Decadal variability of northern Asian winter monsoon shaped by the 11-year solar cycle
Climate signals associated with 11-year sunspot cycle have been extensively studied in various regions of the northern hemisphere, but the precise mechanisms remain elusive. Asian winter monsoon (AWM) is the most powerful circulation system on the Earth, yet its relationship with the 11-year solar cycle has not been explored. Here the response of AWM to the 11-year solar forcing is explored by analysis of numerical experiment results obtained from the Community Earth System Model-Last Millennium Ensemble (CESM-LME) modeling project. We show that a strong 11-year solar cycle can excite a resonant response of the intrinsic leading mode of the AWM variability, resulting in a significant signal of decadal variation. The leading mode, characterized by a warm Arctic and cold Siberia, responds to the maximum solar irradiance with a peculiar 3 to 4-year delay. We propose a new mechanism to explain this delayed response, in which the 11-year solar cycle affects the AWM via modulating Arctic sea ice variation during the preceding summer. At the peak of the accumulative solar irradiance (i.e., 4 years after the maximum solar irradiance), the Arctic sea ice concentration reaches a minimum over the Barents–Kara Sea region accompanied by an Arctic sea surface warming, which then persists into the following winter, causing Arctic high-pressure extend to the Ural mountain region, which enhances Siberian High and causes a bitter winter over the northern Asia.
Keywords11-year solar cycle Asian winter monsoon Decadal variation Arctic sea ice Arctic warming
The relationship between solar activity and Earth’s weather and climate has attracted enormous attention in climate research community over the past century (Siscoe 1978). Numerous model studies have suggested that the spectrum solar irradiance interacting with ozone can strongly affect stratospheric temperature and circulation, which then propagate downward to alter tropospheric general circulation (Haigh 1996; Matthes et al. 2006). Statistical correlations between the 11-year solar activity and the decadal climate variability have been found in the observations (Currie 1993; Soon 2005; Van Loon and Meehl 2012). At the peak years of the 11-year solar cycle, the climatological precipitation maxima in the tropical Pacific were found to be strengthened, further modulating the Pacific climate system, i.e., Pacific Decadal Oscillation (PDO) and El Niño-Southern Oscillation (ENSO) (Van Loon et al. 2007; Meehl et al. 2008, 2009). While for the Northern Hemisphere (NH) winter, links between the solar 11-year variability and the North Atlantic Oscillation (NAO) have been extensively investigated, which result in diverse views (Ineson et al. 2011; Scaife et al. 2014; Thiéblemont et al. 2015; Gray et al. 2016; Chiodo et al. 2019).
Understanding the physical processes involved in solar-climate connections is crucial to interpretation of the observed climate variability and predictability and to the projection of future climate change. A major process of solar influence that has been identified is via a ‘top-down’ mechanism. The ultraviolet radiation variability on heating rates in the tropical upper stratosphere may affect the meridional temperature gradients and the zonal mean wind anomalies, which then migrate poleward and downward through wave-mean flow interaction (Haigh 1996; Andrews et al. 2015). Another mechanism is the ‘bottom-up’ coupled air-sea mechanism. Increased total solar irradiance (TSI) over cloud-free regions of the subtropics translates into greater evaporation, and the resulting moisture is carried to the convergence zones by the trade winds, thereby strengthening the intertropical convergence zone (ITCZ), affecting the Pacific climate system (Meehl et al. 2008, 2009).
Asian winter monsoon (AWM) is the most powerful circulation system in the NH winter. Comparing with the complex structure of the Asian summer monsoon (Wang et al. 2001; Zhou et al. 2009), the surface circulation of the AWM is simply dominated by the gigantic Siberian–Mongolian High (Chang et al. 1983). Jin et al. (2019) had demonstrated that the decadal variation of the East Asian summer monsoon is possibly affected by the 11-year solar cycle through changing North Pacific decadal oscillation. It is expected that the 11-year solar cycle might also have a footprint in AWM variability. So far, the relationship between AWM and the 11-year solar cycle has not been explored. Therefore, we are curious about to what extent the 11-year solar cycle may affect AWM. In particular, what are the characteristics of AWM that are modulated by the 11-year solar cycle? If there is a linkage between them, how can an enhanced solar irradiance affect NH winter climate when the spot of direct sunlight moves toward the Tropic of Capricorn of the southern Hemisphere? The present work aims at addressing these questions.
