Summer Onset in Northern East Asia: Feature, Mechanism and Variability

Abstract

Summer in the East Asian monsoon region is characterized by heavy rainfall and high temperature. Its onset, depicted by monsoon rainfall and/or airflow as well as surface air temperature, has been well documented. However, the onset of summer season is rarely addressed in northern East Asia (NEA) around the northern edge of the East Asian summer monsoon. This study investigates the feature, mechanism, and variability of the summer onset in NEA based on the ERA-5 reanalysis dataset for 1979–2020. Results show that, in climatology, the onset of summer in NEA occurs in pentad 31 when the spring-to-summer warming process is decelerated at the highest rate. The change in the warming rate is mainly attributed to a decrease in the diabatic heat, mostly surface sensible heat, and temperature advection plays a small role. After the onset of summer, regional low-level northwesterly winds are weakened, and a local NEA cyclonic low forms. The latter, coupled with monsoon southerly airflow to the south, advects more moisture into NEA and increases regional rainfall. Furthermore, a temperature threshold of 17 °C, the climatological regional mean surface air temperature in pentad 31, was proposed to depict summer onset in NEA. Based on the temperature threshold, the year-to-year variability of summer onset timing in NEA is revealed, ranging from pentad 29 (late May) to 34 (middle June), with the standard deviation of 1.2 pentads. It advanced by 0.6 pentads, on average, after the late 1990s. This study provides a new method to objectively quantify the timing of summer onset in East Asia, which is thermodynamically explainable and may help us to depict and monitor summer onset in different latitudes and topography.

1. Introduction

Annual cycle is the dominant variability in the monthly and daily time scales for many climate variables in regions located far away from the tropics [1,2]. Seasonal temperatures increase at a more or less rapid rate in spring to a variable summer peak. The spring-to-summer transition corresponds to remarkable changes in temperature and rainfall, which can lead to consequent changes in regional ecosystem functions, agricultural practices, and environmental–social characteristics [3,4,5,6]. In addition, the summer season timing and duration become more important, accompanied by the recent increases in heat-wave events that caused large significant damage worldwide [7,8,9,10]. Thus, it is vital to identify the timing of summer onset.

Summer is characterized by heavy rainfall in East Asia. Many studies have investigated the sub-seasonal evolution in summer rainfall and monsoon in East Asia and obtained similar features in advance of summer rainy season onsets, with different variables (rainfall and/or lower-tropospheric monsoon air flow) and datasets, e.g., [11,12,13,14,15,16,17,18,19,20,21,22,23,24]. Specifically, in climatology, the summer rainy season begins in South China in mid-to-late May, then advances northward to the Yangtze River valley and Japan in mid-June, reaching North China in early July and Northeast China in mid-late July. The northward advances in the rainy belt are accompanied by the northward march of monsoon southerly airflow. Northern East Asia (NEA) is located around the northern edge of the East Asian summer monsoon region. The annual rainfall amounts to less than 500 mm in most regions and is even lower (<250 mm) in the west (semi-arid/arid region), which is controlled mainly by dry westerly air flow and partially by wet East Asian summer monsoon in the east [25]. Therefore, metrics other than rainfall and monsoon air flow should be proposed to depict summer onset in NEA, especially outside monsoon regions.

