Influence of the Northwest Pacific tripole mode on the mid-summer precipitation in North China and the regulation by the North Atlantic

The North China mid-summer (July) precipitation (NCJP) contributes the largest proportion of total annual precipitation in North China, with significant interdecadal and interannual variability. The interannual variability of the NCJP was further investigated on the basis of a study of its interdecadal variability and found that a sea surface temperature (SST) pattern in July located in the northwest Pacific, defined here as the northwest Pacific SST tripole (NWPT), can significantly influence the interannual variability of the NCJP, and that this relationship is regulated by the decadal northern North Atlantic SST (NNASST). Diagnostic analysis and the linear baroclinic model experiment indicate that the positive (negative) NWPT in July can excite an anomalous anticyclone (cyclone) in the region centered on the Korean peninsula and an anomalous cyclone (anticyclone) in the northwest Pacific off southeast Japan, thereby strengthening (weakening) the NCJP. When the decadal NNASST is in a significantly positive phase, the positive geopotential height anomalies it excites in the northwestern region off North China are not favorable for the connection between the NWPT and the NCJP. When the decadal NNASST is in a negative or insignificantly positive phase, the July NWPT and the NCJP have a significant positive correlation on interannual timescale.


Introduction
North China (35 • -44 • N, 110 • -120 • E) is a densely populated area and one of the major political, economic and agricultural centers of China . Water resources in North China are relatively scarce, precipitation is an important source and the rainy season is very concentrated compared to other regions in China (Chen et al 2013, 2020, Ge and Huang 2001, Pei et al 2015, Chen and Sun 2019, Li and Lei 2022. Precipitation in North China is mainly concentrated in summer, so many scholars have conducted numerous studies on summer precipitation in North China (i.e. Wu et al 2003, Liu et al 2011, Wei et al 2014, Song and Wei 2021, Du et al 2022. Nevertheless, fewer studies have focused on mid-summer (July) precipitation in North China alone. The North China July precipitation (NCJP) accounts for the largest proportion of total summer precipitation and also the largest proportion of total annual precipitation (Xie et al 2023). Meanwhile, the precipitation in August accounts for the second highest proportion of total annual precipitation in North China after July, and the precipitation in North China in July and August has different variability characteristics (Xie et al 2023). Therefore, it is necessary and scientifically logical to consider the July precipitation in North China as a separate study rather than together in summer, and can provide very important theoretical basis for improving the operational forecasting capabilities of the NCJP.
There are relatively few studies on the July precipitation in North China, apart from individual case studies (Jiang et al 2015. Li et al (2021) studied the co-variability between the NCJP and the July precipitation in Kazakhstan-Xinjiang region and it is influenced by the atmospheric signals over the Kara and Mediterranean Sea and the Tibetan Plateau, but did not analyze the mechanism of change in the NCJP itself. Yang et al (2017) analyzed the modulation of monthly precipitation patterns over East China by the Pacific Decadal Oscillation (PDO) and found that the contribution of the PDO to the NCJP was not significant. Xie et al (2023) found that the NCJP shows a significant interdecadal variability of a cycle of about 20 years for at least from the last 40 years, and is related to the interdecadal variability of the northern North Atlantic sea surface temperature (SST) (NNASST). And revealed the physical mechanism of the NNASST influencing the interdecadal variability of the NCJP by modulating the mid-to-high latitude atmospheric circulation from the upstream weather systems for North China.
The climate system in North China is complex and can be influenced synergistically by the mid-tohigh latitude atmospheric circulation from upstream and the mid-to-low latitude subtropical climate system from downstream (i.e. Wu et al 2009, 2012, Wei et al 2014, Zheng et al 2014, Li and Ruan 2018, Zhang et al 2018, Xie et al 2019, Hu et al 2022. This means that the potential influences of the downstream mid-to-low latitude climate system on the NCJP cannot be ignored, especially the regional climate factors in the northwest Pacific such as the SST, the East Asian summer monsoon, the west Pacific subtropical high, etc, which have a significant influence on the climate system of North China. Meanwhile, it is also necessary to further explore the interannual variability of the NCJP on the basis of its interdecadal variability. In this paper, we explore the interannual variability of the NCJP on the basis of our previous studies on the mechanism of its interdecadal variability and found that the interannual variability of the NCJP is related to a July SST pattern over the northwest Pacific, defined here as the northwest Pacific SST tripole (NWPT), and that the connection between the July NWPT and the NCJP is regulated by the decadal NNASST. This paper is structured as follows: section 2 describes the data and methods used in this research. Section 3 describes the results, including the relationship of the NCJP with the July NWPT in section 3.1, the mechanism of the July NWPT influencing the interannual variability of the NCJP in section 3.2, and the modulating role of the decadal NNASST in the NWPT-NCJP linkage in section 3.3. Finally, section 4 gives the conclusions and discussion.

