Extreme Rainfall Events in July Associated with the Daily Asian-Pacific Oscillation in the Sichuan-Shaanxi Region of China

Abstract

Rainfall variability and its underlying physical mechanisms are crucial for improving the predictive accuracy of July rainfall patterns in the Sichuan-Shaanxi (SS) region of Southwestern China. This study utilized observational 24 h accumulated rainfall data from China in conjunction with reanalysis products sourced from the European Centre for Medium-Range Weather Forecasts (ECMWF). The purpose of this study was to elucidate the relationship between daily variations in the daily Asian-Pacific Oscillation (APO), atmospheric circulation, and daily rainfall patterns in the SS region, and to evaluate the impact of atmospheric circulation anomalies on these relationships. The results reveal a discernible intensification in the sea–land thermal contrast associated with atmospheric circulation anomalies transitioning from the daily extremely low APO (ELA) to the extremely high APO (EHA) days. These conditions lead to an increased presence of water vapor and widespread anomalies in rainfall that exceed normal levels in the SS region. Concurrently, the increase in stations experiencing extreme rainfall events (EREs) accounts for 21.3% of the overall increase in stations experiencing rainfall. The increase in rainfall amount contributed by EREs (RA-EREs) accounts for 73.5% of the overall increase in the total rainfall amount (TRA) across the SS region. Specifically, heavy rainfall (HR) and downpour rainfall (DR) during EREs accounted for 65.7% (HR) and 95.3% (DR) of the overall increase in the TRA, respectively. Relative to the ELA days, there was a substantial 122.6% increase in the occurrence frequency of EREs and a 23.3% increase in their intensity. The study suggests that the daily APO index emerges as a better indicator of July rainfall events in the SS region, with EREs significantly contributing to the overall increase in rainfall in this region. These findings indicate the importance of improving predictive capabilities for daily variability in the APO index and their correlation with rainfall events in the SS region. The results may inform the development of effective adaptation and mitigation strategies to manage the potential impacts of EREs on agriculture, water resources, sustainable development, and infrastructure in the region.

1. Introduction

During the summer months, the Sichuan-Shaanxi (SS) region and its surrounding regions frequently experience extreme rainfall events (EREs), which can result in devastating floods. These events have a profound impact on local agriculture, industry, communication and transportation networks, ecosystems, and socio-economic development. Tragically, they can also lead to loss of life [1,2]. Encompassing an area of over 0.4 million km², this region has witnessed a significant increase in economic losses due to floods, rising from 0.05 to 14 billion RMB between 1984 and 2010 [3]. Furthermore, the rate of economic loss growth surged by 80.6% from the 1980s to the 2010s [4]. Notably, the impacts of extreme rainfall and resulting floods are particularly pronounced in July compared to other months, with economic losses exceeding 1.2 billion RMB and affecting an area of 142.6 thousand hm² and 55.4 thousand hm² of crops, respectively [4,5,6].

Moreover, the Sichuan-Shaanxi (SS) region and its surrounding regions encompass the Ta-pa and Qinling Mountains (QMs) [7,8], which serve as a geological, geochemical, and physical geographical boundary between northern and southern China. These mountains form a climatic transition zone and exert a significant influence on regional weather patterns, particularly in terms of rainfall. Consequently, they hold significant scientific importance in the field of Chinese climate change research [8,9]. Furthermore, comprehending the variability of extraordinary rainfall events and their underlying physical mechanisms in July is crucial for refining the predictive capabilities concerning extreme rainfall occurrences. These insights play a pivotal role in mitigating the impacts of climate-related disasters arising from abnormal rainfall events.

Numerous prior studies have shed light on the distinctive regional characteristics of extreme rainfall occurrences in Sichuan-Shaanxi, which are closely linked to its topography, particularly the elevation disparity between western and eastern Sichuan-Shaanxi [1]. The terrain experiences a notable decrease in elevation in central Sichuan and southwestern Shaanxi, demarcating the boundary of the eastern Tibetan Plateau. This topographic feature is associated with substantially more summer rainfall in the SS region compared to its western counterpart [1,10,11,12]. Research by Jiang et al. (2015) documented interannual variations in the frequency of heavy rainfall events in the Sichuan region [13]. Additionally, Cheng et al. (2016) underscored the unique topography of the Sichuan basin, which funnels low-level water vapor around the eastern slope of the Tibetan Plateau, resulting in intensified rainfall within the Sichuan basin and its bordering areas in Shaanxi [14,15,16]. Qi et al. (2021, 2022) delineated the spatiotemporal distribution of rainfall in the Sichuan basin, attributing it to interactions between the distinct topography of the basin and various modes of water vapor transport from lower latitudes [17,18]. Furthermore, Liang et al. (2021) identified summer (July–September) as the peak season for hazards triggered by extreme rainfall in Shaanxi Province, particularly in the southern QMs [10]. Lai et al. (2017) revealed the impact of warm, moist air transport from low-latitude oceans on summer rainfall anomalies in the Sichuan region and its surrounding areas [19].