In addition to the rainy season, East Asian summers are also characterized by a high temperature. Temperature is important to the structure and function of biological systems because of its pervasive effects on behavior, physiology, distribution, and migration [3]. The seasonal evolution of temperature is closely tied to agricultural production and phonological phenomena in China [26,27]. In 104 BC, based on the periodic change that occurs as the Earth orbits the sun, the Twenty-Four Solar Terms were laid down as a calendar in Tai Chu Li. After the summer begins (the first solar term in summer), for example, crops grow quickly after being planted in spring. Though the Twenty-Four Solar Terms provide excellent guidance for agricultural activities [28], the unified criterion across China is inappropriate outside the Yellow River basin due to significant differences in climate conditions across different regions. In combination with phonological phenomenon and agricultural production, Chang [29] proposed using two fixed thresholds (10 °C and 22 °C) of pentad surface air temperature to depict the timing of the four-season onset in China. Summer onset timing is defined as the pentad when the surface air temperature first exceeds 22 °C [30]. Based on this criterion, Miao and Wang [31] pointed out that, in climatology, summer begins in South China in April, in the Yangtze and Huanghe River valleys in May, and in northern North China and southern Northeast China in June. However, there is no summer season in northern Northeast China and the Tibetan Plateau [32], where crops such as rice and highland barley grow. The absence of summer suggests that the threshold of 22 °C is probably unsuitable to detect the onset of summer in midlatitudes and high mountains. Considering the topography effect, Tang et al. [33] suggested a temperature threshold of 17 °C, instead of 22 °C, to represent summer onset over the Tibetan Plateau. In addition, Qian et al. [34] applied the ensemble empirical mode decomposition (EEMD) method to isolate the annual cycle from surface air temperature at daily time scales. Four seasons were identified, with two fixed temperature thresholds for the isolated annual cycle components. However, the two thresholds are somewhat subjective and cannot objectively depict the onset of summer based on temperature in East Asia, especially in mid-latitude.

Trenberth [35] referred to summer as the warmest quarter of the year and quantified the summer-onset temperature threshold as the 75th percentile of multi-year daily temperature values. Similarly, Yan et al. [36] described the summer as the hottest 90 days in a long-term period. This definition may well depict summer season in mid and high latitudes [37,38,39] but not in monsoon regions where heavy rainfall leads to a decrease in summer temperature. Considering the spatial heterogeneity across Europe and the long-term trend in temperature, Peña-Ortiz et al. [40] defined the onset of summer in Europe using the threshold of the average climatological temperature between June 1 and 15. Though this metric is applicable for Europe, the period for average temperatures may not be suitable for NEA. As revealed in this study, summer onsets in NEA in the first pentad (June 1 to 5) in NEA in climatology. In addition, the mechanism and effect of summer onset in NEA remain unclear.

All the definitions of summer mentioned above that used surface air temperature are statistical and empirical in nature. In fact, the temperature change follows the thermodynamic energy equation. Considering the onset of summer as the transition from a rapid warming process in spring to a slow warming process in summer, Chyi et al. [41] recently proposed a new way to objectively define the onset of summer in high-latitude northern Asia based on changes in surface air temperature, which is thermodynamically explainable. They proposed climatological summer onset as the pentad with the strongest deceleration during the warming period from early April to late June. Based on this criterion, they identified that the onset of summer in northern Asia occurs in the second pentad of June. Consequently, they chose a fixed threshold of 7.8 °C, the climate mean pentad temperature averaged over northern Asia in the second pentad of June, to describe the onset of summer in northern Asia for each year. Moreover, they discussed the mechanism for the summer onset in northern Asia and highlighted the importance of the positive feedback between surface warming and snowmelt. However, their analysis was confined to the high-latitude northern Asian region. It is unclear whether this method can be extended to East Asia in the low and middle latitudes.

In this study, we attempt to identify the timing of summer onset in NEA, using the objective definition method based on the change in surface air temperature proposed by Chyi et al. [41]. Moreover, we further explore the mechanism for the change in the warming rate related to summer onset. We chose NEA as the first step because, in East Asia, it is closest to northern Asia that was the focus of Chyi et al. [41]. The data and method used in this study are introduced in Section 2. In Section 3.1, we show that the climatological temperature warming decelerates in the first pentad of June (P31), one pentad earlier than that in northern Asia identified by Chyi et al. [41]. Then, we further reveal the underlying thermodynamic processes responsible for summer onset in NEA based on climatological heat budget analysis in Section 3.2. Changes in associated atmospheric circulations and rainfall are examined in Section 3.3. Finally, based on the temperature threshold, the climatological regional-mean surface air temperature in NEA in P31, we identify the timing of the onset of summer for each year from 1979 to 2020 in Section 3.4. Discussion and conclusions are presented in Section 4 and Section 5, respectively.

