Understanding the weakening patterns of inner Tibetan Plateau vortices

This study focuses on changes in the Tibetan Plateau vortices (TPVs) by using ERA5 reanalysis, covering the summers from 1979 to 2022 within the Tibetan Plateau (TP) region. These TPVs were identified using a geopotential height analysis. We discovered that the central-western TP had the most TPV activity and observed a clear decreasing trend in both the intensity and frequency of the TPVs in this region. This decrease was also accompanied by a decline in the strength of the associated vertical upward motion. To better understand this change, we employed the quasi-geostrophic omega equation. This allowed us to examine the dynamic, diabatic, and topographic factors contributing to the vertical motion during different phases of TPV activity in this region. Our results indicate that the main reason behind the weakened TPVs is the diminishing upper-level jet stream, which exerts dynamic forcing on the system. In the later stage, we observed that intensive moisture transport induces heightened diabatic vertical motion. However, this effect is not potent enough to counterbalance the diminishing dynamic influence. Therefore, our findings suggest a significant shift in TPV activity, transitioning from a dynamic-dominated regime to a latent heating-dominated diabatic regime. This new insight enhances our understanding of the complex mechanisms that influence TPV behavior.


Introduction
The Tibetan Plateau (TP), known as the highest plateau on Earth, is often referred to as the Asian water tower and the third pole of the world (Ma et al 2008, Qiu 2008).During the summer months (June-August), the Tibetan Plateau vortices (TPVs)-lowpressure weather systems active at 500 hPa with a horizontal scale of 400-800 km and extending vertically 2-3 km above the ground level-are prevalent over the TP.These TPVs are responsible for causing severe hydro-climatic extremes, leading to significant losses of life and property over the TP (Curio et al 2019, Zhang et al 2019, Lin et al 2020, Li and Zhang 2023).Therefore, it is crucial to investigate the synoptic-scale conditions associated with TPVs to mitigate the impacts of weather and climate disasters.
Previous studies have examined the physical processes behind the activity of TPVs from both climate and weather perspectives.These studies have attributed TPV activity to both dynamic and thermodynamic responses (Boos and Kuang 2010, Sugimoto and Ueno 2010, Curio et al 2019, Shou et al 2019, Lin et al 2020).Zhang et al (2019) specifically highlighted the importance of diabatic heating, dynamic effects associated with the upper-tropospheric jet stream, and topographic forcing in the formation and evolution of TPVs.Synoptic-scale dynamic effects, including the divergence related to the eastward-moving upper-level westerly jet stream and the convergence at 500 hPa resulting from northwesterly (in the westerly flow) and southerlies (from the monsoon trough over the Bay of Bengal), have been identified as significant factors influencing TPV activity (Feng et al 2014, Li et al 2019, 2021).Recently, based on potential vorticity analysis, Wu et al (2022) and Ma et al (2022) pointed out that the warmer TP surface leads to the surrounding isentropic surfaces tilting towards the TP surface.When the air parcel slides down the isentropic surface associated with the warm TP surface, the stability decreases.Meanwhile, the vertical relative vorticity is enhanced with potential vorticity restructuring and the condensational latent heating is released, which facilitates the intensification of the TPV.From the diabatic perspective, some studies suggested that the condensational latent heating plays a significant role in the development of low-pressure vortex over the TP, underscoring the significance of moisture vertical transport (Zhang et al 2019, Lin et al 2021).Some studies also highlighted that the penetrating southwesterly flow brings abundant moisture to the western TP, inducing increasing air ascent and TPV activity due to the vertical inhomogeneity of diabatic heating associated with the wet process from the perspective of the potential vorticity budget (Wu et al 2018(Wu et al , 2022)).The influence of topographic forcing has also been investigated, revealing that terrain primarily affects vertical upward motion in the middle to lower levels of the troposphere, thereby enhancing the activity of the mid-latitude vortices.Numerical sensitivity experiments removing the terrain demonstrate a noticeable decrease in precipitation intensity and moisture transport (Zhao et al 2020(Zhao et al , 2022)).Overall, the vertical motion is the direct response to the aforementioned forcings, playing a crucial role not only in cloud and precipitation formation but also in shaping the vertical profiles of horizontal divergence and convergence associated with TPV activity processes.
The overarching objective of this study is to gain a comprehensive understanding of the activity of TPVs during different periods over the past few decades.To accomplish this, we employ the quasi-geostrophic (QG) omega equation to investigate the multitude of factors influencing vertical motion.While the QG omega equation has not been previously utilized in studies investigating the formation and evolution of TPVs, it has been extensively applied in the examination of dynamic and thermodynamic contributions to mid-latitude cyclones and precipitation (Li et al 2020, Zhao et al 2022).In mid-latitude regions, the QG omega equation, in conjunction with the PV method, has served as a useful diagnostic tool for the exploration of low-pressure vortex phenomena (e.g.Lackmann 2011).Although previous studies have explored TPVs from different perspectives, such as dynamic actions and diabatic influences, the specific causes of variations in TPV intensity and frequency during different periods in the TP remain unclear.
To address this gap, we quantify the contributions of diabatic forcing, dynamic forcing, and topographic effects.We will investigate the physical mechanisms that drive changes in these factors to shed light on the conditions that create unfavorable synopticscale situations for TPV activity in the central-western part of the TP.By achieving these objectives, we can gain valuable insights into the factors influencing TPV activity and provide a deeper understanding of the underlying processes responsible for variations in TPV intensity and frequency over time.Ultimately, our findings will significantly contribute to enhancing our understanding of water resources and their variability in this critical region, aiding the development of more effective disaster management strategies.
The paper is structured as follows: In section 2, we provide a detailed description of the data and methods used in this study.Section 3 focuses on the synoptic-scale background of TPVs and investigates the changes in the influencing factors across various periods.Specifically, we quantify the contributions of diabatic forcing, dynamic forcing, and topographic effect to TPV activity and investigate the potential physical mechanisms that drive changes in these forcings.Finally, in section 4, we summarize our findings and engage in a discussion of the results.We highlight the key factors that influence TPV activity and discuss how our results contribute to a deeper understanding of the underlying processes responsible for variations in TPV intensity and frequency.

