The analysis of primary circulation flow patterns and environmental parameter characteristics of short-duration heavy precipitation during the warm seasons in the Western Tianshan Mountains, Xinjiang

Based on a decade of observational and reanalysis data, this study systematically analyzes the main atmospheric circulation patterns and environmental conditions linked to warm-season short-duration heavy precipitation in the Western Tianshan Mountains, Xinjiang. Three dominant circulation patterns are identified: the Low-Trough pattern, the Low-Vortex pattern, and the Eastward-moving Waves pattern. The Low-Trough and Low-Vortex patterns are further divided into subtypes based on the location of the trough or vortex, while the Eastward-moving Waves pattern is characterized by a zonal flow with eastward-moving short-wave disturbances. Short-duration heavy precipitation shows a clear daily cycle across all patterns, with the highest frequency occurring from afternoon to early evening. Rainfall tends to concentrate on windward slopes and in valleys. The exact location of precipitation is influenced by the pressure gradient and wind direction—shifting northward under stronger westerly winds and southward under weaker southwesterly flows. The Low-Vortex pattern exhibits the strongest coupling between upper- and low-level jets and the most extensive jet coverage, providing optimal lifting conditions for intense and widespread rainfall. The Low-Trough pattern shows moderate jet coupling, while the Eastward-moving Waves pattern relies on weak upper-level forcing, resulting in more limited precipitation. Moderate vertical wind shear is common during these events. All patterns display a consistent vertical structure of pseudo-equivalent potential temperature, characterized by a “high–low–high” pattern. The Low-Vortex pattern features a prominent upper-level high-value layer and a shallow mid-level cool layer, favoring intense convection. The Low-Trough pattern shows strong heating in both the upper and lower levels but a more distinct mid-level cool zone, enhancing atmospheric instability. In contrast, the Eastward-moving Waves pattern has a thick mid-level cool layer and weak low-level heating, leading to the weakest convective potential. The Low-Vortex and Low-Trough patterns are associated with higher low-level moisture, while the drier Eastward-moving Waves pattern requires greater convective available potential energy and lower convective inhibition to initiate convection.

1 Introduction

Short-duration heavy precipitation (SDHP) refers to a meteorological phenomenon characterized by significant rainfall intensity over a brief period, often leading to flash floods, debris flows, and other secondary disasters (Chen et al., 2013; Li et al., 2017). While national standards in China define SDHP as hourly rainfall ≥ 20 mm—a threshold suited to humid monsoon regions (Yu, 2013). But the arid and semi-arid climate of Xinjiang necessitates a regionally adapted threshold. Consequently, the average annual rainfall is less than one-fourth of the national average, and an hourly rainfall of ≥ 10 mm is adopted as the SDHP criterion, as it represents a hydrologically critical intensity that readily exceeds local infiltration capacity and triggers hazardous runoff (Yan et al., 2022).

Previous research on SDHP has evolved from statistical characterization to process-based understanding. Key advancements include the development of monitoring and nowcasting techniques using radar and dense networks (Ducrocq et al., 2002; Hao et al., 2012; Kato et al., 2017; Markowski and Richardson, 2010; Osborn et al., 1979; Yin, 2007), which have improved early warning. Studies on spatiotemporal patterns, such as those by Chen et al. (2013) (Li et al., 2019), have been crucial in identifying regional diurnal cycles and links to Mesoscale Convective Systems (MCSs). Critically, research into triggering mechanisms has established that this phenomenon is not solely a mesoscale event but arises from the interaction of systems across scales, from large-scale circulation to microphysical processes (Fujibe et al., 2002; Iii, 1987). A foundational framework for diagnosing such events is the “ingredient-based” methodology, which posits that heavy rainfall requires the co-location of moisture, instability, and lift (Doswell et al., 1996). This approach has been successfully extended to orographic precipitation regimes. For instance, Lin et al. (2001) refined the methodology for mountainous regions, emphasizing the critical roles of moist upslope flow and terrain-forced ascent. Subsequent modeling studies (Houze, 2012; Lin et al., 2013) have further detailed how terrain geometry modulates convective organization and precipitation efficiency, highlighting that in complex terrain, the spatial coupling between synoptic flow and local topography becomes a decisive ingredient for extreme rainfall.

This fundamental understanding underscores the challenge of prediction and is reflected in the advancement of high-resolution numerical models aimed at simulating these complex events (Alila, 2000; Zhang et al., 2017). Although SDHP often manifests through meso- and micro-scale systems, it is frequently triggered and organized by large-scale disturbances such as fronts and cyclones, resulting from complex multi-scale interactions (Ding, 2013). Globally, the interplay between large-scale disturbances and mesoscale terrain effects has been shown to dictate the location and intensity of heavy precipitation in regions such as the European Alps (Ducrocq et al., 2008) and the Sierra Nevada (Neiman et al., 2014).

In Central Asia, the westerly jet acts as a key linkage among high-, mid-, and low-latitude systems, modulating summer precipitation over the Tianshan region (Jiang and Wang, 2015; Yang et al., 2018). However, most existing SDHP research has focused on eastern China or other monsoon-affected areas, with little attention paid to the distinct meteorological orographic setting of the Western Tianshan. Systematic identification of the terrain-modulated ingredients—such as orographically enhanced moisture convergence, elevated instability zones, and wind-terrain interactions—for SDHP in this high-altitude, arid to semi-arid region remains notably lacking.