It is of great difficulty to distinguish the 11-year solar cycle signal from the short-term observations of the AWM, because the amplitude of 11-year cycle is relatively small (Haigh 1996) whereas the AWM variability is affected by a variety of factors, including natural internal variability (e.g., ENSO, NAO) and other external (e.g., volcanic, anthropogenic) influences (Frame and Gray 2010; Gray et al. 2013). To isolate the impacts of the 11-year solar cycle, we analyze numerical simulation results derived from four solar-only forcing experiments and one control experiment (CTRL), which were conducted by the Community Earth System Model-Last Millennium Ensemble (CESM-LME) modeling project (Otto-Bliesner et al. 2016).
In this paper, we first explore the spatiotemporal variability of the AWM by using the monthly surface temperature data derived from the 20th Century Reanalysis V2 data provided by the NOAA/OAR/ESRL PSD (Compo et al. 2011), and then validate the model performance of the CESM in reproducing the AWM variability in Sect. 3. In Sect. 4, we explore the variation of the AWM modulated by the solar activity on decadal time scale, and examine the linkage between the decadal variation of AMW and the 11-year solar cycle. Then, the possible processes by which the 11-year solar cycle impacts the northern Asian winter are discussed in Sect. 5. In Sect. 6, we elaborate a new mechanism by which the 11-year solar cycle could affect Eurasian winter temperature by modulating the preceding summer and autumn Arctic sea ice variation. The last section presents major conclusions and discusses remaining issues.
2 Numerical experiments and observational data
The CESM-LME simulation data provided by NCAR’s Climate Change Research Section is an ensemble of fully-forced, and single-forced simulations. In the CTRL experiment, with the fixed external forcings at AD 850, climate variations are solely arising from the coupled dynamics internal to the Earth’s climate system. In the four solar-only forcing experiments, the varying solar forcing was applied but all other external forcings were fixed at the condition of AD 850. Changes in the total solar irradiance (TSI) were prescribed by the reconstructed solar forcing (Vieira et al. 2011). The four-member ensemble average derived from the solar-forcing experiments can largely reduce the noises of the internal variability, so that the impacts of solar forcing on AWM variations can be better identified by comparison with the CTRL experiment. Due to the low-top atmospheric model with no prognostic ozone chemistry in these numerical experiments, there is no response of the stratospheric ozone change to the solar intensity variation, the response to solar forcing shown here is coming from the “bottom-up” mechanism.
In addition to the simulation results, we have used the surface air temperature derived from the 20th Century Reanalysis Project version 2 dataset during 1871–2000 to explore the characteristics of the Asian Winter Monsoon and verify the simulation results of the CESM_LME.
3 Characteristics of the AWM in the observation and control experiment
To validate the model’s capacity in reproducing AWM variability and make a fair comparison, we randomly selected a 130-year sample from the CTRL experiment. The spatial patterns of the EOFs in the control run are consistent with the corresponding observed EOF modes (Fig. 2c, d), while there exist large differences in the periodicity of the PC1 (Fig. 3). In the CTRL (Fig. 3b), the interannual energy peaks are evidently stronger than those in the observed data, and the decadal variation has a significant quasi-15-year peak. The leading mode in the CTRL represents an intrinsic internal climate variability in the model without any external forcing, which is different from the observation. Since the various forcing factors are in action in the observed record, the temporal behaviors of the leading modes in the CTRL and observation are not directly comparable. However, the model from the CESM-LME simulates realistic spatial structure of the leading modes of the AWM, suggesting that the most frequently observed spatial pattern of the SAT in the model is realistic.