Data and methods
The Global Precipitation Climatology Project (GPCP) Version 2.3 Combined Precipitation Dataset for the period 1979-2020 was used, which has a horizontal resolution of 2.5 • × 2.5 • grid derived from the US National Oceanic and Atmospheric Administration (NOAA) (Adler et al 2003). The atmospheric data including geopotential height and winds were derived from the NCEP-DOE reanalysis 2 (NCEP2) dataset for the period 1979-2020 with a horizontal resolution of 2.5 • × 2.5 • grid (Kanamitsu et al 2002). The SST data were derived from the Hadley Centre Sea Ice and SST dataset for the period 1979-2020 with a horizontal resolution of 1 • × 1 • grid (Rayner et al 2003). Precipitation data from the daily meteorological dataset of basic meteorological elements of China National Surface Weather Station version 3, and SST data from the NOAA Extended Reconstructed SST version 5 (Huang et al 2017) were also used to reexamine our results.
The NCJP anomaly was calculated as the areaweighted mean July precipitation over the North China region (35 • -44 • N, 110 • -120 • E) relative to the base period 1979-2010 following Xie et al (2023). The NWPT index was defined as the normalized areaweighted mean SST anomalies difference amongst the three regions of the Northwest Pacific shown in figure 1(b), and is formulated as follows: where nSSTa A1 , nSSTa A2 and nSSTa A3 are the normalized SST anomalies for the regional average of The definition for the NNASST also follows Xie et al (2023) and was calculated as the area-weighted mean July SST over the northern North Atlantic There is significant interdecadal variability in the NCJP (Xie et al 2023). To better investigate the high frequency signal in the NCJP, the 7 year running mean was applied to obtain the low frequency interdecadal series, and then the high frequency series for the NCJP were obtained by removing the interdecadal variability. The linear baroclinic model (LBM; Watanabe and Kimoto 2000) was used to explore the influence of the NWPT on the NCJP. In this study, a dry version of the LBM and the July climatology with a T21 horizontal resolution and 20 vertical layers derived from the NCEP-NCAR reanalysis dataset were used. The LBM was integrated for 27 d, and the results after stabilization were used. Further details about the LBM can be found in Watanabe and Kimoto (2000).  Figure 1(a) shows the correlation map of the highpass filtered NCJP anomalies with the contemporaneous SST anomalies in the northwest Pacific during the period 1979-2020. It can be seen that the NCJP anomaly is significantly correlated with the July SST anomalies in the northwest Pacific, and the region of significant correlation is characterized by a tripole pattern distribution: a significant positive correlation region located east of the Philippines, a significant negative correlation region located between the Western Mariana Trench and the solar boundary, and a significant positive correlation region located in the waters adjacent to eastern China. This suggests that a combined consideration of the SST anomalies signals from the above three highly correlated regions may be able to provide an effective predictability source for the NCJP anomalies, because the dynamic climate models, which are important tools for the current short-term climate prediction service, are better at predicting SST than precipitation. Thus, an index that combines the SST anomalies of the above three regions (A1 region: 5 • S-15 • N, 132 • -158 • E, A2 region: 20 • -30 • N, 145 • -175 • E, and A3 region: 25 • -35 • N, 122 • -138 • E) is defined and named the NWPT. As shown in the correlation map between the NWPT index and the SST anomalies in the northwest Pacific ( figure 1(b)), the tripole feature of the NWPT is more pronounced, indicating that the NWPT index is defined in a feasible way.

Relationship of the NCJP with the NWPT
Then, the high-pass filtered NCJP anomalous sequences and the July NWPT indices are shown in figure 1(c). It can be seen that the NCJP anomalies and the NWPT indices show relatively consistent interannual variations for the periods 1979-2002 and 2012-2020, while between 2003 and 2011 they do not show a pronounced relationship, suggesting that the relationship between the NCJP and NWPT may not be stable during the period 1979-2020. As shown in figure 1(d), the 7 year sliding correlation between the NCJP anomalies and the July NWPT indices shows that there is interdecadal variation in the relationship between the NCJP and the July NWPT. During the period 1979-2002, the NCJP anomalies and the July NWPT indices show significant positive correlations, with correlation coefficients even exceeding 0.8, during the period 2003-2012, there are no significant correlations, and during the period 2013-2020, there are again to be significant positive correlations. This is because the relationship between the NCJP and NWPT is regulated by the decadal NNASST, which will be described in a later section. In addition, it should be noted that the low-pass filtered series of the NWPT does not show an interdecadal variability similar to that of the NCJP, but rather an upward trend from 1980 to 1987, followed by a downward trend until 2020.