Therefore, this study aims to bridge this gap by integrating an ingredient-based perspective with a comprehensive analysis of the Western Tianshan. Specifically, we seek to identify the dominant synoptic-scale circulation patterns responsible for warm-season SDHP, analyze the spatiotemporal characteristics of SDHP under each pattern; and investigate the associated environmental parameters, including jet structures, vertical wind shear, thermal stability, moisture, and convective energy that govern SDHP initiation and intensity. By integrating a decade of high-resolution station data, soundings, and reanalysis products, this work establishes a systematic, terrain-aware understanding of SDHP mechanisms in this complex region, with the ultimate goal of improving local forecasting and early warning capabilities.

2 Study region

The Western Tianshan region is situated in the western part of the Tianshan Mountains in Xinjiang, China. It is encircled by mountains to the north, east, and south, resulting in a distinctive “funnel-shaped” topography. This region experiences a temperate continental climate with abundant precipitation, earning it the designation of the “Wet Island of Central Asia.” The region’s growing economic and strategic importance underscores the need for an improved understanding of its high-impact weather events is needed.

The terrain within the Western Tianshan Mountains is marked by significant elevation variations (ranging from 330 to 4,220 m). The underlying surface exhibits complexity due to an intermingling of forests and grasslands. During summer months, localized uneven heating combined with high moisture content near ground can easily trigger convection processes that lead to SDHP events. In the context of global warming, disasters induced by SDHP have caused significant damage to local infrastructure, including transportation, energy, and water conservancy projects. These events also pose serious risks to human safety and property (Shein, 2016). Therefore, conducting research on SDHP in this region is of significant importance for enhancing meteorological forecast accuracy and advancing regional meteorological scientific inquiry (Figure 1).

Figure 1

Figure 1. Topography and spatial distribution of warm-season mean precipitation in the Western Tianshan Region.

3 Data

The study period from May to September was selected to encompass the core of the warm season in the Western Tianshan Mountains. This period is defined by a rapid spring warming that establishes a thermally unstable boundary layer by May and a slow autumn cooling that maintains convective potential through September. Consequently, this 5-month window captures the vast majority of short-duration heavy precipitation events, which are critically dependent on the warmth and moisture available during this thermodynamically active season. The data from the inactive colder months are excluded due to the extreme rarity of such convective events.

A combined dataset from 10 national stations and 148 regional automatic weather stations was used. While the national stations adhere to the most rigorous calibration and maintenance standards of the China Meteorological Administration (CMA), the core instrumentation for measuring hourly precipitation is consistent across both station types. The integration of the regional stations was essential for achieving the high spatial resolution required to resolve the complex topographic influences on precipitation. To ensure homogeneity, a normalization procedure was applied to the event frequency data, as detailed in the Methods section.

The hourly observational data were from the “Sky Engine” big data platform, which is the official, unified data sharing and application platform of the China Meteorological Administration (CMA) (Xu et al., 2023). It integrates real-time and historical data from all national-level meteorological observation stations and thousands of regional automatic weather stations across China. The platform implements rigorous, operationally-applied quality control procedures, including temporal and spatial consistency checks and extreme value control. Therefore, it serves as the primary and most authoritative source for meteorological data within China’s operational and research communities.

The distribution of annual mean precipitation from May to September in the past 10 years at each station shows that the areas with high precipitation during the warm season in the western Tianshan Mountains are mainly distributed on the windward slopes of the north-south boundary mountains and in the eastern region where the boundary mountains converge, while the precipitation in the central plain is relatively small (Figure 1).

For the reanalysis data, the ERA5 hourly dataset was selected, featuring a spatial resolution of 0.25° × 0.25°. For the sounding data, this study utilizes information from the sole sounding station in the region, the Yining Meteorological Sounding Station, which records observations at 00:00 and 12:00 daily. All times referenced in this paper are in Greenwich Mean Time (GMT), which is approximately 5 h behind the local time.

4 Methods

We define a systematic SDHP event as one that meets either of the following criteria: 1. two adjacent stations within the same region record hourly precipitation (R) ≥ 10 mm on the same day, or 2. a single station records R ≥ 10 mm for at least two consecutive hours, or 3. Simultaneously satisfying the conditions in (1) and (2).

Given the significant disparity in the number of SDHP samples associated with different circulation patterns, a normalization method has been employed for statistically comparing the spatial distribution characteristics of these events’ frequencies under each Circulation pattern. The method involves calculating the frequency of SDHP at each station based on the different circulation patterns, using a formula that standardizes the results to a range of [0–1] (Wilks, 2011).

${\mathrm{X}}^{*}=\left(\mathrm{X}-{\mathrm{X}}_{\min }\right)/\left({\mathrm{X}}_{\max }-{\mathrm{X}}_{\min }\right)$

In this formula, X represents the actual frequency at a specific station, X_min and X_max denote the minimum and maximum occurrences of SDHP among all stations, and X* signifies the normalized outcome.

5 Research results

5.1 The primary circulation flow patterns for SDHP

Based on the identification criteria, a total of 122 SDHP weather processes were selected in the West Tianshan region. Analysis reveals that three dominant circulation patterns at 500 hPa are responsible for the majority of these events: the Low-Trough (LT) pattern, the Low-Vortex (LV) pattern, the Eastward-moving Waves (EW) pattern, Northern-Transverse-Trough (NTT, hereinafter) pattern, and Northwest-Airflow (NA, hereinafter) pattern. The data indicate that the LT, LV and EW pattern collectively account for over 86% of the SDHP cases, while the remaining two types have fewer instances and are not specifically analyzed in this paper. The average duration of SDHP processes across the five circulation patterns ranges from 1.0 to 1.5 h, with the Low-Trough type being the longest. The number of occurring days during the process indicates that under favorable atmospheric conditions, SDHP process exhibit characteristics of multiple centers and time periods. Detailed information is presented in Table 1.