Data
To achieve the objective of identifying TPVs and investigating the synoptic-scale atmospheric circulation features, along with their physical influences on vertical motion, we employed the 6 hourly reanalysis data from the European Center for Medium-Range Weather Forecasting.This data is part of the fifth generation of atmospheric reanalysis (ERA5; Hersbach et al 2020) and covers the summer period (June-July-August) from 1979 to 2022.The ERA5 datasets provide a spatial resolution of 1.0 • × 1.0 • and span 27 pressure levels from 1000 hPa to 100 hPa.These datasets supply crucial variables, including horizontal winds, vertical velocity, specific humidity, geopotential height (GPH), surface pressure, air temperature, and topographic height.The climatological field refers to the mean from 1979 to 2022.

TPVs identification
Historically, a TPV was defined as a low-pressure system over the TP, characterized by closed contour lines or cyclonic winds at three observation stations at 500 hPa (Lhasa group for Tibetan Plateau meteorology research 1981).In this study, we refer to the TPV detection method proposed by Lin et al (2020) for the identification of TPV which builds on gridded data to identify low pressure systems with closed isolines.The results for TPV center identification at the 500 hPa height over the region (70 1(a) (purple dots).The specific detection steps involved can be found in the supporting information.It should be noted that Lin et al (2020) conducted a study on the interannual variations of TPVs across the entire TP region.In contrast, our research is centered on the high-activity region of TPVs, which is a subset of the TP.Furthermore, the reanalysis datasets employed for detection in our study differ from those used by Lin et al (2020).Our methodology also diverges in terms of predefined max pooling and the limited radius of TPVs.Therefore, our results are not directly comparable to those of Lin et al (2020).

The QG omega equation
The structures of TPVs are characterized by the presence of strong upward motion (Li et al 2011(Li et al , 2020)).To better understand the factors influencing change in TPV activity on an interannual scale, the vertical motion can be analyzed by considering atmospheric dynamic forcing, diabatic forcing, and topographic forcing, which can be derived using the QG-omega equation (Nie and Fan 2019, Li and O'Gorman 2020): (1) The first two terms on the right-hand side of equation ( 1) are the geostrophic wind variation with height and the distribution of thermal wind maximum which stands for the dry dynamic forcing with corresponding vertical motion of ω Vd and ω Td .The third right-hand term is the diabatic heating forcing term.The diabatic heating term, can be indirectly obtained by atmospheric apparent heat source (J) with the thermodynamic equation in the hydrostatic approximation in pressure coordinates (Yanai et al 1973, Wang and Qian 2000, Zhao et al 2021): The QG vertical motion (ω QG ) is iteratively solved by the successive overrelaxation method with a relaxation factor varying in pressure in equation (1).However, equation (1) does not include topographic forcing.
The individual vertical motion is driven by dynamic forcing and diabatic forcing with vertical velocity set to zero at the surface boundary.ω Dia is the vertical motion term triggered by diabatic forcing.The individual vertical motion of the ω Vd , ω Td , and ω Dia is showed in equation (3a).
A, B, and C in equation (3a) are the three terms of the right-hand side of equation ( 1).Finally, the ω QG can be expressed as equation ( 4): ω Dyn is the sum of ω Vd and ω Td , which presents the dry dynamic vertical velocity.The meanings of other mathematical symbols used in equations are explained in table 1.