Mechanism of the NWPT influencing the NCJP
From the composites of the July geopotential height anomalies and the horizontal wind anomalies at 500 hPa for high NCJP years, there are positive geopotential height anomalies in the region east of North China to the Kuril Islands, and negative geopotential height anomalies in the region west of North China to the western Mongolian Plateau and the northwest Pacific of southeastern Japan, and this circulation pattern is clearly favorable to precipitation in North China, which can also be seen in the horizontal wind field ( figure 2(a)). As shown in figure 2(b), the circulation pattern is opposite to that in figure 2(a) when the NCJP is a significant negative anomaly, which is not favorable to precipitation in North China. As shown in figure 2(c), from the composite difference of the July geopotential height anomalies and the horizontal wind anomalies at 500 hPa between the years of significantly high and low NCJP, the circulation pattern favoring increased NCJP is more evident. North China is located to the west of the anomalous anticyclone, which can bring large amounts of warm and humid air from the sea to North China, and the anomalous cyclone of southeastern Japan can further enhance this moisture transport, enabling a significant increase in NCJP when this warm and humid air meets the cold air brought by the anomalous cyclone in northwest China.
Regressing the July geopotential height anomalies and horizontal wind anomalies at 500 hPa on the NWPT index after normalization shows that the regression pattern is similar to the circulation pattern as in figures 1(a) and (b), which favors precipitation in North China. This suggests that the July NWPT may be able to excite an anomalous anticyclone in the region centered on the Korean peninsula and an anomalous cyclone in the northwest Pacific off southeast Japan, thereby strengthening the NCJP, which was also confirmed in an LBM experiment. As shown in figures 3(a), a circulation pattern similar to that in figure 2(d) is obtained when an SST forcing source such as the NWPT pattern is set up in the northwest Pacific. It should be noted that the position of the positive geopotential height anomalies obtained from the LBM experimental forcing is slightly shifted from that in figure 2(d), perhaps due to that the LBM only considers the linear processes involved and there may also be non-linear effects. In addition, the circulation patterns obtained when the heat or cold forcing source is set up in A1 region, A2 region, and A3 region are shown in figures 3(b)-(d), respectively, indicating that the cold source in the A2 region may play the largest role in the NWPT influencing the NCJP interannual variability, but cannot be separated from the synergistic effects coming from the A1 and A3 regions, because in the LBM experiment with a cold source in the A2 region alone, the other regions are equivalent to the heat source in relation to the A2 region.