2. Data and Methods

2.1. Study Area

This study focused on NEA. The NEA region (40°–50° N, 105°–135° E, depicted by the box with red dashed lines in Figure 1) includes the eastern Mongolian Plateau and Northeast China, as well as part of Southeast Russia. The Mongolian Plateau, which includes the Gobi and areas of dry short-grass steppe, ranges in elevation from 1000 to 1500 m above sea level. The Northeast China Plains and Amur River Basin are surrounded by Great Khingan Mountains in the west and Changbai Mountains and Sikhote-Alin Mountains in the east, and they are divided by the Lesser Khingan Moutains. The average annual mean temperature for NEA is 4 °C.

2.2. Data

We used daily mean data from the European Center for Medium-Range Weather Forecast (ECMWF) reanalysis data version 5 (ERA-5) with a horizontal resolution of 1° × 1° during the 1979–2020 period [42]. The variables included air temperature at 2 m, total precipitation, surface sensible heat flux, and sea level pressure, as well as geopotential height, air temperature, horizontal winds, and vertical pressure velocity at the 37 pressure levels. Climatology was defined as the 42-year mean for 1979–2020.

To define the onset of summer in NEA, we adopted the pentad data, using the same strategy as Chyi et al. [41] who defined the onset of summer in northern Asia. First, we averaged daily surface air temperature in land grids of NEA (40°–50° N, 105°–135° E). Then, we obtained seasonal-cycle signals by removing high-frequency variability using a 30-day low-pass filter. Finally, the pentad data (Tp) were constructed for the entire study period of 1979–2020 based on the following protocol: A calendar year consists of 72 pentads, and every month has six pentads. One pentad consists of five non-overlapping days, except for the sixth pentads of February, March, May, July, August, October, and December. The sixth pentad consisted of three (day 26 to day 28) or four days (day 26 to 29 in a leap year) for February, while it was averaged as the six-day mean (day 26 to day 31) for March, May, July, August, October, and December.

2.3. Definition of Summer Onset in NEA

We adopted the definition of the onset of summer proposed by Chyi et al. [41]. First, we calculated the pentad-to-pentad increment of Tp (dTp, the first derivation of Tp with respect to time) and then the pentad-to-pentad increment of dTp (dT2p, the second derivation of Tp with respect to time). During the spring-to-summer transition period, a positive dT2p value depicts acceleration of the warming process, and a negative value depicts deceleration of the warming process.

Following Chyi et al. [41], we defined the onset of summer in NEA in climatology as the pentad with the minima in climatological dT2p during the spring-to-summer warming period from mid-April to late June. The summer onset depicts the transition from a rapid surface warming process in spring to a more stable warming process in summer. Consequently, the climatological regional-mean Tp for NEA in the summer-onset pentad was then adopted as a temperature threshold for summer onset in NEA. For each year, a pentad was defined as the summer onset timing when the pentad-mean Tp was closest to the temperature threshold.

2.4. Climatological Heat Budget Analysis

To clarify the major thermodynamic processes contributing to summer onset in NEA in climatology, we analyzed the heat budget at the 850 hPa surface, the pressure level closest to the surface over NEA. The climatological temperature tendency was calculated as

$$\frac{\partial \overline{T}}{\partial t}=-\overline{\overrightarrow{V}·\nabla T}+\overline{\mathsf{\omega }\left(\frac{RT}{p{C}_p}-\frac{\partial T}{\partial t}\right)}+\overline{Q}$$

(1)

where the overbar denotes climatology for 1979–2020. $T$ is temperature, $\overrightarrow{V}$ and$\mathsf{\omega }$depict the horizontal winds vector and vertical pressure velocity, and $Q$ represents diabatic heat.$R$ is the gas constant, $p$ is pressure,$C_p$ is specific heat at a constant pressure and $\nabla$ the horizontal gradient. The term on the left-hand side of Equation (1) represents the temperature tendency. On the right-hand side, the first and second terms depict horizontal temperature advection and temperature changes due to vertical motion. The third term is assessed as a residual term of Equation (1), the approximate diabatic heat term.

All the terms were calculated first using daily mean data and then averaged for pentad data according to the strategy outlined in Section 2.1. Finally, the climatological pentad data were derived for the 42-year mean from 1979–2020.