Furthermore, Zhao et al. (2007) defined the Asian-Pacific Oscillation (APO) as a large-scale zonal tropospheric temperature contrast between Asian land and the North Pacific, and delved into its implications for summer monsoon rainfall in East Asia and South Asia [20]. Subsequent observations, simulations, and predictions have consistently indicated that this thermal contrast effectively reflects the variability of the East Asian monsoon rainfall across various timescales [21,22,23,24,25]. Furthermore, it has been demonstrated that the APO can enhance the predictive capacity for interannual variability of rainfall in China [26]. In particular, Chen et al. (2016) revealed that the improvement in predicting East Asian summer monsoon (EASM) rainfall is significantly influenced by the APO in addition to ENSO [27]. Liao and Zhao (2023) employed the daily APO index to reveal an interdecadal shift in extreme summer rainfall within the Huaihe River basin (HRB) in central-eastern China [28].

While significant progress has been made in previous research on the thermal contrast between the Eurasian landmass and the adjacent oceans, particularly the western North Pacific, and its impact on rainfall in China, there remains a dearth of studies investigating the connection between the thermal contrast and rainfall patterns in Southwestern China (i.e., the SS region) using the daily APO index. The objectives of this study are as follows: (1) to identify the relationships between daily rainfall, daily occurrences of extreme rainfall, and the daily APO during July in the SS region; (2) to assess the impact of atmospheric circulation anomalies on these relationships. By addressing these inquiries, this study seeks to provide valuable insights into the fundamental mechanisms governing atmospheric circulation and its associated rainfall.

The structure of this article is outlined as follows: In Section 2, we provide a description of the datasets and methods employed in this study. Section 3 focuses on the examination of rainfall anomalies associated with the July daily APO index. Section 4 analyzes the atmospheric circulation anomalies linked to the July daily APO index. In Section 5, we present our conclusions and engage in discussions on the implications of our findings.

2. Materials and Methods

2.1. Study Area

The study area is situated in the Sichuan-Shaanxi (SS) region (Figure 1), encompassing central-eastern Sichuan and southwestern Shaanxi (102°–108° E, 29.5°–34.5° N), and includes a total of 89 stations. Sichuan Province (92°21′–108°12′ E, 26°03′–34°19′ N) is located in Southwestern China, with an annual mean temperature ranging from 16.1 °C to 18.5 °C. Eastern Sichuan receives a mean annual rainfall of 1200 mm, with approximately 30% of the total rainfall occurring in July. This region is characterized by a basin, which is one of the four largest basins in China, extending southeastward into Chongqing and surrounded by the Ta-pa and Qinling mountains to the north [1]. Additionally, Shaanxi Province (105°29′–111°15′ E, 31°42′–39°35′ N) is located in Northwestern China, with an annual mean temperature ranging from 6.5 °C to 16.6 °C. Southern Shaanxi receives a mean annual rainfall of 600 mm, with approximately 35% of the total rainfall occurring in July. The southern region, bordered by the Ta-pa and Qinling mountains, features a north subtropical monsoon climate that is much more humid than its central-northern counterpart [10]. Consequently, the area is prone to various hazards due to intensive rainfall and steep slopes.

2.2. Data and Method

This study calculated the APO index utilizing the daily mean air temperatures sourced from the European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis spanning from 1979 to 2014 [29], with a horizontal resolution of 2.5° × 2.5°. The July daily APO index is derived by projecting the projection of anomalous eddy temperature fields onto the July mean APO pattern, following the methodology outlined by Liao and Zhao (2023) [28]. In addition, gauge data from 2419 observational stations across China, alongside gridded rainfall data with a horizontal resolution of 0.25° × 0.25° from the National Meteorological Information Centre (NMIC) of China, are utilized [30]. Daily atmospheric variables are incorporated to elucidate the relationship between the daily APO index and July rainfall patterns in Southwestern China.

