This study employs ozone measurements from urban sites across 17 cities, a background site (Waliguan), four satellite products, and integrate two models (GEOS-Chem CTM and trajectory models) to analyze ozone variations over the TP region. The authors find a notable increase in ozone levels at urban stations, surpassing trends observed at Waliguan and those derived from satellite data. Analysis of model results and emission inventories suggests that this ozone rise is driven by increased local anthropogenic emissions and enhanced contributions from non-local sources. The study offers valuable new insights into understanding ozone changes in the region. The manuscript is well-written and structured. I recommend addressing the following points before publication:

Line 171: Please clarify the rationale behind selecting a 192-hour time frame for analysis.

Figure 2: Are all 30 sites consistently showing an increasing trend? If not, please elaborate on any variations.

Section 4.2: While generally reasonable, the discussions could be more informative. Why does the residence time show an increasing trend, and is it linked to changes in meteorological patterns? Could you present and discuss the non-local emission trends over TP, along with foreign source regions derived from the PHLET-OMI inventory? Additionally, since the PHLET-OMI only quantifies NOx emissions, and VOC emissions can differ significantly, do we have robust estimates of VOC emissions over TP and the surrounding source regions?

Line 341: The increase in NOx emissions is surprising, given that most regions in China have seen significant reductions in NOx emissions. Could the authors discuss possible reasons for this increase, possibly referencing relevant literature?

Lines 359-369: I am uncertain about the robustness of this method and its interpretation. First, it overlooks the non-linear interactions between meteorology, local, and non-local emissions. Second, Lines 364-367 assume that Waliguan and the 17 MEE cities share similar ozone formation mechanisms. Is this assumption reasonable? If the same method is applied individually to each of the 17 cities, would we obtain similar values for a and b for each city pair?

Publisher’s note: this comment was edited on 7 January 2025. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.

The manuscript presents a comprehensive analysis of ozone trends over the Tibetan Plateau, particularly highlighting urban areas, and successfully identifies the contributions of local emissions and regional transport to ozone changes, based on ground observation data, different satellite data, CTM model and HYSPLIT model. The authors concluded that the ozone trends at urban areas of TP were caused by a combination of local and non-local factors, especially from the local anthropogenic emissions. The multiple methods adopted in this study were appreciated. However, there were some logistic issues the authors need to address.

In this study, the authors did not use the CTM to quantify the contribution of non-local factors on ozone issues in TP but the HYSPLIT (the QNR method), quoting that there were large uncertainties for the VOCs (line 305-306). However, the QNR method also had uncertainties due to the fact it did not consider the nonlinearity in ozone formation chemistry (line 179). So how the authors justify their choices of one method over another? At least by using the CTM, the comparisons will be consistent. So in quantifying their contributions of ozone trends for cities and Waliguan (line 359-369), the simplified linear model were not acceptable since the three factors (𝑇_𝑚𝑒𝑡, 𝑄_{𝑛𝑜𝑛𝑙𝑜𝑐𝑎𝑙,} 𝑄_{𝑙𝑜𝑐𝑎𝑙} ) were not derived at the same ground.

Minor comments:

1. Explain QNR.
2. Line 93: put the abbreviation of NO_X where it was first introduced in the main content. The same for O₃ in line 102.
3. Line 132: Please double check the OMI website. I think OMI ozone products start from 2004.
4. Line 171: The authors chose 192 hours for their back-trajectory simulations; what was the basis for this time choice? Please provide a clear explanation of why 192 hours is sufficient to capture the majority of air mass transport influencing TP. For example, is this duration chosen based on previous studies, the lifetime of ozone precursors, or the distance from major emission source regions? Or discuss whether shorter or longer durations would significantly change the results of the transport analysis.
5. Add a note on the accuracy of model simulations.
6. Line 182: change to (Chen et al., 2022a)
7. Please double check the different MERRA2 resolution used for the HYSPLIT and GEOS-Chem model.
8. From Fig. 4, the authors stated that none of the satellite products captured the rapid ozone growth observed at the MEE stations. So it looks to me that these multiple satellite products were redundant, and not did not add extra merits to this study. Section 2.2 were really not necessary.  
9. In discussing QNR, the role of NOx as ozone precursors has been emphasized. However, it is recognized that NOx can also reduce ozone concentrations through NO titration under certain conditions, especially in high NOx environments or at night. Consider adding analysis (satellite data: HCHO and NOx) or references that illustrate the dependence of O₃ production on NOx in the study area.
10. Please unify the abbreviation/full name of the journal in the References as required.