Table 1

Table 1. Statistics of SDHP processes associated with different circulation flow patterns in the Western Tianshan Mountains.

The LT pattern is defined by a Two-Ridge-One-Trough configuration at 500 hPa over Eurasia. Based on the position of the trough axis, it is categorized into Central Asian Low-Trough (CALT, hereinafter) pattern and West Siberian Low-Trough (WSLT, hereinafter) pattern. Under the condition of CALT, the trough line is located within Central Asia (60°–100°E, 35°–50°N), placing the Western Tianshan region in the southwesterly flow ahead of the trough (Figure 2a). For the WSLT pattern, the trough line is positioned further north over West Siberia (60°–100°E, 50°–70°N), resulting in a stronger meridional flow and westerly dominance over the study area (Figure 2b). In both patterns, the area ahead of the trough experiences warm, moist advection and positive vorticity advection, promoting upward motion which, when combined with orographic lifting, effectively triggers SDHP.

Figure 2

Figure 2. Composite 500 hPa geopotential height and temperature fields for primary circulation patterns associated with SDHP in the Western Tianshan Mountains. (a) CALT (b) WSLT (c) CALV (d) WSLV (e) EW (The blue solid lines represent geopotential height, unit: dagpm; The red dashed lines represent temperature, unit: °C; The red solid line represents geographic boundary of the study area).

The LV pattern is characterized by a Two-Ridge-Two-Trough configuration at 500 hPa, featuring at least two closed isopleths and an associated cold core within a persistent low-pressure system. It is subdivided into Central Asian Low-Vortex (CALV, hereinafter) and West Siberian Low-Vortex (WSLV, hereinafter) pattern. In the CALV pattern, the vortex is centered within Central Asia (60°–100°E, 35°–55°N), as shown in Figure 2c. For the WSLV pattern, the vortex develops over West Siberia (60°–100°E, 55°–70°N), as shown in Figure 2d. The closed circulation of the low vortex fosters stronger and more persistent convergent ascent, facilitating the organization of mesoscale convective systems and resulting in widespread and heavy precipitation bands.

The EW pattern is distinct from the trough/vortex patterns and is characterized by a zonal circulation across Eurasia. A steady westerly flow prevails between 40°N and 50°N, which is perturbed by eastward-propagating short-wave disturbances (Figure 2e). Although these short waves are associated with weaker moisture transport, they can still trigger SDHP when their dynamic lifting interacts with afternoon thermal instability over favorable topography.

In summary, the specific dynamic and thermodynamic configurations inherent to each of these three circulation patterns collectively determine the location, intensity, and diurnal variation of SDHP events in the Western Tianshan region.

5.2 Characteristics of spatiotemporal distribution of SDHP

The analysis of SDHP events across various weather systems reveals that the LV pattern occurs most frequently at 348 station times, followed by the LT pattern, which has 221 station times, while the EW pattern records the least frequency at 108 station times.

5.2.1 Temporal distribution characteristics of SDHP

An examination of the diurnal variation was conducted on cumulative values (total station times) and average frequency (total station times divided by the number of stations experiencing SDHP within the total sample) for such events occurring under three weather influence systems, as illustrated in Figure 3. Notably, the LT pattern of SDHP displays a multi-peak feature, with a peak period from the afternoon to early evening (8:00–16:00), peaking around 13:00. The LV pattern shows a unimodality, with similar peak times to the LT pattern. In contrast, the EW pattern exhibits a three-peak pattern, with peak periods slightly delayed, occurring at 11:00–12:00, 14:00–16:00, and 19:00–20:00, with peaks at 11:00, 16:00, and 20:00, respectively.

Figure 3

Figure 3. Diurnal variations of accumulated frequency and station-mean SDHP frequency for the three circulation patterns in the West Tianshan region. (a) LT (b) LV (c) EW.

From these observations, it is evident that stations with higher average frequencies tend to have a more concentrated spatial distribution of SDHP, indicating that heavy rainfall is more likely to occur in specific areas during those times. In the West Tianshan region, the diurnal variation peaks for the three weather influence systems show significant overlap, suggesting that areas experiencing frequent SDHP are relatively concentrated in this region.

5.2.2 Spatial distribution characteristics of SDHP

The northern boundary mountains composed of The Borokonu Mountains and The Keguqin Mountains, as well as the southern boundary mountains composed of the Halkataw Mountains and The Narati Mountains, have a significant impact on the precipitation in the Western Tianshan Mountains. “The Two Valleys” refer to the Ili River Valley (Northern Valley) and the Kunes River Valley (Southern Valley). The Southern Valley has a higher elevation and covers a relatively smaller area compared to the Northern Valley (Figure 1). Through the homogenization calculation of SDHP events, the precipitation distribution under the influence of three patterns of weather circulation airflow is identified, as shown in Figure 4.

Figure 4

Figure 4. Normalized spatial frequency distribution of SDHP under different circulation patterns in the Western Tianshan Mountains (a) CALT (b) WSLT (c) CALV (d) WSLV (e) EW (f) Total Frequency.