Extreme rainfall events (EREs) are identified through the utilization of rainfall percentiles, a method extensively discussed in previous research [28,31,32,33]. Initially, the daily rainfall gauge data from 1 July to 31 July for each year are arranged in ascending order ($x_1,x_2,x_3,\dots ,x_n$). The probability $\left(p\right)$ of selecting a value equal to or less than the rank of a specific value $x_{my}$ is estimated using the following formulae.

$$p={\displaystyle \frac{m-0.31}{n-0.38}}$$

(1)

$$x_{ym}={\displaystyle \frac{{\sum }_{t=1}^y{x}_{mt}}{y}}$$

(2)

$$x_i \ge { x}_{ym}$$

(3)

In the calculation, $p$ represents the 95th percentile; $n$ stands for the number of samples (which, for the period of 1 July to 31 July, equals 31 days);$y$ denotes the number of years (spanning 30 years from 1981 to 2010); $m$ represents the record number within the sample size $n$; $x_{mt}$ indicates the values of the rainfall specified by percentile ranks $p$ [31,32,34]; and $x_{ym}$ signifies the 30-year mean value of $x_m$. At each monitoring station, the 95th percentile ($p$ = 95%) of the daily rainfall distribution was estimated from 1 July to 31 July, 1981–2010. If the daily rainfall ($x_i$) at any station surpassed this threshold ($x_{ym}$), it qualified as a daily extreme rainfall event (ERE). Unlike fitting a statistical distribution such as gamma, Equation (1) is straightforward and may circumvent assumptions about the underlying distribution [28,31,32,34,35,36]. Each day, at every rainfall gauge station, the frequency of extreme rainfall is defined as the percentage of days with EREs out of the total rainfall days within a given area. Meanwhile, the extreme rainfall intensity is described as the average daily rainfall on days with EREs [8,28,34,37,38].

Utilizing observational rainfall data from 2400 stations, we established gridded extreme daily rainfall datasets with a horizontal resolution of 0.10° × 0.10°. This was achieved through climatological optimal interpolation (OI), following the methodology outlined by Shen et al. (2010) [30]. To explore rainfall features, we categorized daily rainfall amounts into six types: light rainfall (LR; 0.1–9.9 mm), moderate rainfall (MR; 10.0–24.9 mm), heavy rainfall (HR; 25.0–49.9 mm), torrential rainfall (TR; 50.0–99.9 mm), downpour rainfall (DR; 100.0–249.9 mm), and extraordinary rainfall (ER; ≥250.0 mm).

In the realm of statistical analysis, correlation and composite analyses are employed to scrutinize the relationships between two variables. Statistical significance is determined utilizing Student’s t-test, ensuring robustness in our findings. Consistent with the conclusions drawn by Liao and Zhao (2023) [28], all significances surpassed the 90% confidence level, unless explicitly stated otherwise.

3. Rainfall Anomalies Associated with the July Daily APO Index

The positive or negative APO index serves as an indicator of a stronger or weaker tropospheric thermal contrast between the East Asian land and the North Pacific, thereby exerting influence on the East Asian summer monsoon rainfall [20]. Figure 2 shows the characteristics of the standardized time series of the daily APO index in July for the period 1979–2014. In this figure, the daily APO index clearly shows the multiple timescale variations, such as the sub-seasonal and interannual variations. The present study focuses on the effect of the extreme high or low daily APO index. Consequently, the daily APO indices higher (lower) than 1.5 (−1.5) standard deviations were selected to perform further analyses, in which there are 68 days with extremely high daily APO cases (referred to as the EHA days hereinafter) and there are 47 days with extremely low daily APO cases (referred to as the ELA days hereinafter).

During periods of the EHA days (Figure 3a), the large rainfall was predominantly observed in the central-northeastern Sichuan and southwestern Shaanxi regions (102°–108° E/29.5°–33° N), with the central value exceeding 20 mm·d⁻¹. Conversely, the rainfall amounts below 4 mm·d⁻¹ were observed to the southeast of the Chongqing region. In contrast, during periods of the ELA days (Figure 3b), reduced rainfall was observed compared to the EHA days, with larger rainfall occurring in the southeast (102°–105° E/28°–30° N) and northern (104°–105° E/31°–33° N) Sichuan regions, with the central value exceeding 10 mm·d⁻¹. Conversely, rainfall amounts below 4 mm·d⁻¹ were observed north of the 30° N region. These findings indicate that there is a northward shift of July rainfall from the southern Sichuan regions to the central-northeastern Sichuan and southwestern Shaanxi regions during the EHA days. This observation aligns with previous studies that have highlighted significant summer rainfall over the Sichuan-Chongqing-Shaanxi region and its surrounding areas [1,13,17,18].