4 Evidence of 11-year solar cycle influence on decadal variation of the AWM
5 Possible processes by which the 11-year solar cycle impacts the northern Asian winter
The 3–4 years phase-delayed response is a peculiar feature. This peculiar feature is further perplexed by the fact that during winter, solar irradiance flux has little impacts on the northern high latitudes. The solar irradiance cycle should have strongest climate consequence during boreal summer when the northern hemisphere, especially the northern high latitudes, receive maximum solar radiation heating. Thus, the direct effects of solar forcing cannot explain this simulated winter phenomenon. In this section, we explore the two possible processes that can elucidate the aforementioned dilemmas.
According to the thermodynamic equation, the maximum solar irradiance heating rate should result in a maximum rate of temperature change not the temperature itself. Assuming the solar irradiation flux is proportional to the diabatic heating rate, the accumulative TSI represents the accumulative solar heating effects. The result shown in Fig. 6b indeed indicates a generally simultaneous correlation (but most significant correlation with slight 1–2 years delay) between the accumulated TSI and the leading mode of the AWM on decadal timescale during the strong 11-year solar cycle epoch.
Ocean is an enormous heat reservoir. For the phase-delayed and seasonally-delayed responses, we hypothesized that the impacts of the enhanced summer solar forcing may be ‘memorized’ by ocean warming, which then feedback to the atmosphere later, affecting the northern Asian winter temperature. The temperature response should correspond to the accumulative solar heating effects. Evidences have shown that a small La Niña-like response in the peak solar year, followed a couple years later by a larger central Pacific El Niño-like response (Meehl and Arblaster 2009; Misios et al. 2016, 2019), which could induce a cooler-than-normal winter over Eurasia as shown by an AGCM experiment (Zhang et al. 2019). Other studies have shown that the cold Eurasia could be a consequence of the rapid decline of the preceding summer Arctic sea ice, especially over the Barents–Kara sea region where sea ice has largest seasonal and interannual variability (Honda et al. 2009; Cavalieri and Parkinson 2012; Inoue et al. 2012).
6 Key mechanisms linking the 11-year solar cycle, Barents Sea ice, and cold winter in the northern Asia
Second, the “Arctic memory” hypothesis can also explain the seasonally phase-delayed response. With the reduction of the sea ice over the Barents Sea during summer and autumn, the adjacent ocean is generally warmer than normal, which can persist into the ensuing winter due to the large heat capacity of the ocean, so that the DJF mean SAT exhibits abnormal warming over the Barents Sea and the Greenland Sea region (Fig. 7a). The winter Arctic warming features high-pressure anomalies over the entire Arctic (Fig. 8c). The winter Barents Sea warming can extend the Arctic high-pressure to Ural mountain region, which then perturbs westerly jet stream and further enhances and shifts the downstream EA trough westward (Kug et al. 2015) (Fig. 8c). On the surface, the Mongolian–Siberian High shifts northwestward over Siberia, favoring the cold Arctic air mass intrusion into the high latitudes of Asian continent (Luo and Wang 2018). The Asian coldness extends from the Ural mountain all the way to the central North Pacific (Fig. 7a).
The explanation of the seasonal delayed response supports the observations. In fact, Cavalieri and Parkinson (2012) have suggested that the cold Siberian winter is significantly influenced by the diminishing Barents Sea ice in the preceding autumn. Luo and Wang (2018) have used the September–October mean SST anomalies in the Barents Sea region to predict the ensuing winter extremely cold days over the northern East Asia. The importance of the variation of the autumn Barents Sea ice in the winter climate system over Eurasia was also emphasized by Zhang et al. (2018).
7 Conclusion and discussion
The leading mode of the Asian winter monsoon (AWM), characterized by a cold northern Asia between 40 and 70°N, is an intrinsic mode of AWM variability. However, the external forcing provided by strong 11-year solar cycle can excite a significant decadal signal in the leading AWM mode.
The response of the leading mode of the AWM reaches the strongest phase 4 years after the maximum solar irradiance. Thus, the 11-year solar cycle-excited decadal variation of the AWM is approximately in phase with the accumulated total solar irradiance.
