Impact of Solar Activity and ENSO on the Early Summer Asian Monsoon During the Last Millennium

The Asian Monsoonal rainfall accounts for the majority of annual regional precipitation in East and South Asia and could be remotely regulated by El Niño‐Southern Oscillation (ENSO). Besides, several paleoclimate records and simulations have indicated solar signals in the Asian Monsoon, implying the impact of solar activity on the regional monsoon precipitation. By conducting multi‐linear regression analysis to the solar irradiance forced single‐forcing experiment in the last millennium, this study presents the comparison of solar and ENSO effects on monsoonal precipitation in South and East Asia during early summer (May–June). Increased total solar irradiance during high solar activity years tends to trigger a favorable environment for developing monsoon onset, leading to more precipitation against ENSO‐related patterns over Southeast and South Asia before peak‐summer (July–August). The result supports reconstructed terrestrial records and underlines considerable influences of the solar cycle on the variation of the Asian Summer Monsoon.

during high (low) activity years.Besides, Jin et al. (2019) provided explanations for the impacts of the decadal solar cycle on the East Asian Summer Monsoon and argued that the physical processes might result from a coupled decadal mode in the Asia-Pacific region.Notably, for improving our understanding of AM system, the investigation of early summer (May-June) and peak-summer (July-August) rainfall should be particularly separated because there are mechanistic differences in precipitation patterns between the two periods (Su et al., 2014;Wang et al., 2009) due to the complex geographical structure over the monsoonal Aisa.Furthermore, the two major subsystems of AM, the Indian and East Asian Monsoon, have distinct timing of monsoon onset during early summer (Yihui & Chan, 2005).In the aforementioned studies, the influences of solar activity on the early summer AM remain unclear.
In addition to the decadal cycle, solar activity is suggested to trigger multi-centennial variations of the Asian Summer Monsoon (Sun, Liu, Wang, Chen, Wan, & Wang, 2022).The changes in solar irradiance could modulate the sea surface temperature in the tropical Pacific, resulting in anomalous western North Pacific anti-cyclone and delayed precipitation anomalies over East Asia.The study highlighted the importance of exploring the potential effects of long-term the solar cycle and its potential influence on regional precipitation and hence water resources in the future.
In this study, we conducted a multi-linear regression analysis of the dataset from a model simulation of a solaronly forcing experiment that contained transient decadal and centennial solar cycles in the last millennium to explore the temporal and spatial complexity of AM.We attempt to investigate the precipitation responses to the solar activity over South and East Asia, with a focus on early summer, to offer plausible physical processes and provide a more comprehensive understanding of solar cycle effects on the monsoon system.

Data
The model simulation data set utilized in this study is from the Community Earth System Model-Last Millennium Ensemble (CESM-LME) modeling project (Otto-Bliesner et al., 2016) that prescribed CMIP5 (Coupled Model Intercomparison Project Phase 5) climate forcing reconstructions to conduct ensembles of last millennium simulations, containing "full-forcing" experiments with all forcings and "single-forcing" experiments with each forcing individually (Text S1 in Supporting Information S1).The CESM-LME provides fairly good simulations of ENSO for us to investigate ENSO in the last millennium (Han et al., 2023;Stevenson et al., 2019) although the amplitude of ENSO is overestimated relative to observation.To single out the influence of solar activity, we use the data from the solar-only forcing (changes in total solar irradiance (TSI) only) experiment during AD 850-1850 and analyze the ensemble mean of four ensembles.However, one may consider greenhouse gases (GHG), the key forcing of global warming increasing dramatically after industrialization, as another forcing candidate although the concentrations were relatively stable before the 1850s.In light of this, we investigate the response of monsoonal precipitation to pre-industrial GHGs and compare it with the influences from solar activity in the Supporting Information S1 (Text S2).
To determine the months of Asian Summer Monsoon that we focus on, we compare the total precipitation rate and 850-hPa wind data set from the CESM-LME 850 control experiment (Figures 1a-1c) and the precipitation data set from the Climate Prediction Center (CPC) (Chen et al., 2008) (Figures 1d and 1e).Although the CESM can just simulate precipitation rate, its consistent patterns to the observation validate the CESM simulation of the Asian monsoonal precipitation.Both of them show sub-seasonal changes in precipitation distribution before and during the typical rain season of Asia Summer Monsoon, regarded as the months from May to August (Wang & LinHo, 2002).During March and April, precipitation mainly concentrates in eastern China and the maritime continent (Figure 1a).Then, strengthened south-westerly winds during early summer enhance the precipitation on the windward side of South Asia (Figure 1b).While stronger winds during peak-summer correspond to more precipitation over South Asia, the precipitation over the leeward side of Southeast Asia is relatively less due to the topographical blocking (Figure 1c).Based on the results, separating early-and peak-summer precipitation would benefit our understanding of the Asian Summer Monsoon response.