Under conditions characterized by CALT, SDHP in the Western Tianshan region reveals two prominent precipitation belts: one in the western part of the Northern Valley and another along the northern boundary mountains, which appear as patchy and linear patterns respectively, with multiple high-frequency precipitation centers dispersed within them (Figure 4a). However, when affected by WSLT, there is a notable alteration in spatial distribution. Although Northern-Southern rain belts persist, the size of these belts significantly decreases, and take on a more patchy appearance. The northern rain belt predominantly occupies central regions of the boundary mountains while its southern counterpart is situated within central parts of both Southern Valley and its adjacent southern foothills of Usun Mountains, with fewer precipitation centers (Figure 4b).

In contrast to previous observations under different weather circulation patterns, strong precipitation frequencies associated with CALV patterns exhibit widespread distribution across all regions except for the eastern part of the Western Tianshan Mountains and the central section of the northern boundary mountains. Multiple high-frequency precipitation centers are situated in the central regions of both the Northern and Southern Valleys, as well as in the western part of the northern boundary mountains (Figure 4c). In the case of the WSLV, the regions with frequent SDHP shift northward, primarily concentrated in the northern boundary mountains and flanking both sides of the Usun Mountains, with a reduction in the size of the precipitation belts. The high-frequency precipitation centers are mainly found in the northern boundary mountains (Figure 4d).

The occurrence of SDHP associated with EW is predominantly observed in the southern boundary mountains, the Southern Valley, and the eastern part of the Western Tianshan Mountains. These areas exhibit a multi-center distribution, with the highest intensity recorded in the eastern part of the Usun Mountains (Figure 4e). Overall, the frequency of SDHP in the Western Tianshan area reveals the presence of two prominent precipitation belts: one located at the southern foothills of the northern boundary mountains, and the other in the Southern Valley. The high-frequency precipitation centers are extensively distributed, primarily located in lower mountainous regions or on windward slopes at higher elevations. Conversely, plain areas experience relatively low frequencies of SDHP, indicating that orographic uplift plays a significant role in triggering these events in the Western Tianshan Mountains. The above analysis further underscores that SDHP exhibits distinct regional characteristics throughout this area.

The spatial distribution of precipitation is closely linked to the synoptic-scale circulation (Figure 2). Under the influence of the Central Asian Low-Trough/Vortex (CALT/CALV), which features a weak horizontal pressure gradient, the region experiences southwesterly flow ahead of the trough. This flow favors SDHP in the south, particularly over the southern boundary mountains and the southern foothills of the Usun Mountains. Conversely, during events associated with the West Siberian Low-Trough/Vortex (WSLT/WSLV), a stronger pressure gradient results in a dominant westerly flow. Consequently, the primary precipitation zone shifts northward, affecting the northern boundary mountains and the northern foothills of the Usun Mountains.

5.3 Principal atmospheric environmental parameters

SDHP is classified as a form of convective weather, necessitating local atmospheric instability, abundant water vapor, substantial potential energy, and dynamic conditions. Regarding the above three circulation airflow patterns, the influence of the atmospheric environment on the occurrence and sustaining mechanisms of SDHP in the West Tianshan region is investigated.

To ensure the relevance of analysis to actual conditions, the meteorological stations within a 100 km radius of the sounding station (Yining) are selected in the West Tianshan region as observational points. The specific time periods chosen correspond to those immediately preceding the occurrence of SDHP. As most convective events occur in the afternoon, the soundings from 00:00 UTC are utilized, along with the maximum ground temperature (T_g, hereinafter) and dew point temperature (T_d, hereinafter) recorded during the afternoon to refine the calculations of Convective Available Potential Energy (CAPE, hereinafter) and Convective Inhibition Energy (CIN, hereinafter). Similar adjustment methods are applied to convective events occurring at other times.

5.3.1 Dynamical conditions

The initiation and organization of SDHP are critically dependent on the presence and configuration of upper-level jets (ULJ) and low-level jets (LLJ). The LLJ functions primarily as a moisture conveyor and a source of dynamic lifting, thereby initiating convective development. In contrast, the ULJ promotes large-scale ascent and the organization of convection through associated divergence patterns. The coupled effect of ULJ and LLJ, particularly when interacting with the complex topography of the Western Tianshan, can significantly enhance the organization and persistence of Mesoscale Convective Systems (MCSs), leading to SDHP events.

In this study, the LLJ and ULJ are defined as wind speed maxima exceeding 12 m s^-1 at 850 hPa and 30 m s^-1 at 200 hPa respectively. The LLJ primarily triggers precipitation where its flow intersects with terrain features or where low-level speed convergence is enhanced. Conversely, the ULJ favors SDHP occurrence on the right side of its entrance region or the left side of its exit region, where upper-level divergence effectively promotes sustained upward motion.

The spatial distributions of ULJ and LLJ for each of the three primary circulation patterns are shown in Figure 5. Under the influence of LT, jet streams are observed from 200 hPa to 700 hPa over the western Tianshan Mountains. Prominent southwesterly ULJ occur at 200 hPa and 500 hPa, while a westerly jet is identified behind the trough at 700 hPa. In contrast, the study area is characterized by weak Easterly flow at 850 hPa. In the case of LV, jet streams are present on all isobaric surfaces, exhibiting greatest intensity and broadest spatial extent, with pronounced eastward LLJ at 850 hPa. The fluctuating jet stream in the EW pattern is the weakest. Southwesterly jets appear only above 500 hPa, accompanied by weak westerly flow between 700 hPa and 850 hPa.

Figure 5

Figure 5. Distributions of low-level and high-level jets under different circulation patterns in the Western Tianshan Mountains (a) LT (b) LV (c) EW (The blue, red, green, and orange solid lines with arrows from top to bottom denote jet streams at 200, 500, 700, and 850 hPa, respectively).