The flowchart for the methodology is presented in Figure 2. In this study, we first processed daily surface air temperature data to obtain pentad data and averaged pentad surface air temperature (Tp) for NEA (40°–50°N, 105°–135°E). Then, we identified climatological summer onset timing in the i^th pentad (Pi) in NEA based on the minimum value in dT2p. We further analyzed related circulation and rainfall change to reveal the effect of the summer onset in NEA and investigated the summer-onset mechanism by diagnosing climatological heat budget. Moreover, we proposed the climatological pentad temperature in Pi as the temperature threshold for summer onset in NEA. Based on this threshold, we revealed summer-onset variability in NEA year by year from 1979 to 2020.

3. Results

3.1. Summer Onset in NEA

Figure 3 shows the climatological surface air temperature in May and June. In May, surface air warms more quickly in continental NEA than in the adjacent oceans in the east at 40°–50°N (Figure 3a). The average temperature in the NEA region (40°–50°N, 105°–135°E) is 13.4 °C. Surface air continues warming in June (Figure 3b), with a regional mean of 18.8 °C, which is five degrees higher than that in May. The change is remarkably different over NEA in the meridional direction (Figure 3c). The increase in temperature is higher to the north (>6 degrees) than the south (~4 degrees), which is likely due to the positive feedback between surface warming and snowmelt to the north [41]. On the other hand, the zonal difference is small, with slightly stronger warming in the west (5.5 °C) over the eastern Mongolian Plateau (40°–50° N, 105°–120° E) than east (5.2 °C) in the Northeast China plain and surrounding mountains (40°–50° N, 120°–135° E).

The spring-to-summer warming process in climatology is clearly illustrated in Figure 4a through the evolution of regional-mean pentad surface air temperature in NEA from P21 (the third pentad in April) to P36 (the last pentad in June). The average warming rate is approximately 1.2 °C per pentad before pentad 31 (P31, the first pentad in June). It drops one third later, with a warming rate being less than 0.8 °C per pentad after P31. We tested the significance of the change using a running t test. The results show that the surface air warming rate indeed experiences the most significant change around P31 (Figure 4b). Accordingly, the warming process is decelerated, with the dT2p minima in P31, suggesting that surface air in NEA transits from a fast-warming phase to a relatively slow-warming phase. Following Chyi et al. [41], we therefore defined P31 as the summer onset pentad in climatology over NEA. Consequently, the climatological pentad Tp (17 °C) in P31 is referred to as the temperature threshold for summer onset in NEA.

We further showed the pentad-to-pentad evolution of spatial distribution for Tp and the warming tendency (dTp) from P27 to P32 (Figure 5). Surface air warms first in the south NEA, with two maxima in the Northeast China Plains and to the south of the Mongolian Plateau in P27 (the third pentad in May) (Figure 5a). The warm surface air advances northward, with an almost uniform warming rate exceeding 1 °C per pentad over the whole NEA from P27 to P30 (Figure 5a–d). After that, the warming rate drops to less than 1 °C per pentad in most of NEA in P31 (Figure 5e) except in the northwest and P32 (Figure 5f). The results also indicate that surface warming transits from a fast to slow warming stage after P31.

Given the difference in the warming rate between west and east NEA as shown in P31 (Figure 5e), we further calculated the regional-mean dTp and dT2p for east and west NEA, respectively. Surface air experiences the most significant change and reaches the minimum acceleration around P31 in both east and west NEA (Figure 4), the same as that for the whole of NEA, though differences exist in the magnitudes. Meanwhile, the temperature thresholds, i.e., the climatological Tp at P31 in the two sub-regions (17.1 °C for east NEA and 17.2 °C for west NEA), are also close to the whole NEA region (17 °C). Considering the similarity in temperature change and thresholds in the west and east portions, hereafter in this study we focused on the whole NEA region (40°–50° N, 105°–135° E).