Figure 4a shows the composite difference in July rainfall anomalies between the EHA and ELA days. In this figure, significantly positive rainfall anomalies are predominantly observed in the central-eastern Sichuan and southwestern Shaanxi regions (102°–108° E, 29.5°–34.5° N), commonly referred to as the SS region, with a central value exceeding 10 mm·d⁻¹. Conversely, negative rainfall anomalies are generally observed in the southeast of Sichuan and Chongqing regions, referred to as the SC region, with a central value below −6 mm·d⁻¹.

This pattern illustrates that, in correspondence with the interannual variability of the July daily APO index between the EHA and ELA days, there is a significant increase in rainfall in the SS region, while a decrease is observed in the southeast of the SC region. During a positive APO phase, rainfall intensifies between 29.5° N and 34.5° N, coupled with a reduction in rainfall between 28° N and 32° N across Southwestern China. Moreover, these observations suggest a significant northward shift in the positive anomalies of July daily rainfall across Southwestern China, corresponding to the positive daily APO index in July. This trend is consistent with prior research indicating that, during periods of a positive APO index on both the interannual and interdecadal scales, the rain belt in central-eastern China experiences a northward shift [20,28].

Figure 4b offers further insight into the composite difference in the July daily extreme rainfall anomalies between the EHA and ELA days. The patterns reveal significant positive and negative anomalies of extreme rainfall resembling those depicted in Figure 4a. The positive extreme rainfall anomalies are observed in the SS region, while the negative extreme rainfall anomalies are evident on the eastern peripheries of the SS region. These findings suggest a heightened occurrence of heavy rainfall over the SS region during the EHA days.

As both the July total rainfall amount (TRA) and the rainfall amount contributed by extreme rainfall events (RA-EREs) in the SS region have experienced an interannual increase, as depicted in Figure 4, it raises the question of whether the interannual increase of RA-EREs has significantly influenced the overall increase in TRA in this area, which consists of 89 stations (Figure 1). As illustrated in

Table 1. The July total rainfall amount (TRA; unit: mm·day⁻¹), rainfall amount contributed by extreme rainfall events (RA-EREs; unit: mm·day⁻¹), stations experiencing rainfall in the SS region, and associated differences between the EHA and ELA days. (Note: ** $p=0.05$, * $p=0.1$).

	Period (Day)	Light (LR)	Moderate (MR)	Heavy (HR)	Torrential (TR)	Downpour (DR)	Extraordinary (ER)	TOTAL
TRA (mm·day−1)	EHR	72.1	123.0	159.4	155.3	143.9	27.0	680.7
	ELR	61.3	76.7	78.2	80.2	20.3	-	316.7
	Difference	10.8	46.3 **	82.2 **	75.1 *	123.6 *	-	364.0 **
RA-EREs (mm·day−1)	EHR	-	10.1	94.9	137.9	131.4	27.0	401.3
	ELR	-	6.0	40.9	73.3	13.6	-	133.8
	Difference	-	4.1	54.0 **	64.6	117.8 *	-	267.5 **
Rainfall (stations)	EHR	1765	508	303	156	70	6	2808
	ELR	1037	232	103	55	7	0	1434
	Difference	728	276	200	101	63	6	1374
EREs (stations)	EHR	-	34	167	139	64	6	410
	ELR	-	14	49	50	5	0	118
	Difference	-	20	118	89	59	6	292

, from the ELA to the EHA days, the number of stations experiencing rainfall increased by 1374, while the stations experiencing extreme rainfall events (EREs) increased by 292, accounting for 21.3% of the overall increase in rainfall stations. Moreover, the July TRA significantly increased by 364 mm·day⁻¹. Concurrently, the July RA-EREs significantly increased by 267.5 mm·day⁻¹, representing 73.5% of the overall increase in TRA.

To gain deeper insights into rainfall characteristics, we conducted an examination of various rainfall classifications in the SS region. As shown in Table 1, for each type of the July TRA, there is a notable disparity in station counts between the EHA and ELA days. Specifically, there was a significant increase in station numbers during the EHA days compared to the ELA days, with increments of 728 (LR), 276 (MR), 200 (HR), 101 (TR), 63 (DR), and 6 (ER), respectively. Additionally, there is also a notable disparity in station counts between the EHA and ELA days for each type of the July EREs, with increments of 20 (MR), 118 (HR), 89 (TR), 59 (DR), and 6 (ER), respectively. These increments account for 2.7% (LR), 42.8% (MR), 44.5% (HR), 58.4% (TR), 93.7% (DR), and 100% (ER) of the overall increase in rainfall stations, respectively. In addition, the July TRA experienced a significant increase, with increments of 46.3 mm·day⁻¹ (MR), 82.2 mm·day⁻¹ (HR), 75.1 mm·day⁻¹ (TR), and 123.6 mm·day⁻¹ (DR), respectively. Simultaneously, the July RA-EREs significantly increased by 54.0 mm·day⁻¹ (HR) and 117.8 mm·day⁻¹ (DR), representing 65.7% (HR) and 95.3% (DR) of the overall increase in TRA.

