries. However, due to the chemical "memory" of the tracer and the smudging effect caused by temporal averaging, rapid dynamical events (such as eddy shedding or daily synoptic variability) are completely lost. The CO cloud thus ends up representing a cumulative and fuzzy area, making the validation of dynamical boundaries that vary on the synoptic scale less stringent. I think a short discussion of this point would be important to add.
Specific Comments
Lines 79-80: I think it would be better to clarify what is meant by “graph partition problem”
Lines 109-111: Can you explain what criteria led to the choices of using 1 or 2 or 3 days for the time window?
Lines 160-163: In figure 1b there is also a noticeable discrepancy on the West side of the ASMA, why is it not discussed?
Lines 159-171: I think the readers would benefit of a discussion that goes beyond the sole discussion of the differences in the results of the methods, but also of the physical reasons why the different methods give differences in the contours in those outlier case studies.
Figure 4: I think that the figure would be more readable if the histogram were overlapped in transparency (or with simple 1-D lines) rather than stacked.
Figure 9: The colorbar is missing