3.2. Mechanism

To understand the mechanism for the change in temperature tendency in NEA around summer onset, we analyzed the evolution of climatological pentad heat budget terms in Equation (1) at 850 hPa averaged over the NEA region. As shown in Figure 6a, the warming rate is apparently attributed to the positive diabatic heat forcing, which is largely offset by the cooling effect of advection. Here, the advection consists of both horizontal temperature advection and temperature change due to vertical motion, i.e., the combination of the first two terms in the right-hand side of Equation (1). Moreover, around the onset of summer (P31), the decrease in the warming rate is synchronized with the weakening of diabatic heat forcing. Compared with P30 before the summer onset, the temperature tendency in P32 reduces by 0.35 °C per pentad. Two thirds of the reduction are contributed by the change in diabatic heat (–0.22 °C per pentad) and only one third by the change in the advection (–0.12 °C per pentad). The result suggests that the abrupt change in the warming rate around the summer onset is mainly caused by the reduced diabatic heat over NEA. Moreover, the evolution of diabatic heat in NEA is coincidental with the regional surface sensible heat flux (Figure 6b); both demonstrate a peak in pentad 30. The slow increase in surface sensible heat flux before P30 is due to the effect of increasing solar insolation (figure not shown). After P30, the shortwave radiation absorbed at the surface decreases as rainfall increases, which is discussed later. Consequently, sensible heat flux and diabatic heat at the surface decrease.

We further showed the spatial distribution of the three heat budget components at 850 hPa in NEA before and after the summer onset. Air warms before the summer onset, with a warming rate of 1.0~1.2 °C per pentad in pentads 29–30 (Figure 7a). This warming is mainly caused by positive diabatic heat (Figure 7e) and is offset by cold advection (Figure 7c). The diabatic heat shows a similar spatial pattern to surface sensible heat flux (Figure 7g). More sensible heat flux in the west NEA is consistent with the stronger diabatic heat in the west than east NEA. The stronger diabatic heat in the west is offset by stronger cold advection due to strong northwestern winds (Figure 8a). In east NEA, weak warm advection is likely due to southwesterly winds in the southeast that advects warmer air from land in the west to sea in the east (Figure 5). After the summer onset, the warming rate weakens (Figure 7b). The maximum warming rate is less than 1 °C per pentad in pentads 31–32, which is consistent with the result of average temperature tendency for NEA (Figure 6a). For diabatic heat (Figure 7f) and cold advection (Figure 7d), the spatial patterns resemble those prior to the summer onset (Figure 7c, e). Meanwhile, the spatial pattern of diabatic heat is also similar to the surface sensible heat flux (Figure 7h). The similarity between the diabatic heat and surface sensible heat flux in NEA in the sub-seasonal evolution of regional mean (Figure 6) and spatial pattern (Figure 7) indicates that diabatic heat in NEA is probably contributed to by surface sensible heat flux. In other words, the change in the warming rate in NEA is mainly caused by the change in regional surface sensible heat flux.

3.3. Related Circulation and Rainfall Changes

The summer onset in NEA in climatology is related to remarkable change in low-level circulation. Before the summer onset, northwesterly winds dominate over NEA in pentads 29–30 (Figure 8a) and divert eastward, forming a strong East Asian trough in the east (Figure 8a,b). Meanwhile, a strong westerly jet core in the upper troposphere appears over the western North Pacific south of Japan (Figure 8b). These lower-, middle-, and upper-tropospheric circulation features are closer to the winter circulation pattern [43,44,45], though an isolated low is evident at the sea level pressure over NEA (Figure 8a) due to surface warming (Figure 5).

After the summer onset, the northwesterly winds are greatly weakened, with a closed cyclonic circulation over NEA (NEAL) at 850 hPa (Figure 8c). The sea level pressure over NEA expands southward and westward (Figure 8c), merging with a continental-scale subtropical Asian low. In addition, the East Asian trough becomes shallower (Figure 8d). The northern edge, depicted by the zero contour in eddy components of geopotential height at 500 hPa, shrinks to the south of 50°N and is detached from the mid-high trough in northern Asia by a weak ridge. In the upper troposphere, the westerly jet weakens in the core region (Figure 8d). The changes in the circulation are characterized by a see-saw pattern in sea level pressure (Figure 8e) and geopotential height at 500 hPa (Figure 8f), with a positive anomaly to the northeast of NEA and a negative anomaly to the southwest. The positive anomaly weakens the East Asian trough (Figure 8b vs. Figure 8d), and the negative anomaly enhances the Asian low (Figure 8a vs. Figure 8c). The formation of NEAL in the lower troposphere, shallower East Asian trough, and weakened westerly jet all indicate the transition to summer circulation [25,46,47,48].