Diagnosis of interdecadal changes in TPVs
Figure 1(a) illustrates the spatial distribution of summer TPVs, which are primarily located in the central , providing strong confirmation of the accuracy of our results.Furthermore, we observe that the activity region of TPVs is strongly influenced by low-level wind shear in the transition zone, where the westerlies and southerlies converge, creating favorable conditions for TPV activity.Figure 1(b) also demonstrates a distinct banded pattern of vertical upward motion, accompanied by a convergence field near the surface of the TP.Three reasons can explain why the centers of vertical upward motion and convergence do not perfectly coincide.Firstly, the vertical upward motion may not align well with the TPV center due to the small-scale structure of the latter.Secondly, it is generally understood that the vertical upward motion is primarily situated over the precipitation region, which is slightly ahead of the cyclonic vortex in the mid-latitude region.Thirdly, and most importantly, the vertical motion in the key region is determined whenever there is TPV activity.This could lead to spatial location mismatches, with additional vertical motion factored into the results where no corresponding vorticity occurs, or where the vertical motion is neutralized.In any case, vertical upward motion plays a significant role in TPV activity.Therefore, to gain a deeper understanding of TPVs, it is crucial to investigate the changing trend of TPV frequency and intensity in the key region over time, as well as how the physical factors influencing vertical motion, particularly the primary determinants, change.
In the context of global climate warming, changes in the westerlies and the south Asian summer monsoon can lead to fluctuations in surface pressure, directly impacting the activity of TPV.To investigate such variations in the key region, we construct the probability distribution functions (PDFs) of the synoptic-scale GPH intensity background using all grid points in the key region from 1979 to 2022, as shown in figure 2(a).The climatological field of GPH is partitioned into four equal timeslots spanning the entire period, and the peak probability of interdecadal changes in low-level GPH intensity reveals an increasing trend every 11 years, suggesting a progressive rise in surface pressure over time.This suggests that the climatological background is becoming increasingly unfavorable for TPV activity.However, the climatological regional average GPH in the key region does not show significant changes on an interannual scale (figure S1), implying that the main body of surface GPH remains relatively stable, but the GPH when a TPV occurred in the key region has increased, potentially contributing to the weakening of TPVs. Figure 2(b) shows the interannual variations and long-term trends in TPV intensity and frequency, indicating unfavorable conditions for TPV activity, as evidenced by the significant trend of increasing GPH intensity and decreasing TPV frequency.Although the rise in low-level GPH in the westerncentral part of the TP offers a plausible explanation for the observed changes in TPV frequency, it is essential to discern and elucidate various factors that have distinctly contributed to the gradual weakening of TPVs over time.
In the introduction, we highlighted the importance of vertical upward motion in characterizing TPV.As depicted in figure S2, the analysis reveals a weakening trend in the interannual variation of vertical motion associated with TPV activities in the key region, which is consistent with the changes in TPV intensity and frequency shown in figure 2(b).Although the regional average of GPH associated with TPV activities in the key region (figure S1) exhibits a slightly weaker trend than that in figure 2, the difference in GPH intensity remains significant between the earliest and the latest 11 year periods, indicating a weakened state of low vortex activity.To study the reasons for the changes in TPVs while ensuring an adequate sample size for analyzing changes in vertical motion, we will focus on the early phase (1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989) and late phase (2012-2022).These two phases capture the most significant interdecadal differences in TPV intensity and frequency, as well as the corresponding GPH intensity in the key region.