These findings reveal that EREs play an important role in dominating the interannual increase in TRA in the SS region. EREs contributed to more than half of the observed interannual surge in the TRA. Furthermore, the HR, TR, and DR in EREs individually contributed to more than half of the increase in the TRA with their respective types. Additionally, there was a significant increase in the occurrences of HR and DR during EREs, as well as a significant increase in the occurrences of MR, HR, TR, and DR within overall rainfall from the ELA to the EHA days in the SS region.

To delve deeper into the characteristics of EREs, we performed a comparative analysis of their frequency and intensity variations in the SS region between the EHA and ELA days. We categorized daily extreme rainfall into three types, that is, moderate to heavy rainfall (MHR; 10.0–49.9 mm), above torrential rainfall (ATR; ≥50.0), and above downpour rainfall (ADR; ≥100.0 mm). As shown in

Table 2. As in Table 1, but for the frequency (unit: %·day⁻¹) and intensity (mm·day⁻¹) of EREs in the SS region. (Note: *** $p=0.01$, ** $p=0.05$, * $p=0.1$).

	Period (Day)	MHR	ATR	ADR	TOTAL
Frequency (%·day−1)	EHR	61.8	48.5	23.5	66.1
	ELR	21.7	14.9	6.4	29.7
	Difference	34.1 ***	33.6 *	17.1 *	36.4 **
Intensity (mm·day−1)	EHR	36.4	77.4	131.2	51.4
	ELR	32.5	69.8	125.1	41.7
	Difference	3.9 *	7.6 *	6.1 *	9.7 **

, the occurrence frequency during the ELA days was 29.7%·day⁻¹, whereas it rose to 66.1%·day⁻¹ during the EHA days, marking a significant increase of 122.6% (p = 0.95) relative to the ELA days. Correspondingly, the intensity during the ELA days was 41.7 mm·day⁻¹, whereas it rose to 51.4 mm·day⁻¹ during the EHA days, marking a significant increase of 23.3% (p = 0.95) relative to the ELA days. For each type of extreme rainfall, the differences in the occurrence frequency during the EHA days were 61.8%, 48.5%, and 23.5%·day⁻¹ for the HHR, ATR, and ADR, respectively. These represent increases by 157.1%, 225.6%, and 267.2% relative to the ELA days, respectively. Likewise, the intensity of these extreme rainfall types also shows a remarkable change between the EHA and ELA days; the differences in the intensity during the EHA days were 36.4 mm·day⁻¹, 77.4 mm·day⁻¹, and 131.2 mm·day⁻¹ for the HHR, ATR, and ADR, increasing by12.0%, 10.9%, and 4.9% relative to the ELA days, respectively.

The aforementioned findings reveal that EREs provide a significant contribution to the TRA during the EHA days. Accordingly, corresponding to the interannual shift of the APO from the ELA to the EHA days, there is a significant increase in both the occurrence frequency and intensity of EREs in the SS region.

4. Atmospheric Circulation Anomalies Associated with the July Daily APO Index

Extreme rainfall events associated with the July daily APO index in the SS region are influenced by atmospheric circulation systems [20,28]. Considering that the interannual variation in EREs is a widespread phenomenon in the SS region, investigating the role of large-scale atmospheric circulation anomalies in this variability is essential. Figure 5a depicts the composite middle tropospheric $T^{\prime }$ for the EHA days. Significantly positive anomalies are evident over Eurasia, spanning between 10° N and 50° N, with a central value exceeding 2.0 °C in East Asia. Conversely, significantly negative anomalies are observed over the central North Pacific, with a central value dropping below −2.0 °C in the extratropical central North Pacific. A similar pattern is observed in the composite middle tropospheric $T^{\prime }$ for the ELA days, as illustrated in Figure 5b, albeit with opposite polarity compared to Figure 5a. Figure 5c depicts the composite difference in the middle tropospheric $T^{\prime }$ between the EHA and ELA days. In this figure, the central value of the significantly positive anomalies over Eurasia, spanning from 10° N to 50° N, exceeds 4.0 °C. Conversely, the central value of the significantly negative anomalies over the extratropical central and eastern North Pacific falls below −4.0 °C. Clearly, relative to the July mean high/low APO (figure omitted), the daily extremely high/low APO can delineate the intensified/weakened tropospheric thermal contrasts between East Asia and the North Pacific [20,28].