Modulating role of the NNASST
The relationship between the NCJP and the July NWPT and the physical mechanisms by which the July NWPT influences the NCJP have been analyzed previously, and it has also been noted that the relationship between the NCJP and the July NWPT was not stable, especially in the period 2003-2011, and that there may be other climatic factors influencing the relationship between the NCJP and the July NWP T. Xie et al (2023) analyzed the interdecadal variability of the NCJP and found that July NNASST can influence the interdecadal variability of the NCJP by modulating the atmospheric circulation at mid-to-high latitude. Further analysis showed that the interannual variability of the NCJP with respect to the July NWPT was modulated by the July decadal NNASST, which is described in detail below.
As shown in figure 4(a), during the period 2003-2011, when the NCJP was not significantly related to the July NWPT, the decadal NNASST was in significant positive anomalies. When the decadal NNASST is in the positive phase, there are positive geopotential height anomalies in North China and its northwestern region, which are not conducive to the precipitation in North China (Xie et al 2023). Combined with the mechanistic analysis of the influence of the July NWPT on the interannual variability of the NCJP, the positive geopotential height anomalies in North China and its northwestern region due to the decadal NNASST are speculated to be an important cause of the insignificant relationship between the July NWPT and the NCJP on the interannual timescale in 2003-2011. Figure 4(b) shows the correlation coefficients between the July NWPT index and the NCJP anomalies after high-pass filtering in the period of significant positive phase and other of the decadal NNASST. It can be seen that there is a significant positive correlation between the NCJP and the July NWPT when the decadal NNASST is in a negative or insignificantly positive phase, with a correlation coefficient of 0.66 (significant at the 95% confidence level), while when the NNASST is in a significantly positive phase the relationship is no longer significant, with a correlation coefficient of −0.34. A consistent finding can also be obtained from the scatter plots as shown in figures 4(c) and (d), the distribution of the NCJP anomalies corresponding to the July NWPT index is disordered when the July decadal NNASST is in a significantly positive phase and becomes ordered when the July decadal NNASST is in a negative or insignificantly positive phase. These suggest that the relationship between the NCJP and the July NWPT on interannual timescale is modulated by the July decadal NNASST.
As shown in the composite plots of the 500 hPa geopotential height anomalies and the horizontal wind anomalies in July corresponding to the positive (figures 5(a) and (d)) and negative (figures 5(b) and (e)) anomalies of the NWPT index for the period when the decadal NNASST not in a significant positive phase (figures 5(a) and (b)) and in a significant positive phase (figures 5(d) and (e)), when the decadal NNASST is in a significant positive phase, there are positive geopotential height anomalies in the northwestern area of North China regardless of whether the NWPT is positive (figure 5(d)) or negative (figure 5(e)) anomalies, which is consistent with the study of Xie et al (2023), that positive NNASST anomalies can generate anomalous positive geopotential heights in the aforementioned area. In particular, when the NWPT is a positive anomaly, the positive geopotential height anomalies excited in the region centered on the Korean peninsula are linked to the positive geopotential height anomalies excited by the NNASST in the northwest China, resulting in a dumbbell-shaped positive geopotential height anomaly field in the region from Lake Baikal to Japan ( figure 5(d)). When the NWPT is a negative anomaly, the negative geopotential height anomalies excited in the region centered on the Korean peninsula squeeze the positive geopotential height anomalies excited by the NNASST in the northwest China to the northwest, resulting in a dumbbell-shaped negative geopotential height anomaly field in the region from North China to southern Japan (figure 5(e)). Both of the above circulation patterns (figures 5(d) and (e)) have no significant effect on the NCJP anomalies, as can also be seen from the composite difference (figures 5(c) and (f)). These results above verify that the relationship between July NWPT and NCJP at interannual timescale can be modulated by the decadal NNASST. On the basis of the mechanistic analysis of the decadal NNASST regulating the relationship between the July NWPT and the NCJP anomalies, the corresponding mechanism schematics are summarized as in figure 6. As shown in figures 6(a) and (b), when the decadal NNASST is in a negative phase, a geopotential height anomalies field with the structure 'positivenegative-positive-negative' at 500 hPa can be excited from the North Atlantic to East Asia. At this time, the positive NWPT anomaly that can excite the positive geopotential height anomalies in the region centered on the Korean peninsula, in conjunction with the NNASST-excited negative geopotential height anomalies located in the northwestern area of North China, can enhance the NCJP through enhanced water vapor transport (figure 6(a)). As shown in figure 6(b), the negative NWPT anomaly that can excite the negative geopotential height anomalies in the region centered on the Korean peninsula, and connect with the NNASST-excited negative geopotential height anomalies, North China area is controlled by the anomalous westerly flow, which suppressed the NCJP. As shown in figures 6(c) and (d), when the decadal NNASST is in a significant positive phase, the difference in positive or negative NWPT anomalies on the NCJP is not significant.

Conclusion and discussion
In this paper, we found an SST pattern in the northwest Pacific in July that can influence the NCJP interannual variability, defined as the NWPT, and analyzed the mechanism by which the July NWPT influences the NCJP interannual variability and the regulation role of the decadal NNASST. The main conclusions can be summarized as follows. The highpass filtered NCJP has a strong correlation with the July NWPT on interannual timescale, but the relationship between the NCJP and the July NWPT is not stable and is regulated by the decadal NNASST. The positive (negative) NWPT in July can excite an anomalous anticyclone (cyclone) in the region centered on the Korean peninsula and an anomalous cyclone (anticyclone) in the northwest Pacific off southeast Japan, thereby strengthening (weakening) the NCJP. When the decadal NNASST is in a significantly positive phase, the positive geopotential height anomalies it excites in the northwestern region of North China are not favorable for the connection between the July NWPT and the NCJP interannual variability. When the decadal NNASST is in a negative or insignificantly positive phase, the July NWPT and the NCJP have a significant positive correlation on interannual timescale.
Here, based on the study of the interdecadal variability of the NCJP by Xie et al (2023), we further explored the interannual variability of the NCJP and its ocean signal. In the future, further constructing an empirical model of the NCJP based on the NWPT and NNASST, and combining the output data from dynamic climate models can effectively improve the prediction capability of the NCJP.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).