Multi-Linear Regression (MLR)
To separate the influence of solar activity from other forcing factors, we utilize the MLR analysis that has been frequently used to isolate the impacts of variability sources on climate variables (Brugnara et al., 2013;Hood et al., 2013;Lean & Rind, 2008;Ma et al., 2019;Wang et al., 2021).A climate variable T can be expressed as the function of a space vector x and time t through the following equation: C indicates the regression coefficient of each climate index and the following four climate indices are included in the equation.SA denotes the solar activity, usually represented by the extended winter (October-March) mean of the sunspot number in previous studies.Given the lack of observational data in the last millennium, we replace the sunspot number with the annual mean of TSI to represent solar activity.ENSO is the Nino3.4index that is derived using the seasonal (December-February) mean of averaged sea surface temperature in the Nino3.4region (5°S-5°N; 120°W-170°W).AOD is the bimonthly (December-January) mean of stratospheric aerosol optical depth over the Northern Hemisphere, representing volcanic influences.TREND denotes a linear trend term that roughly captures anthropogenic radiative forcing, including GHGs, reflective aerosols, land cover changes, and forced cloud changes.
Because of the experimental design in the solar-only forcing experiment, the AOD and TREND terms can be ignored and the contribution of the solar activity and ENSO will be mainly discussed.The residual term is represented by the parameter ε.To ensure the regression coefficient can be tested by Student's t-test, we deal with the potential autocorrelation in the residual term by using the prewhitening procedure that is repeatedly employed until most of the grids satisfy the Durbin-Watson test (Chen et al., 2015).Subsequently, we utilize a two-tailed Student's t-test to measure the statistical significance of regression coefficients.Besides, a similar MLR analysis is conducted to the GHG-only forcing experiment with a slight adjustment to the GHG-based function (Text S2 in Supporting Information S1).