The attribution of TPVs changes in selected two phases with the quasi-geostrophic omega equation
We utilized the QG-omega equation to delve into the physical factors that contribute to the disparity in vertical motion between the early and late phases.This approach provides insights into the reasons behind the observed decrease in low vortex intensity and frequency.As shown in figure 3, while differences exist between ω and ω QG , the latter can account for most of the former.The unexplainable part is presumably due to the ignorance of non-QG processes, computation errors, and boundary layer friction (Zhao et al 2023).This finding suggests that the QG dynamics can reasonably explain the inverted vertical motion associated with the low vortex.
Upon comparing Figures 3(a) and (b), we observe that the topographic factor primarily affects low-level vertical motion in the TP, while diabatic forcing is mainly observed at mid-levels, likely due to the latent heating of condensation.Interestingly, ω Topo changes little between the two phases, with a relative change of only 7.3%.In contrast, the relative change of ω Dia decreases by 28.6% between the early phase and the late phase (figure 3(c)).This indicates that under the global warming background, moisture transport may have been enhanced, leading to more diabatic heating in the late phase.
In the early phase, we found that ω Dyn is the most crucial component of ω QG in determining low vortex activity, particularly in the upper-level troposphere (figure 3(a)).This highlights the importance of the upper-level dynamic.Furthermore, the most significant change between the two phases is in ω Dyn , which exhibited a relative change of 161.9%.The difference between ω Dyn and ω Dia can also verify the above results and the specific details can be found in figure S4.This change is likely the primary reason behind the decrease in vertical motion and  c) and (d) Are the anomalous specific humidity (shading; 10 −4 kg kg −1 ) and divergence field (contours; 10 −5 s −1 range from −5 to 5) along the meridional average (84 • -95 • E) in the early phase and the late phase.(e) and (f) Are the same as (c) and (d), but for the anomalous latent heat of condensation calculated by the atmospheric apparent moisture sink (shading; 10 −3 J kg −1 s −1 ) with 95% confidence test by gray slashes.The black shading is the topography.The key region is located in between the two dashed green lines in (c) and (d).
low vortices during the late phase.Thus, based on the aforementioned analysis, it becomes imperative to identify the dynamic systems responsible for the alteration in vertical motion.
Figure 4 provides valuable insights into the potential synoptic-scale atmospheric circulation that influences changes in vertical upward motion.In the early phase, the vertical motion is strongly influenced by the presence of a prominent upper-level jet (ULJ) stream, as evidenced by the strong divergence field in the central-east part of the key region at the equatorial entrance of the ULJ (figure 4(a)).However, during the late phase, there is no significant ULJ activity, and the divergence field at the upper-level troposphere is primarily observed to the south of the key region (figure 4(b)).This suggests that the dynamic effect plays a more substantial role in driving vertical motion in the early phase compared with the late phase, which is consistent with the findings in figure 3. It is important to note that both the location and intensity of the ULJ, which are determinants of the weakening of vertical velocity over the study period during TPV phases, influence the dynamic role effect.To further investigate this, we have computed the full wind field difference between the early and late phases.As depicted in figure S5, there is a significant presence of westerlies near the central-eastern region of the TP.This suggests that the observed weakening of vertical velocity is primarily due to a decrease in wind field intensity, rather than a shift in the wind field's location.
To further confirm the importance of moisture in the late phase, we depict the distribution of specific humidity in both the early and late phases in figures 4(c) and (d) as well as the contribution of the atmospheric apparent moisture sink in figures 4(e) and (f), which is related to the latent heating of condensation (Wei et al 2014).The results reveal apparent divergence (convergence) structures in the upper-level (low-level) regions during both phases, indicating strong upward motion.Moreover, the contribution of moisture with latent heating of condensation is notably stronger in the late phase compared to the early phase, and the latent heating of condensation exhibits an obviously enhancing trend (figures 4(c)-(f) and S3).This finding reinforces the notion that diabatic heating, facilitated by increased moisture content, plays a more significant role in vertical motion during the late phase.
Overall, the decrease in upper-level wind speed is identified as the primary dynamic factor leading to the weakening of vertical velocity in the late phase compared to the early phase.Additionally, the enhanced contribution of moisture, which amplifies the diabatic heating effect, promotes enhanced diabatic upward motion in the late phase.However, it is important to note that the increased vertical motion driven by moisture contribution in the late phase is not sufficient to offset the weakened vertical motion caused by the decreasing effect of the ULJ.Consequently, this leads to a decrease in the frequency and intensity of TPVs.In summary, the combined effect of the weakened ULJ and the enhanced contribution of moisture leads to a weakening of vertical motion, which is the primary reason for the decrease in TPV activity.However, latent heat may progressively emerge as the predominant factor if global warming continues to intensify.The TPV activity appears to be transitioning from a dynamicaldominated regime into a latent heating-dominated diabatic regime.