This analysis indicates that the combination of upper-level divergence and low-level convergence introduced by the LLJ associated with the LV pattern is highly favorable for triggering widespread SDHP. The LT pattern also enhances this effect, though to a lesser extent. In contrast, the EW pattern relies primarily on weak suction from ULJ to lift low-level airflow, resulting in relatively weaker precipitation intensity and limited spatial coverage.

Among the three patterns of SDHP weather processes, the LV pattern has the highest proportion of both ULJ and LLJ, followed by the LT pattern, while the EW pattern has the lowest proportion. According to the calculated divergence fields at different levels, all three patterns of SDHP weather processes exhibit a synoptic-scale configuration conducive to uplift, characterized by upper-level divergence and lower-level convergence, which is then reinforced by latent heat release from the ensuing convection. For more details, please refer to Table 2.

Table 2

Table 2. Occurrence frequency of key dynamic conditions for SDHP under different circulation patterns in the Western Tianshan Mountains.

Vertically, the change in wind field with height, known as vertical wind shear, greatly influences the organization, structure, and evolution of convective storms. When sufficient moisture, static instability, and lifting triggers are present, vertical wind shear has the most profound impact on the organization and characteristics of convective storms, making it an important parameter for forecasting severe convective weather. Typically, deep-layer wind shear is represented by the vector difference in wind from the surface to a height of 6 km (Doswell, 2001).

The intensity of vertical wind shear is an important parameter for forecasting severe convective weather, including heavy rainfall, as it critically influences the organization, structure, and longevity of convective storms (Doswell and Evans, 2003; Doswell et al., 1996; Markowski and Richardson, 2010). This paper categorizes the intensity of vertical wind shear in the 0–6 km range into three levels: weak (<12 m·s^-1), moderate (12–20 m·s^-1), and strong (≥20 m·s^-1) (Johns and Doswell, 1992). As illustrated in Figure 6, the median values of vertical wind shear for the three patterns of SDHP are 17.0 m·s^-1 for the LV, 13.0 m·s^-1 for the LT, and 16.0 m·s^-1 for the EW. This indicates that during SDHP events in the Tianshan Mountains, moderate vertical wind shear intensity is more prevalent. Among these patterns, the LV and EW patterns are classified as exhibiting moderately high deep-layer wind shear, while the LT pattern belongs to moderately low deep-layer wind shear. In terms of maximum values, all three circulation airflow patterns reach a strong level.

Figure 6

Figure 6. Distribution of 0–6 km vertical wind shear under different circulation patterns in the Western Tianshan Mountains.

The boxplot for the LT pattern shows the widest distribution, whereas that for the EW pattern is the narrowest. This suggests that the vertical wind shear range for LT is greater, whereas the SDHP processes influenced by the EW have smaller variations.

5.3.2 Atmospheric stability

Atmospheric stability is a crucial factor for assessing the occurrence of thunderstorm weather during the summer months. It is specifically indicated by the temperature difference (△T₈₅) between high altitudes (500 hPa) and low (850 hPa) altitudes under conditions of deep moist convection. A larger value of △T₈₅ signifies stronger conditional instability, while a smaller value indicates weaker instability.

Figure 7 illustrates the corresponding △T₈₅ values for three patterns of influencing systems: 28 °C for LV, 23 °C for LT, and 28 °C for EW, all of which are significantly higher than the 20 °C–21 °C observed under moist neutral stratification. Furthermore, the maximum (minimum) values and the upper (lower) quartiles exceed the temperatures corresponding to neutral stratification, presenting a clear conditional instability. The box-plot for the LT pattern is the widest and located at a significantly lower position, indicating a greater dispersion of △T₈₅ values. However, the median △T₈₅ for LT is substantially smaller than that of the other two patterns.

Figure 7

Figure 7. Distributions of △T₈₅ (850–500 hPa) and dew point temperature (T_d) under different circulation patterns.

In order to gain a deeper understanding of the atmospheric dynamic and thermal characteristics triggered by short-duration heavy rainfall under different circulation airflow patterns, three typical short-duration heavy rainfall processes were selected for analysis, as shown in Table 3.

Table 3

Table 3. Overview of three short-duration heavy rainfall processes.

Figure 8 illustrates the vertical distribution of pseudo-equivalent potential temperature (θ_se) across three distinct circulation airflow patterns. A consistent “high–low–high” pattern in θ_se is observed from the top to the bottom of the atmosphere. In the upper levels, the LV pattern exhibits the most prominent high-value region (θ_se > 340 K), followed by LT, whereas EW displays a relatively thinner high-value layer. In contrast, the mid-level low-value regions (θ_se < 330 K) show an opposite trend: EW features the thickest low-value zone, followed by LT, while LV shows no significant large-area low-value region in the middle atmosphere.

Figure 8

Figure 8. Distribution of θse under different circulation patterns in the Western Tianshan Mountains (a) 28 June 2015 (b) 19 June 2016 (c) 28 June 2016.

The secondary high-value region below 700 hPa (330 K ≤ θ_se ≤ 340 K) is markedly thinner than the counterpart of upper high-value layer, with LV registering the highest θ_se within this range. This analysis indicates that the LT pattern experiences considerable heating in both the upper and lower atmospheric layers. However, the relatively weak cold air in the mid-level poses a significant risk for atmospheric instability and subsequent convection. Although the lower temperature zone is situated in the mid-level under LV pattern, the layer’s limited thickness allows warmer air from below to easily penetrate and merge with the warmer air above, potentially triggering intense convective weather.