As the sea level pressure low in NEA merges with the subtropical Asian low (Figure 8c), the southwesterly winds in the front of the NEAL are coupled with the monsoon southerly winds to the south. The coupling brings more moisture to NEA, which is consistent with the anomalous northwestward transport of moisture by the southeasterly wind anomalies (Figure 8e). Consequently, rainfall increases in NEA after the summer onset (Figure 9).

3.4. Year-to-Year Variability

Based on the temperature threshold (17 °C) for summer onset in NEA, which was proposed in Section 3.1, we revealed the summer onset pentad in NEA for each year from 1979 to 2020 (Figure 10). The onset pentads range from P29 (late May) to P34 (middle June), with a standard deviation of 1.2 pentads. In addition, a weak downward trend was identified during 1979–2020 due to a change in the onset pentads around the late 1990s. For the first 21 years (1979–1999), there were nine (five) years in which summer onset in NEA is later (earlier) than P31 (the climatology). In contrast, there were six (nine) years of late (early) summer onset in NEA for the next 21 years (2000–2020). The 21-year mean was 31.3 for 1979–1999 and 30.7 for 2000–2020, that is, summer began 3 days (0.6 pentads) earlier after 2000 in NEA.

4. Discussion

This study provides a new way to objectively quantify the timing of summer onset in East Asia in view of temperature change. We proposed 17 °C as the surface air temperature threshold for summer onset in NEA. The threshold is lower than the threshold of 22 °C for summer onset in China proposed by Chang [29]. The lower threshold in NEA is consistent with the lower climatological summer temperature due to the weaker absorption of solar radiation at the surface in higher latitudes (40°–50° N). On the other hand, based on the fixed temperature threshold 22 °C proposed by Chang [29], Miao and Wang [31] showed that no summer season exists in northern Northeast China, which is consistent with the result that the climatological maximum pentad temperature (21.8 °C) in NEA is less than 22 °C. Meanwhile, the threshold 17 °C in NEA is the same as that modified by Tang et al. [33] for summer onset in the Tibetan Plateau, where surface air temperature decreases as the altitude increases.

The 17 °C threshold is closely tied to agricultural production and phonological phenomenon in NEA. It is close to the lower threshold (18 °C) for the fast summer growth of spring corn in the last vegetative stage and the lower threshold (15~17 °C) for rice tillering in Northeast China. Therefore, it can be used as a meteorological indicator for regional agricultural production and phonological phenomenon in summer. In other words, the 17 °C threshold is appropriate to represent summer onset in NEA.

In this study, we showed that summer onsets in P31 in NEA, one pentad earlier than in mid-high-latitude northern Asia (P32). The spring-to-summer warming process decelerates earlier in NEA than northern Asia. However, it should be noted that the temperature thresholds used are different in NEA (17 °C) and northern Asia (7.8 °C). The lower temperature threshold in northern Asia is likely due to weaker insolation and greater reflection by snow in higher latitudes [41].

This study showed that, in climatology, the regional surface air warming rate in NEA is 1.2 °C per pentad before summer onset. The warming rate in NEA is close to that in northern Asia (Chyi et al., 2021); both are contributed to by positive diabatic heat release and are largely offset by cold temperature advection. The diabatic heat rate is weaker in high-latitude northern Asia (2.5 °C per pentad) than mid-latitude NEA (5 °C per pentad), due to great reflection and less absorption of solar shortwave radiation by snow cover in high latitudes. However, because of the larger meridional gradient of temperature in middle latitudes compared to high latitudes, the effect of cold advection is also stronger in NEA than northern Asia. As the result of the combined effect of diabatic heat and temperature advection, the warming rates in NEA and northern Asia are close.

The warming rate in NEA decreases significantly from 1.2 °C per pentad before the summer onset to 0.8 °C per pentad after. The change in the warming rate is mainly attributed to the decrease in diabatic heat, similar to that found in northern Asia (Chyi et al., 2021) but related to different physical processes. In northern Asia, the decrease in diabatic heat agrees with the weakening in regional snowmelt process, which reduces the change in surface sensible and latent heat release and then decelerates the warming process in northern Asia. Alternatively, due to the absence of snow in NEA in late spring (see Figure 7 in Chyi et al., 2021), local dry surface air is warmed mainly by surface sensible heat. After P30, the surface sensible heat flux is suppressed, which is likely related to less solar radiation obtained and absorbed at the surface as rainfall increases (Figure 9). Consequently, the decrease in surface sensible heat slows down the warming process in NEA and summer onsets.