Figure 6a depicts the latitude–height cross-section of the composite $T^{\prime }$ difference between the EHA and ELA days along the longitude 105° E. Generally speaking, significantly positive anomalies of $T^{\prime }$ are evident north of 25° N, with a central value exceeding 6 °C around 38° N between 300 and 200 hPa. Conversely, weak negative anomalies of $T^{\prime }$ are predominantly observed south of 20° N, with the central value dropping below −0.5 °C, mainly confined to the lower troposphere (below 600 hPa). Figure 6b depicts the longitude–height cross-section of the composite $T^{\prime }$ difference along the latitude 32° N. In this figure, significantly positive anomalies of $T^{\prime }$ manifest west of 130° E, with the central vale exceeding 2.0 °C over the eastern slope of the Tibetan Plateau around 105° E below 200 hPa. Conversely, significantly negative anomalies of $T^{\prime }$ are evident in the troposphere over the central and eastern North Pacific to the east of 140° E, with the central value dropping below −2.0 °C in the mid-upper troposphere (between 700 hPa and 200 hPa). It is evident that the positive/negative daily APO index distinctly reflects the intensified/weakened zonal and meridional tropospheric thermal contrasts between the East Asian landmasses and the adjacent oceans [28].

Figure 7a depicts the composite difference in the daily 150 hPa geopotential height between the EHA and ELA days. From this figure, significantly positive anomalies of geopotential height can be observed over the mid-high latitudes of East Asia, ranging from 25° N to 65° N, with a central value exceeding 160 gpm. Simultaneously, significantly positive anomalies of geopotential height are also noted in the mid-high latitudes of the North Pacific, spanning between 40° N and 60° N, with the central value surpassing 80 gpm. These results indicate an increase in local geopotential height during the EHA days. Conversely, significantly negative anomalies of geopotential height can be observed over the western North Pacific, spanning between 20° N and 35° N. The combination of positive anomalies of geopotential height in the west and negative ones in the east amplifies the zonal pressure gradient between East Asia and the western North Pacific in the upper troposphere, thereby favoring the northeasterly wind anomalies between the positive and negative anomalies of geopotential height. As a result, the mid-upper tropospheric cold air may intrude southward into the SS region.

In Figure 7b, we present the composite difference in the daily 500 hPa geopotential height between the EHA and ELA days. In this figure, significantly positive anomalies of geopotential height appear over the mid-high latitudes of East Asia, ranging from 30° N to 45° N, as well as over the middle latitudes of the North Pacific, spanning between 38° N and 45° N, with the central value exceeding 30 gpm. These findings suggest an increase in the local geopotential height during the EHA days, consistent with the anticyclonic circulation anomaly near 40° N, 110° E at 500 hPa (Figure 8b). Simultaneously, significantly negative anomalies of geopotential height are observed over the mid-low latitudes of East Asia and the western North Pacific, spanning between 5° N and 30° N, with the central value dropping below -20 gpm. Correspondingly, in line with the positive negative anomalies of geopotential height observed in South Asia and the western North Pacific, cyclonic circulation anomalies emerge between 5° N and 30° N, with the anomalous cyclonic center near 25° N, 132° E (Figure 8b). This reflects a weaker Western Pacific Subtropical High (WPSH) during the EHA days than that during the ELA days.

In the lower troposphere (850 hPa) (Figure 7c), it can be seen from this figure that there are significant negative anomalies of geopotential height observed across East Asia and the adjacent oceans (western North Pacific), spanning from 0° N and 65° N. The central value dropping below −30 gpm, particularly between 40° N and 55° N and between 18° N and 27° N, These findings indicate an decrease in the local geopotential height during the EHA days, consistent with the cyclonic circulation anomalies observed near 50° N, 128° E and 20° N, 110° E at 850 hPa (refer to Figure 8c). This pattern favors the strengthening of the easterly winds at the northern flank of the anomalous cyclone circulation.

Figure 8a shows the composite 200 hPa wind anomalies between the EHA and ELA days. In the upper troposphere, anticyclonic circulation anomalies manifest with an anomalous center near 40° N, 105° E. Anomalous northeasterly winds dominate the eastern flank of the anomalous anticyclone, spanning from 30° N to 45° N of East Asia. Governed by the above circulation anomalies, strong northeasterly wind anomalies extend from North China to the SS region, potentially facilitating the southward intrusion of the upper tropospheric cold air from higher latitudes to the SS region. In Figure 8b, corresponding to the positive 500 hPa geopotential height anomaly over East Asia (as shown in Figure 7b), anticyclonic circulation anomalies with the anomalous center near 40° N, 110° E are observed, alongside a cyclonic anomaly emerging in the region, centered near 25° N, 132° E. These circulation anomalies drive the prevalence of strong easterly wind anomalies over South China and the western North Pacific, ranging from 25° N to 35° N, potentially amplifying water vapor transport towards the SS region.