In comparison, EW pattern is characterized by a thicker low-temperature zone in the mid-level and less pronounced lower-level heating, resulting in weaker convective intensity. This interpretation is further supported by the maximum hourly rainfall intensity data presented in Table 3.

5.3.3 Water vapor

Water vapor content and convergence are essential for the generation of SDHP. Key indicators, including the surface dew point (T_d), specific humidity, and water vapor flux divergence—collectively reveal a clear moisture gradient among the three circulation patterns.

As illustrated in Figure 7, the median T_d is 14 °C for the LT pattern and 13 °C for the LV pattern, while the EW pattern is significantly drier, with a median T_d of only 8 °C. The wider box-plot range for EW also reflects greater variability in moisture conditions, likely due to variations in the system’s position and intensity.

Specific humidity further highlights differences. The LV pattern shows values of 12–14 g·kg^-1, the LT pattern ranges between 10 and 12 g·kg^-1, and the EW pattern remains below 10 g·kg^-1. In terms of dynamic moisture transport, both LT and LV patterns exhibit consistent water vapor flux convergence, approximately −2 to −5 g·hPa^-1·m^-2·s^-1, whereas the EW pattern shows no significant convergence.

In summary, the LT and LV patterns are characterized by higher moisture content and more favorable convergence conditions, making them more likely to produce SDHP compared to the drier and less organized EW pattern.

5.3.4 Potential energy

The Convective Available Potential Energy (CAPE) and Convective Inhibition Energy (CIN) of three circulation airflow patterns of SDHP were analyzed. These reveal their significance as indicators of the potential occurrence and intensity of deep moist convection. From the box plot illustrating the distribution of CAPE (Figure 9a), it is evident that the maximum value, median and upper quartile of the LV, LT and EW patterns increase successively. This trend indicates that the EW pattern of SDHP requires the highest CAPE value, while the LV pattern only needs a relatively small CAPE value to trigger the occurrence of SDHP, followed by an intermediate requirement for the LT pattern. Additionally, the sample distribution trend indicates the box representing the EW pattern exhibits greater width, reflecting a more dispersed distribution of CAPE. In contrast, the Box corresponding to the LT pattern is narrower, indicating a more concentrated distribution.

Figure 9

Figure 9. Distribution of potential energy parameters under different circulation patterns (a) CAPE (b) CIN.

CIN serves as a prerequisite for the lifting trigger intensity necessary for the formation of deep moist convection. As illustrated in Figure 9b, which depicts CIN distributions associated with three distinct patterns of SDHP events, it is clear that median CIN values are lowest for the EW pattern and highest for the LT pattern, with intermediate values observed in relation to the LV pattern. In terms of minimum CIN value distributions, the LV pattern exhibits the highest value, followed by the LT pattern, while the EW pattern has the lowest value. This finding suggests that the strong synoptic-scale lifting and low-level convergence inherent to the LV pattern can effectively erode a stronger capping inversion. This allows convection to initiate in an environment with higher residual CIN, which, once broken, releases the built-up instability, contributing to the sustainment of stronger convective weather.

6 Conclusions and discussion

Based on a decade of observational and reanalysis data, this study systematically analyzes the primary circulation flow patterns and environmental parameter characteristics associated with short-duration heavy precipitation (SDHP) during the warm seasons in the Western Tianshan Mountains, Xinjiang. The main conclusions are as follows:

1. Three dominant circulation patterns—Low-Trough (LT), Low-Vortex (LV), and Eastward-moving Waves (EW)—are responsible for most SDHP events. The LT and LV patterns can be further classified into Central Asian (CALT/CALV) and West Siberian (WSLT/WSLV) types based on the location of the trough or vortex.

2. SDHP exhibits a pronounced diurnal cycle, peaking from afternoon to early evening across all patterns. The primary precipitation belts are concentrated on windward slopes and in valleys. Their specific location shifts northward under the stronger westerlies of the WSLT/WSLV pattern and southward under the weaker southwesterlies of the CALT/CALV pattern.

3. The LV pattern provides the most favorable dynamic forcing with the strongest upper- and low-level jet coupling and the most extensive jet coverage, supporting the most widespread and intense precipitation. The LT pattern offers moderate dynamic support, while the EW pattern relies on weak upper-level forcing, resulting in the most limited precipitation. Moderate deep-layer vertical wind shear is prevalent during SDHP events. The LV and EW patterns are characterized by moderately high shear, while the LT pattern exhibits moderately low shear.

4. A consistent “high-low-high” vertical structure of θse is observed across all patterns. However, The LV pattern, with a prominent upper-level high θse and a shallow mid-level cool layer, is most favorable for intense convection. The LT pattern features significant upper- and lower-level heating with a distinct mid-level cool zone, enhancing instability. The EW pattern has a thick mid-level cool layer and weak low-level heating, resulting in the weakest convective potential. Low-level moisture and energy parameters distinctly separate the patterns. The LT and LV patterns are associated with higher low-level moisture content. The drier EW pattern requires greater CAPE) and lower CIN to initiate convection.