Summary and discussion
This research presents findings on the activity of TPVs and the physical mechanisms influencing their changes between 1979 and 2022.By using an objective method for identifying TPV activity, the study confirms that TPVs tend to occur in the central-west region of the TP, which serves as a transition area between the westerlies and southerlies.However, an analysis of TPV center variations reveals a decreasing trend in both intensity and frequency.
While diabatic vertical motion is enhanced in the late stage due to increased moisture transport to the TP, it is insufficient to counterbalance the effect of dynamic forcing.The study also suggests that the activity of TPVs, accompanied by strong vertical motion, is influenced by the ULJ-dominated dynamical regime transitioning into a latent heating-dominated diabatic regime with the enhanced latent heat of condensation as a response to global warming, signifying the intensified warming and humidification of the TP.
The study emphasizes the need for further exploration of additional factors, such as the impact of boundary layer friction in future research.Additionally, this investigation does not account for the significant influence of land-air interactions, like surface sensible and latent heat fluxes.The QGomega equation, widely used in atmospheric dynamics to study vertical motion, may suffer from numerical calculation errors during QG-omega inversion.Although the equation incorporates topographic effects to explain vertical motion as much as possible, discrepancies between calculated and observed values still persist.Notably, previous research has indicated a higher presence of low clouds over the Plateau during summer, which may be attributed to the pronounced effect of strong boundary layer friction.Therefore, an investigation into the impact of boundary layer friction and other associated factors is imperative to attain a thorough comprehension of the underlying physical mechanisms that drive alterations in TPV activity, which holds significant importance in the realms of disaster management and water resource management in this region.
In addition, this study used a traditional omega equation to analyze the TPV activity, and the two dynamic terms may counteract with each other, and the interaction of different forcings cannot be obtained.However, Wu et al (2020) divided the vertical motion into the isentropic gliding term, isentropic displacement term, and diabatic term by combining the quasi-geostrophic potential vorticity equation and vertical motion equation.The developed equation can not only overcome the cancellation of different terms but also highlight the interaction of dynamic and diabatic terms.It is very helpful for reference in the following works.
All data that support the findings of this study are included within the article (and any supplementary files).

Figure 1 .
Figure 1.(a) The distributions of the TPVs (represented by dots) with their frequency density (shading; the number of each grid) and the corresponding wind fields (vector; m s −1 ) (b) The difference of vertical uplift motion (shading; Pa s −1 ), convergence field (contours; 10 −5 s −1 ), and their climatologies at 500 hPa as long as when a TPV was identified within the key region.The gray line represents the TP.The key region of TPVs is indicated by the green rectangle (30.5 • -35 • N, 84 • -95 • E).

Figure 2 .
Figure 2. (a) The probability distribution functions (PDF) of interdecadal GPH constructed by each grid in the key region for the climatological background (1979-2022), and (b) the interannual change of TPV centers intensity (black curve line; m) and frequency (bars) in the key region as shown in figure 1(b) during 1979-2022.Note that no TPVs were detected in 2022.

Figure 3 .
Figure 3. (a), (b) The vertical profiles of ω (purple solid), ωQG (red solid), ωDyn (blue dashed), ω Dia (green dashed), and ω Topo (gray dashed) averaged over the key region when a TPV was identified.(c) The sum of different vertical velocity profiles from 500 to 100 hPa in the early phase (blue bars) and in the late phase (green bars) as well as the relative change between the early phase and late phase (purple bars).Black error bars denote one standard error.

Figure 4 .
Figure 4.The composite of divergence field (contours; 10 −6 s −1 ) and anomalous full wind speed (shading; m s −1 ) associated with TPV activities in the key region in the early (late) phase and climatological field during 1979-2022 at 200 hPa with 95% confidence level by white dots in the early phase (a) and the late phase (b).(c) and (d) Are the anomalous specific humidity (shading; 10 −4 kg kg −1 ) and divergence field (contours; 10 −5 s −1 range from −5 to 5) along the meridional average (84 • -95 • E) in the early phase and the late phase.(e) and (f) Are the same as (c) and (d), but for the anomalous latent heat of condensation calculated by the atmospheric apparent moisture sink (shading; 10 −3 J kg −1 s −1 ) with 95% confidence test by gray slashes.The black shading is the topography.The key region is located in between the two dashed green lines in (c) and (d).

Table 1 .
List of mathematical symbols.