In the lower troposphere (at 850 hPa and 950 hPa) (Figure 8c,d), anomalous cyclonic circulations are observed over the mid-high latitudes of East Asia, with the center of this anomalous circulation located near 50° N, 120° E. Concurrently, prevailing northeasterly wind anomalies extend from North China to the SS region, spanning between 30° N and 40° N. Additionally, easterly wind anomalies dominate from the western North Pacific, passing through South China and into Southwest China. This wind pattern facilitates the transport of warm and wet air masses from the East China Sea (around 25° N, 130° E) towards the SS region.

Corresponding to the positive (negative) anomalies in geopotential height and wind patterns, the water vapor transport field also exhibits consistent characteristics. Figure 9 shows the composite difference in water vapor transport between the EHA and ELA days. In Figure 9a, an anomalous cyclonic circulation can be observed south of 30° N, centered near 18° N, 110° E. Eastward water vapor transport on the northern flanks of the anomalous cyclonic circulation is evident, facilitating the transport of warm and wet air masses from the adjacent oceans towards the SS region. This transport pattern can be observed in both Figure 9a and the lower troposphere (at 850 hPa and 950 hPa) as depicted in Figure 9b,c. Consequently, the northward water vapor flux transports wet air masses from the East China Sea (ECS) towards the SS region, merging with northeasterly water vapor transport anomalies from north of 32° N in the lower troposphere (Figure 9b,c), potentially enhancing rainfall in the SS region (as depicted in Figure 4).

Meanwhile, significantly negative anomalies in water vapor transport divergence were observed between 28° N and 35° N over the SS region in the entire troposphere (Figure 9d–f), with the central value dropping below −0.2 m·day⁻¹. This indicates an increased convergence of water vapor over the SS region. Simultaneously, significantly positive anomalies in water vapor transport divergence were observed in the northern and southern parts of the SS region, facilitating the conveyance of wet air masses towards the SS region and resulting in a higher occurrence of rainfall events in this region.

In addition, significant changes were also observed in the tropospheric vertical motion. Figure 10a presents the longitude–height cross-section of the composite difference of the zonal vertical circulation anomalies and pseudo-equivalent temperature anomalies between the EHA and ELA days, averaged from 30° N to 35° N. In this figure, it is evident that the maximum positive anomalies of 8 K occur between 800 and 600 hPa around 103°–106° E (i.e., the SS region). Concurrently, significant upward motion anomalies manifest in the mid-lower troposphere between 102° E and 108° E. Figure 10b displays the latitude–height cross-section of the composite difference of the meridian vertical circulation anomalies and pseudo-equivalent temperature anomalies between the EHA and ELA days, averaged from 103° E to 108° E. It is notable from this figure that positive pseudo-equivalent temperature anomalies emerge below the 150 hPa level, with the center value exceeding 10 K near the 300 hPa level. Meanwhile, significant upward motion anomalies occur in the mid-lower troposphere between 30° N and 35° N. Furthermore, the lower tropospheric airflow ascends and forms an anomalous upward flow around 30°–35° N (i.e., the SS region), then turns and moves southward in the upper troposphere before eventually descending around 25°–27.5° N.

Upon analyzing the aforementioned data, it becomes evident that the lower tropospheric air mass contains higher internal energy and latent heat energy compared to the air mass situated above it, which possesses higher gravitational potential energy. This energy imbalance prompts the air mass with higher internal energy and latent heat energy to ascend, thereby catalyzing more vigorous convective upward motions over the SS region, approximately situated between 30° and 35° N. These dynamics are further supported by observed vertical circulation anomalies within this region, contributing to a general increase in rainfall over the SS region. Consequently, the extreme daily APO can be considered a reliable indicator of rainfall anomalies within this locale.

5. Discussion and Conclusions

This study utilizes gridded and gauge rainfall data from China, as well as the ERA-Interim reanalysis, to construct the daily APO index following the methodology outlined by Liao and Zhao (2023) [28]. Specifically, this study aims to examine the intricate relationship between the daily APO, atmospheric circulation patterns, and rainfall in the SS region of Southwestern China. Distinguishing itself from previous research that predominantly focused on the correlation between the monthly or seasonal APO and rainfall [20,39,40], this study focuses on analyzing the occurrence of both extremely high and low daily APO events during July, coupled with the corresponding anomalies in atmospheric circulation and rainfall patterns. Furthermore, this research seeks to evaluate the extent to which atmospheric circulation anomalies contribute to this relationship within the SS region.