The influence of the 11-year solar cycle on the AWM is mainly via modulating Arctic sea ice variation over the Barents–Kara Sea region during the preceding summer and fall. The Arctic sea ice melting reaches a maximum after a 4-year delay following the peak of the solar irradiance. The summer melting of sea ice causes winter Arctic warming north of 70°N, which, in turn, generates extensive coldness over northern Asia and North Pacific through strengthening the Ural mountain pressure ridge and enhancing the Siberian High in the high latitude of Asia.
The new mechanism proposed in the present study emphasizes that the 11-year solar forcing can affect decadal variability of the northern hemisphere winter climate via its accumulative heating effects that are “memorized” by the Arctic sea ice melting during the preceding summer. This mechanism can explain two dilemmas, one is why the solar forcing variability in summer can affect winter climate, and the other is why there is a delay between the maximum solar irradiance and the AWM response, i.e., the Barents Sea warming and Siberian cooling. Our analysis is consistent with the result of Kug et al. (2015) in the sense that the Barents–Kara Sea warming induces the Ural mountain high anomaly, which can further shift the Mongolian–Siberian High northwestward, favoring intrusion of the cold Arctic air mass into the high latitudes of Asian continent. However, the Arctic high-pressure anomalies were not produced by the Barents–Kara Sea warming in the coupled global climate model experiments. This is not surprising because the sea ice and SST anomalies are not confined in the Barents–Kara Sea region (Fig. 8a). More appropriate experiments are required to distinguish the complex local atmosphere–ocean–sea ice interaction and remote atmospheric forcing. Meanwhile, the role of the 11-year solar cycle-induced Pacific warming on the cold northern Eurasia cannot be entirely ruled out, that calls for further studies.
Since the simulations do not include the ‘top-down’ effect of solar variability (Otto-Bliesner et al. 2016), the mechanism proposed in this paper depends almost entirely on the increased solar forcing acting on the ocean surface. However, in the presence of the ‘top-down’ mechanism, what is the resultant response of AWM calls for further investigation. The results of our study are based on numerical experiments performed using a single climate system model with a four-member ensemble. More ensemble members are desirable for a more reliable detection of the decadal signals. In addition, multi-model simulation results should be investigated to verify the results obtained from the single model used here.
This work was supported by the National Key Research and Development Program of China (2016YFA0600401), National Natural Science Foundation of China (41420104002, 41671197 and 41631175), Natural Science Foundation of Jiangsu Higher Education Institutions (14KJA170002), Open Funds of State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, CAS (SKLLQG1820), Program of Innovative Research Team of Jiangsu Higher Education Institutions of China, and Priority Academic Program Development of Jiangsu Higher Education Institutions (164320H116). Financial support from the program of China Scholarships Council (NO. 201806860029). The CESM-LME data were generated by the CESM Paleoclimate Working Group at NACR. Support for the Twentieth Century Reanalysis Project dataset is provided by the U.S. Department of Energy, Office of Science Innovative and Novel Computational Impact on Theory and Experiment (DOE INCITE) program, and Office of Biological and Environmental Research (BER), and by the National Oceanic and Atmospheric Administration Climate Program Office. This is publication No. 10770 of the School of Ocean and Earth Science and Technology, publication No. 1399 of the International Pacific Research Center and publication No. 277 of the Earth System Modeling Center.
- Chang C-P, Millard JE, Chen GTJ (1983) Gravitational character of cold surges during winter MONEX. Mon Weather Rev 111:293–307. https://doi.org/10.1175/1520-0493(1983)111%3c0293:GCOCSD%3e2.0.CO;2 CrossRefGoogle Scholar
- Misios S, Gray LJ, Knudsen MF, et al (2019) Slowdown of the Walker circulation at solar cycle maximum. Proc Natl Acad Sci 201815060Google Scholar
- Van Loon H, Meehl GA (2012) The Indian summer monsoon during peaks in the 11 year sunspot cycle. Geophys Res Lett 39:L13701Google Scholar
- Van Loon H, Meehl GA, Shea DJ (2007) Coupled air–sea response to solar forcing in the Pacific region during northern winter. J Geophys Res Atmos 112:D02108Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.