Results
Through the MLR analysis, the regression coefficient of precipitation on solar activity and ENSO is demonstrated in Figure 2. The solar signals in the early summer precipitation (Figure 2a) show an almost opposite pattern to the ENSO signals (Figure 2b) but a similar pattern to climatological rainfall (Figure 1b) over South and East Asia, suggesting enhanced subseasonal rainfall due to increased TSI.The increased rainfall corresponds to enhanced cross-equatorial low-level winds and is mainly located on the windward side of India and Indochina, particularly with statistically significant signals over Southeast Asia.While both Indian and East Asian Monsoon could be responsible for the increased rainfall over Southeast Asia, the consistent winds in climatological and regressed patterns imply the crucial responses of South Asian Monsoon to the variation of TSI.On the contrary, during the early summer that follows a warm ENSO phase, anomalous southward winds take place on the Indian Ocean aligning with decreased precipitation (Figure 2b).An anomalous anti-cyclone is shown over the western Northern Pacific, regarded as the key system connecting ENSO to the AM (Chung et al., 2011;Li et al., 2017;Wang et al., 2000;Zhang et al., 1996), and accounts for increased rainfall around the south of Yangtze River.
Overall, these results suggest that high solar irradiance tends to increase early summer precipitation, especially in Southeast Asia, in contrast to the reduced precipitation during the warm phase of ENSO.
Compared to the precipitation anomalies during early summer, the ENSO signals during peak-summer illustrate a different pattern (Figure 2d), including enhanced rainfall over the northern Indian Ocean and around northern China due to the northward displaced anomalous anti-cyclone.The distinct spatial structures indicate the different impacts of ENSO on AM from early summer to peak-summer.On the other hand, the solar signals during peak-summer intensify the precipitation around Southeast Asia and contrast with ENSO-related rainfall anomalies in other regions, such as China and the Arabian Sea (Figure 2c).Although solar signals indicate enhanced rainfall over Southeast Asia in both early-and peak-summer, statistically insignificant signals in peak-summer and the distinct ENSO signal patterns in the two sub-seasons prompt us to differentiate them, instead of seasonal analysis, to identify the responses of AM to the climate forcing factors.Moreover, given that the opposite precipitation patterns between solar and ENSO signals are more obvious over the northern Indian Ocean during early summer and that the solar activity effects on the early summer AM are less explored by previous studies, we will focus on the solar signals during May and June in the following subsections.
To investigate the change of atmospheric circulations that is concurrent with the anomalous early summer precipitation in Asia, Figures 3a and 3b shows anomalous winds in the upper troposphere, and Figures 3c and 3d illustrates a vertical structure of anomalous temperature and meridional circulations in a meridional section (60°E-90°E).In climatology, an upper-tropospheric anti-cyclone is situated over the southern edge of the Tibetan Plateau (TP) during June-July-August, and is attributed to the development of the South Asian High before the peak-summer (Reiter & Gao, 1982;Wu & Zhang, 1998), the monsoon onset vortex over the Bay of Bengal (Liu et al., 2013), and intense heating associated with the summer monsoon rainfall (Ueda et al., 2022).In Figure 3a, enhanced updraft in the middle troposphere corresponds to the increased precipitation over South Asia (Figure 2a).Meanwhile, an anomalous anti-cyclone takes place above TP, implying an earlier summer monsoon onset around the northern Indian Ocean.Its south branch corresponds to the southward high-level wind anomalies and facilitates the upper divergence pumping, demonstrating a favorable environment for convection over South Asia.The anomalous high-level winds indicate that solar signals benefit the formation of upper-tropospheric anti-cyclone over TP during early summer, suggesting quicker or earlier development of the South Asian Summer Monsoon.
Additionally, Figure 3c displays the vertical profile of anomalous meridional circulations, including landward low-level southerly winds, updrafts over the southern slope of TP, and southward cross-equatorial winds in the high level.Concurrent with the response of meridional circulation that mechanically favors moisture transport from the Indian Ocean, anomalous warming takes place above TP.Climatologically, the temperature increase in the upper troposphere around TP is responsible for the reversal of the meridional temperature gradient that coincides with the onset of the Asian Summer Monsoon (Li & Yanai, 1996;Zhang et al., 2004).Our results suggest that when the Northern Hemisphere receives more direct sunlight during early summer, higher solar activity adds a slight increase in the solar irradiance at the top of the atmosphere and diabatic heating over South Asia.The forced anomalous updraft might lead to the upward transport of diabatic heating, partially accounting for the warming anomalies above TP.The corresponding southerly wind anomalies near the surface (Figure 3c) tend to transport water vapor from the Indian Ocean and favor maintaining or strengthening the diabatic heating over land.Coupling with monsoonal circulations, the anomalous meridional circulation strengthens the onset of South AM during early summer and therefore intensifies regional precipitation.
The ENSO signals around the northern Indian Ocean again show nearly opposite circulation patterns to the solar signals (Figures 3b and 3d), such as an anomalous divergence near the surface of the maritime continent (Figure 2b) and upward motion in the western Indian Ocean (Figure 3b).The associated northward wind anomalies in the upper troposphere converge with the south branch of the anomalous cyclone above TP, causing anomalous subsidence over South Asia.The anomalous descent (Figure 3b) may be related to the low-level warming anomalies due possibly to less cloud coverage and more associated clear-sky influx of solar radiation.The cooling anomalies above TP result from the decrease of upward diabatic heating transport.All the ENSO signals suggest that a warm ENSO phase suppresses the development of the Asian Summer Monsoon during early summer (Liu et al., 2015).
In Figure 4, we display solar and ENSO signals in vertically integrated moisture transports, including the divergence (Figures 4a and 4c) and the meridional transport in a longitudinal section (60°E-120°E) (Figures 4b  and 4d).The solar signals suggest an anomalous convergence of moisture transport over South and East Asia (Figure 4a), consistent with the positive precipitation anomalies (Figure 2a).The cyclonic anomalies in the Bay of Bangel and the enhanced south-westerly winds in the South China Sea (Figure 2a) manifest the development of a monsoon vortex triggered by increased solar irradiance.In particular, the responded low-level winds are critical to the north-eastward transport of water vapor (Abe et al., 2013;Liu et al., 2013;Wu et al., 2012).The positive anomalies of meridional transport quickly diminished from 15°N to 30°N (Figure 4b), reiterating the anomalous convergence of moisture transport in the domain contributed to the intensified local precipitation.In contrast, ENSO signals show negative anomalous meridional transport (Figure 4d) and less water vapor being transported toward the land.The low-level easterly wind anomalies result in decreased moisture transport over South Asia (Figures 2b and 4c), and thereby cause reduced precipitation during early summer following a warmer ENSO phase.