This study focuses on the SS region, which includes central-eastern Sichuan and southwestern Shaanxi. Summer rainfall in the SS region is much greater than that in western Sichuan (WS) [12,41] and is even slightly higher than that in the Yangtze River basin [42]. Heavy rainfall events occur more frequently in eastern Sichuan (ES) [6]. Additionally, the topography of the region, characterized by basins and plateaus, exhibits significant altitude variations between western Sichuan (WS) and the SS region, playing a crucial role in triggering extreme rainfall over central-eastern Sichuan and its surroundings. The regional-scale anomalous circulation patterns at both upper and lower tropospheric levels are closely related to the topography in the SS region [1]. Enhanced extreme rainfall in the SS region is associated with an anticyclonic anomaly in the upper troposphere over China and cooler sea surface temperatures (SSTs) in the equatorial central Pacific [1]. The frequency of heavy rainfall events in the region is influenced by warm (cool) SST anomalies in the western North Pacific (WNP) [13]. During wet (dry) years, southerly (northerly) anomalies prevail in the SS region, often accompanied by the westward (eastward) displacement of the WNP Subtropical High [43]. This study is based on previous research by focusing on daily-scale analysis, providing a new perspective on the daily signals of atmospheric circulation patterns through more detailed detection and synthesis analysis. It may provide an important reference for predicting occurrences of EREs, potentially mitigating significant economic losses and human fatalities.

The results indicate that the occurrence of higher (lower) daily APO cases in July, which reflects a strengthened (weakened) zonal tropospheric thermal contrast between mid-latitude East Asia and the North Pacific, corresponds to increased (decreased) rainfall over the SS region during that month. Moreover, the study reveals a significant increase in the occurrence of daily EREs over the SS region from the ELA to the EHA days. Corresponding to the EHA days, daily rainfall generally exhibits an increasing trend in the SS region. Additionally, the number of stations experiencing EREs accounts for 21.3% of the overall increase in stations experiencing rainfall, and the July RA-EREs account for 73.5% of the overall increase in the TRA. The July RA-EREs account for more than half of the observed interannual surge in the TRA. In general, the interannual variation in these rainfall events is observed as a widespread phenomenon across the SS region, characterized by a higher occurrence frequency and greater intensity of EREs during the EHA days. Conversely, rainfall events exhibit lower frequency and weaker intensity during the ELA days.

As the daily APO index transitions from its negative phase to the positive phase, it strengthens the zonal and meridional tropospheric thermal contrasts between the East Asian landmass and the adjacent oceans, as highlighted in previous studies [20,28]. During the EHA days, the positive geopotential height anomalies and the anomalous anticyclone circulation appear over East Asia in the mid-upper troposphere. These anomalies are accompanied by significantly northeasterly wind patterns that may facilitate the southward intrusion of the upper tropospheric old air from higher latitudes towards the SS region. Concurrently, the negative geopotential height anomalies and the anomalous cyclone circulation are observed over East Asia and the adjacent oceans in the lower troposphere. Significantly easterly wind anomalies dominate from the western North Pacific passing through South China and into Southwest China, potentially transporting warm and wet air masses from the East China Sea towards the SS region. Additionally, upward motion anomalies emerge in the mid-lower troposphere between 102° E and 108° E. These water vapor transport and vertical motion anomalies, along with the higher pseudo-equivalent temperatures in the lower troposphere, facilitate the appearance of large-area and significantly positive rainfall anomalies that appear in the SS region.

This study reveals the relationship between the daily APO index, atmospheric circulation patterns, and rainfall anomalies in the SS region, providing an atmospheric signal for the EHA and ELA days and their associated rainfall anomalies. These findings offer valuable insights that could enhance the prediction of the extreme APO and related rainfall occurrences in the SS region on sub-seasonal timescales. Notably, on sub-seasonal timescales, atmospheric circulation anomalies are influenced by both intrinsic dynamic processes within the atmosphere and external forcing [28,44,45,46,47,48,49,50]. These atmospheric circulation anomalies may in turn affect interannual shifts in rainfall patterns over the SS region. Investigating the potential physical mechanisms through which these factors influence interannual variations in EREs over the SS region, particularly by adjusting atmospheric circulation anomalies at lower latitudes, warrants further exploration in future research. Furthermore, future studies will explore other significant factors affecting rainfall variability in the SS region, such as changes in land use and regional topography. This research will contribute to the development of predictive models and mitigation strategies for extreme weather events in the region.