The circulation patterns identified in this study share dynamical similarities with systems known to produce heavy precipitation in other parts of the world, such as MCSs over the Great Plains of the United States (Houze, 2004; Schumacher and Johnson, 2006). Owing to relying on station-based precipitation and sounding data, this study primarily imposes certain limitations on sample selection and spatial coverage. Future work will incorporate high-resolution gridded data (e.g., CLDAS) to improve spatial representation. Additionally, in subsequent research, we will consider utilizing minute-scale data to screen SDHP processes in arid regions (Jiang et al., 2025; Yu et al., 2007), thereby including non-hourly events and ensuring the comprehensiveness of sample selection. Furthermore, the complex terrain of the western Tianshan Mountains and the configuration of jets and moisture transport under different weather systems warrant more in-depth investigation to better understand the mechanisms maintaining heavy precipitation and to refine forecasting models.

Building upon this foundation of a 10-year analysis of SDHP in the western Tianshan Mountains, future work will expand the dataset to over 30 years. This will facilitate a more comprehensive investigation into the long-term characteristics and climatic trends of these events, ultimately to provide scientific support for regional economic resilience and development.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

JL: Visualization, Conceptualization, Writing – original draft, Formal Analysis, Project administration. XZ: Methodology, Writing – review and editing, Conceptualization, Funding acquisition, Supervision. CJ: Visualization, Validation, Resources, Writing – review and editing, Data curation. ZT: Writing – original draft, Investigation, Validation, Visualization, Software. TB: Validation, Resources, Writing – original draft. YL: Resources, Data curation, Writing – original draft.

Funding

The author(s) declared that financial support was received for this work and/or its publication. The authors are thankful for the assistance and good suggestions of Dr. Zhang Yun Hui, Dr. Chen Jing, and other colleagues who took part in this work. This research was funded by the Open Research Project of the Key Laboratory of Hydrology and Meteorology, China Meteorological Administration (23SWQXM026), Scientific Research Project of Xinjiang Meteorological Bureau (ZD202505), Key Laboratory Opening Foundation of Xinjiang Uygur Autonomous Region (2023D04048), Science and Technology Development Fund of Institute of Desert Meteorology, Urumqi, China Meteorological Administration (KJFZ202504), and Tianshan Mountains Talent Project (2022TSYCLJ0003). The authors declare that they have no financial conflicts of interests.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Alila, Y. (2000). Regional rainfall depth-duration-frequency equations for Canada. Water Resour. Res. 36 (7), 1767–1778. doi:10.1029/2000wr900046

CrossRef Full Text | Google Scholar

Chen, J., Zheng, Y. G., Zhang, X. L., and Zhu, P. J. (2013). Analysis of the climatological distribution and diurnal variations of the short-duration heavy rain and its relation with diurnal variations of the MCSs over China during the warm season. Acta Meteorol. Sin. 71 (3), 367–382. doi:10.11676/qxxb2013.030

CrossRef Full Text | Google Scholar

Ding, Y. (2013). Contributions of prof.Shiyan tao to the study of formation conditions and mechanisms of heavy rainfall in China. Chin. J. Atmos. Sci. 38 (4), 616–626.

Google Scholar

Doswell, C. A. (2001). Severe convective storms—an overview. Am. Meteorological Soc., 1–26. doi:10.1007/978-1-935704-06-5_1

CrossRef Full Text | Google Scholar

Doswell, C. A. I., and Evans, J. S. (2003). Proximity sounding analysis for derechos and supercells: an assessment of similarities and differences. Atmos. Res. 67, 117–133. doi:10.1016/s0169-8095(03)00047-4

CrossRef Full Text | Google Scholar

Doswell, C. A. I., Brooks, H. E., and Maddox, R. A. (1996). Flash flood forecasting: an ingredients-based methodology. Weather Forecast. 11 (4), 560–581. doi:10.1175/1520-0434(1996)011<0560:fffaib>2.0.co;2

CrossRef Full Text | Google Scholar

Ducrocq, V., Ricard, D., Lafore, J., and Orain, F. (2002). Storm-scale numerical rainfall prediction for five precipitating events over France: on the importance of the initial humidity field. Weather Forecast. 17 (6), 1236–1256. doi:10.1175/1520-0434(2002)017<1236:ssnrpf>2.0.co;2

CrossRef Full Text | Google Scholar

Ducrocq, V., Nuissier, O., Ricard, D., Lebeaupin, C., and Thouvenin, T. (2008). A numerical study of three catastrophic precipitating events over southern France. II: mesoscale triggering and stationarity factors. Q. J. R. Meteorological Soc. 134 (630), 131–145. doi:10.1002/qj.199

CrossRef Full Text | Google Scholar

Fujibe, F., Sakagami, K., Chubachi, K., and Yamashita, K. (2002). Surface wind patterns preceding short-time heavy rainfall in Tokyo in the afternoon of midsummer days. Weather 49, 395–405.

Google Scholar

Hao, Y., Yao, Y. Q., Zheng, Y. Y., and Lu, J. (2012). Multi-scale analysis and nowcasting of short-time heavy rainfall. Meteorol. Mon. 38 (8), 903–912.

Google Scholar

Houze, R. A. (2012). Orographic effects on precipitating clouds. Rev. Geophys. 50 (1), RG1001. doi:10.1029/2011rg000365

CrossRef Full Text | Google Scholar

Houze, R. A. (2004). Mesoscale convective systems. Rev. Geophys. 42 (4), RG4003. doi:10.1029/2004rg000150

CrossRef Full Text | Google Scholar

Iii, C. A. D. (1987). The distinction between large-scale and mesoscale contribution to severe convection: a case study example. Weather and Forecast. 2 (1), 3–16. doi:10.1175/1520-0434(1987)002<0003:tdblsa>2.0.co;2

CrossRef Full Text | Google Scholar

Jiang, X. A., and Wang, M. Z. (2015). Characteristics of the raindrop spectrum during a short-time strong precipitation episode in the yili river valley. Desert Oasis Meteorology 9 (5), 56–61.