Discussion and Conclusion
We examine the effects of solar activity on the early summer monsoonal circulation over South and East Asia based on the solar-only experiments from the CESM-LME project.Through MLR analysis, isolated solar signals demonstrate increased precipitation patterns over Southeast Asia, which is opposite to ENSO signals and is attributed to the well-coupling of anomalous monsoonal circulations and corresponding meridional moisture transport over the northern Indian Ocean.Due to the hemispheric asymmetry of solar heating in boreal early summer as well as increased solar irradiance during high solar activity years, an enhanced meridional contrast of surface temperature is set up in the northern Indian Ocean and corresponds to an anomalous meridional flow.This consequential meridional circulation provides positive feedback to the background meridional temperature gradient through the upward transport of diabatic heating and the low-level moisture transport toward South and East monsoonal Asia.Additionally, the upper branch of meridional circulation couples with an anomalous anti-cyclone above TP, accounting for an anomalous divergence pumping in the upper troposphere that intensifies the convection over Southeast Asia.
Our results highlight the significance of separating forced responses of AM into early and peak summer because the two sub-seasons exhibit distinct responses to solar activity (Figures 2a and 2c).We emphasize the effects of solar activity on AM during early summer to complement earlier studies which mostly focus on winter and peak-summer (Jin et al., 2019;Ma et al., 2021;Wang et al., 2021).According to the observation and reanalysis data sets, ENSO and solar signals are both recognized as important factors in the East Asian winter climate (Wang et al., 2021) because of their positive correlation with the precipitation in South China.During boreal summer, higher solar activity is suggested to intensify tropical precipitation over the Indian Ocean and western Pacific (van Loon et al., 2004).Besides, a "northern wet-southern dry" pattern over East Asia tends to be triggered in the warm season (May-September) during peak years of a stronger 11-year solar cycle (Jin et al., 2019).Such a pattern is also exhibited in model simulations (Sun, Liu, Wang, Chen, Wan, & Wang, 2022).In our results, early summer solar signals suggest south-westerly wind anomalies over the South China Sea (Figure 2a), accounting for the intensified precipitation in southern China, but do not display an anomalous cyclone to counteract the anomalous anti-cyclone shown in ENSO signals (Figure 2b), which is suggested to be moderated by solar signals as well (Ma et al., 2021).The differences between our results and previous studies indicate distinct forced responses of the Asian Summer Monsoon in sub-seasonal and seasonal scales, while further comparison is necessary because of the different timeseries and types of data set we used.Notably, our diagnoses agree that differentiating early-and peak-summer is essential to decipher the development of the South Asian monsoon system.
Regarding the physical process that the solar variability affects climate systems, there are two plausible mechanisms (Gray et al., 2010) could amplify the influences from the variation of solar irradiance and leave considerable impacts on the climate.The first is the "top-down" mechanism that results from the stratospheric response to solar activity.The ultraviolet absorption from the stratospheric ozone causes a meridional temperature gradient in the stratosphere, thereby modifying tropospheric circulations through troposphere-stratosphere coupling (Andrews et al., 2015;Kodera, 2004;Kodera & Kuroda, 2002;Lin et al., 2021).The second is the "bottom-up" mechanism that is triggered by the solar heating at the surface, altering the surface energy budget and then amplifying the response of surface temperature through adjusting atmospheric circulations and air-sea coupling (Meehl et al., 2008;Misios et al., 2016;van Loon et al., 2004).The two mechanisms are found to be geographically dependent, but they work together in some places (Meehl et al., 2009).Our work is not meant to pick one mechanism over the other.The increased solar influx during early summer could induce anomalous warming over Indian subcontinent and the northern Indian Ocean through the two amplification mechanisms, resulting in an increased meridional temperature gradient above TP (Figure 3c).We highlight the importance of setting up the anomalous temperature gradient that suggests plausibly earlier monsoon onset during high solar activity years.
The significance of solar activity in modulating the change of AM has been supported by several paleo-records during the late Holocene (Duan et al., 2014;Hong et al., 2001;Tiwari & Ramesh, 2007;Wang et al., 2019).A stalagmite record from north-western Vietnam (Nguyen et al., 2020) reflected the intensity of the Asian Summer Monsoon and reported the evolution during the Heinrich event 3 in the last glacial period, suggesting that the events on centennial timescale may correlate with the solar Suess cycle (Suess, 1980;Wagner et al., 2001), known as an approximately 200-year solar cycle.Our study provides physical insights into how the South Asian Summer Monsoon responds to the variation of solar irradiance, based on the solar-only experiment that involved decadal and centennial solar cycles in the last millennium.Although the variation only accounts for 1 ∼ 2 W/m 2 to the total solar energy, the impact on the hydrological cycle in monsoonal Asia is nontrivial, suggesting the importance of considering solar activity in future climate projections.Additionally, we explain the solar effects that partially counteracts the ENSO-related precipitation around the northern Indian Ocean and point out an implication of how the signal of high solar activity could be recorded by proxy data set although ENSO acts as a significant factor on Asian monsoonal precipitation.Nonetheless, considering the solar forcing has increased since the end of the Maunder Minimum in 1715, further examinations of the variation of AM in these centuries are needed to explore the role of solar activity versus ENSO in Asian precipitation.
Aside from the effects of the solar cycle and the teleconnection with other climate variability, the variation of AM could be influenced by natural or anthropogenic external forcing.For instance, volcanic forcing (Sun, Liu, Wang, Chen, & Gao, 2022) and anthropogenic aerosols (Ganguly et al., 2012) can alter the effective solar radiation and then affect regional surface temperature and precipitation.Besides, GHG forcing can amplify regional circulation responses to solar forcing (Meehl et al., 2003).In terms of the MLR analysis in the GHG-only experiment (Figure S1 in Supporting Information S1), the GHG signals in precipitation during early summer also partially counteract ENSO signals, yet they are geographically and dynamically different from the solar signals (Text S2).Although their radiative magnitudes are similar and both contrast with the ENSO-related precipitation over Southeast Asia, the variation of a solar cycle is more crucial to the local rainfall based on statistical significance.Given the additional forcings as mentioned, while the direct and effective solar radiative forcing could almost explain the centennial variations of the global monsoon precipitation before industrialization (Liu et al., 2009), more analyses are necessary as we further explore the impacts of climate forcings on the variations of Asian Summer Monsoon.