Summer began approximately 0.6 pentads (3 days) earlier in NEA after the late 1990s (Figure 10). The interdecadal change is consistent with the summer warming in NEA after the mid-late 1990s, which was attributed to the local upper-tropospheric anticyclone and positive geopotential height anomaly [49,50]. The upper-tropospheric circulation anomaly was probably modulated by the phase changes in the Atlantic Multi-decadal Oscillation (AMO) [25,51,52,53,54] and the Pacific Decadal Oscillation (PDO) [55], and the increased concentration in greenhouse gases and changes in anthropogenic aerosol emissions [56,57].

This study had several limitations. Firstly, the seasonal variation of surface air temperature is closely related to large-scale atmospheric dynamical processes and surface thermodynamic processes, including shortwave and longwave radiation as well as the turbulent heat exchange between the surface and atmospheric boundary level. In this study, we compared their relative importance to determine the overall effects of dynamic and thermodynamic processes. However, the lack of diabatic heating rate data makes it impossible to quantify the individual impacts of these diabatic heat components, and thus further observations and model simulations are required. Secondly, we demonstrated that the timing of summer onset in NEA varies year by year (Figure 10). Several questions related are posed in need of further investigation: Does the mechanism for climatological summer onset in NEA also work for the year-to-year variability? Is there any difference in the summer onset between west and east NEA on the interannual timescales?

5. Conclusions

We extended the definition of summer onset proposed by Chyi et al. (2021) for high-latitude northern Asia based on temperature change, to the subtropical and middle-latitude East Asia. As the first step, this study quantified summer onset in NEA. Results show that, in climatology, surface air in NEA experiences the transition from a rapid warming stage to a slow warming stage around P31, when the strongest deceleration occurs. The regional mean warming rate drops one third, from 1.2 °C per pentad before P31 to 0.8 °C per pentad after. Therefore, we defined the P31 as the summer onset pentad in NEA in climatology and proposed the climatological regional surface air temperature 17 °C in P31 as the threshold for summer onset in NEA. Based on the climatological heat budget analysis, we concluded that the change in warming rate is mainly attributed to the decrease in diabatic heat due to surface sensible heat flux and partly by temperature advection. Related to the summer onset, the circulation patterns change remarkably. The northwesterly winds in winter and spring are reduced in the lower troposphere, and a closed cyclonic NEAL is formed and coupled with the continental subtropical Asian low to the south. In the middle troposphere, the East Asian trough becomes shallower, and the westerly jet in the upper troposphere weakens. Consequently, rainfall increases in NEA as more moisture is transported into NEA from the south due to the coupling of the NEAL and the southerly monsoon. Furthermore, we identified the summer onset pentad in NEA for each year from 1979 to 2020 based on the temperature threshold 17 °C. The summer onset timing ranged from P29 to P34. In addition, it advanced approximately 0.6 pentads (3 days) after the late 1990s, on average.

In this study, we provide a new method to objectively quantify the timing of summer onset in NEA. The summer onset is characterized by abrupt changes in the warming rate, from a rapid warming process in spring to a slow warming process in summer. The defined summer onset is thermodynamically explainable. It is caused mainly by a decrease in diabatic heating, especially in terms of sensible heat; temperature advection also plays a small role. Therefore, the summer-onset definition in this study was based on temperature change governed by the thermodynamic energy equation, which is different from the previous statistical and empirical definition for summer onset [2,28,29,31,32,35,38]. Moreover, because of the different heat budget contributions in different regions, the threshold for summer onset is variable and highly regional. Thus, the definition posed in this study can serve as a potential metric to objectively define, in a thermodynamically explainable manner, and monitor temperature-based summer onset in different latitudes and topographic heights of East Asia. The application of this metric to depict summer onset in East Asia is ongoing and may further improve our understanding of the seasonal variations of East Asian climate and circulation.