Google Scholar

Jiang, R. Y., Cui, X. P., Zhang, D. L., Zhou, X. Y., and Pei, Z. L. (2025). The dominant synoptic patterns under which frequent extreme hourly rainfall events occurred over southwest China during the warm seasons of 1981–2020. Q. J. R. Meteorological Soc. 151, e5035. doi:10.1002/qj.5035

CrossRef Full Text | Google Scholar

Johns, R. H., and Doswell, C. A. I. (1992). Severe local storms forecasting. Weather Forecast. 7 (4), 588–612. doi:10.1175/1520-0434(1992)007<0588:slsf>2.0.co;2

CrossRef Full Text | Google Scholar

Kato, R., Shimizu, S., Shimose, K. I., and Iwanami, K. (2017). Very short time range forecasting using CReSS-3DVAR for a Meso-γ-Scale, localized, extremely heavy rainfall event: Comparison with an extrapolation-based nowcast. J. Disaster Res. 12 (5), 967–979. doi:10.20965/jdr.2017.p0967

CrossRef Full Text | Google Scholar

Li, H. Q., Cui, X. P., and Zhang, D. L. (2017). A statistical analysis of hourly heavy rainfall events over the beijing metropolitan region during the warm seasons of 2007–2014. Int. J. Climatol. 37 (11), 4027–4042. doi:10.1002/joc.4983

CrossRef Full Text | Google Scholar

Li, J. G., Yang, L. M., Liu, W., and Jiang, C. L. (2019). Spatiotemporal distribution characteristics of mesoscale convective systems producing short-duration heavy rainfall over the tianshan Mountain area. Adv. Meteorology 2019 (1), 1–16. doi:10.1155/2019/9090659

CrossRef Full Text | Google Scholar

Lin, Y., Chiao, S., Wang, T., Kaplan, M. L., and Weglarz, R. P. (2001). Some common ingredients for heavy orographic rainfall. Weather Forecasting 16 (6), 633–660. doi:10.1175/1520-0434(2001)016<0633:scifho>2.0.co;2

CrossRef Full Text | Google Scholar

Lin, Y., Colle, B. A., and Yuter, S. E. (2013). Impact of moisture flux and freezing level on simulated orographic precipitation errors over the Pacific northwest. J. Hydrometeorol. 14 (1), 140–152. doi:10.1175/jhm-d-12-019.1

CrossRef Full Text | Google Scholar

Markowski, P., and Richardson, Y. (2010). Mesoscale meteorology in midlatitudes. Wiley.

Google Scholar

Neiman, P. J., Ralph, F. M., Moore, B. J., and Zamora, R. J. (2014). The regional influence of an intense sierra barrier jet and landfalling atmospheric river on orographic precipitation in northern California: a case study. J. Hydrometeorol. 15 (4), 1419–1439. doi:10.1175/jhm-d-13-0183.1

CrossRef Full Text | Google Scholar

Osborn, H. B., Renard, K. G., and Simanton, J. R. (1979). Dense networks to measure convective rainfall in the Southwestern United States. Water Resour. Res. 15 (6), 1701–1711. doi:10.1029/wr015i006p01701

CrossRef Full Text | Google Scholar

Schumacher, R. S., and Johnson, R. H. (2006). Characteristics of US extreme rain events during 1999–2003. Weather Forecast. 21 (1), 69–85. doi:10.1175/waf900.1

CrossRef Full Text | Google Scholar

Shein, K. A. (2016). Central Asian weather: flying conditions in this continental interior can be ever changing and severe. Brisbane, QLD: Queensland Communications Corporation.

Google Scholar

Wilks, D. S. (2011). Statistical methods in the atmospheric sciences, 100. Academic Press.

Google Scholar

Xu, Y., Luo, B., Li, N., Chen, X., Wang, L., Zhang, W., et al. (2023). Construction and application of an integrated quality control system for multi-source meteorological observation data. Meteorological Environ. Res. 14 (3), 1–7.

Google Scholar

Yan, H., Xiaoding, Y., and Yunhui, Z. (2022). Characteristic analysis of main weather pattern and environmental parameter of short-time heavy precipitation in Xinjiang. Meteorological Sci. Technol. 50 (1), 85–93.

Google Scholar

Yang, L., Guan, X., and Zhang, Y. (2018). Atmospheric circulation characteristics of precipitation anomaly in arid regions in central Asia. Arid Zone Res. 35 (2), 249–259.

Google Scholar

Yin, D. (2007). A comparative analysis of doppler radar products in hail, gale and short-time heavy rainfall in Jiangxi. Meteorol. Mon. 33 (3), 48–53.

Google Scholar

Yu, X. D. (2013). Nowcasting thinking and method of flash heavy rain. Torrential Rain Disasters 32 (3), 202–209.

Google Scholar

Yu, R., Xu, Y., Zhou, T., and Li, J. (2007). Relation between rainfall duration and diurnal variation in the warm season precipitation over central eastern China. Geophys. Res. Lett. 34 (13), L13703. doi:10.1029/2007gl030315

CrossRef Full Text | Google Scholar

Zhang, X., Zwiers, F. W., Li, G., Wan, H., and Cannon, A. J. (2017). Complexity in estimating past and future extreme short-duration rainfall. Nat. Geosci. 10 (4), 255–259. doi:10.1038/ngeo2911

CrossRef Full Text | Google Scholar