Figure 1 .
Figure 1.Bimonthly mean of total precipitation rate (shaded) and 850-hPa winds (vector) during (a) March-April, (b) May-June, and (c) July-August in climatology from the CESM-LME 850 control run (AD 850-1850).(d-f) Show the bimonthly averaged precipitation in climatology based on the data from the Climate Prediction Center (CPC) during 1979-2009.

Figure 2 .
Figure 2. Regression coefficients of precipitation (shaded) and 850-hPa horizontal winds (vector) employed in the multi-linear regression analysis.(a, b) display the solar and ENSO signals during May-June, respectively.(c, d) are similar but display the regression map during July-August.The solid black dots denote the regions where the coefficients (shaded) are statistically significant at 95% confidence level.

Figure 3 .
Figure 3. Regression coefficients of (a, b) 500-hPa omega (shaded: the positive means anomalous downward motion), 200-hPa horizontal winds (vector), (c, d) zonal averaged (60°-90°E) temperature (shaded), omega (contour: red lines denote anomalous downward motion), and meridional winds (vector).We show solar signals during early summer, including their (a) horizontal patterns in higher levels and (c) the vertical structures of temperature and meridional circulations over South Asia.(b, d) Are similar to the left panels but display the patterns of ENSO signals.The solid black dots denote the regions where the coefficients (shaded) are statistically significant at 95% confidence level.

Figure 4 .
Figure 4. Regression coefficients of vertical integrated moisture transport (MT) during early summer.The solar signals in moisture transport includes (a) magnitude, direction (vector), divergence (shaded: the positive means anomalous divergence), and (b) the zonal averaged (60°E-120°E) meridional transport (positive: anomalous northward transport).(c, d) are similar but illustrate ENSO signals.The solid black dots in (a, c) denote the regions where the coefficients (shaded) are statistically significant at 95% confidence level.