1. Introduction
Known as “the roof of the world,” the Tibetan Plateau (TP) covers an area of more than 2 million square kilometers with an average altitude of over 4000 m above mean sea level (MSL), and has profound effects on the climate in the surrounding region and globally via interactions with atmospheric circulations (Schiemann et al. 2009; Zhang et al. 2017). The TP also affects the water supply to more than 1/5 of the world population as the birthplace of many rivers and is renowned as “the water tower of Asia” (Immerzeel et al. 2010). In recent decades, the TP has experienced more rapid warming than other areas at the same latitude (Zhou and Zhang 2021; Duan et al. 2015), which has affected the fragile ecology of the TP. The rapid warming has been attributed to increasing anthropogenic greenhouse gas forcing and snow/ice albedo feedbacks (An et al. 2016; Rangwala et al. 2010). However, our understanding of the natural climate drivers and their mechanisms remains limited.
The Atlantic multidecadal oscillation (AMO), characterized by cycles of about 20–30 and 40–60 years, has a significant impact on the multidecadal climate variability of the Northern Hemisphere (Dima and Lohmann 2007; Wyatt et al. 2012). During the warm phase of the AMO, the warming sea surface temperature (SST) of the North Atlantic triggers the stationary Rossby waves in the middle and upper troposphere over Eurasia, which propagate eastward and affect the local circulation patterns in different regions, subsequently influencing their climates (Li et al. 2008; Luo et al. 2011; Wang et al. 2015), such as inducing cyclonic systems that promote warming in East Asia (Ratna et al. 2020; Yang et al. 2021) or anticyclonic systems that cause cooling in central Asia (Dong et al. 2023). Observations and climate model simulations have shown that the AMO strongly affects the TP surface temperature and heat sources (Feng and Hu 2008; Li et al. 2008; Luo et al. 2011). The effects of the AMO on the temperature of the TP can be traced back thousands of years as shown by proxy data (Wang et al. 2015). The teleconnections between the TP and the Atlantic may be attributed to the propagation of Rossby waves (Wang et al. 2015; Shi et al. 2019). Further research is needed to fully understand the potential mechanisms linking the AMO and temperatures in the TP.
Volcanic eruption is another primary natural forcing influencing the temperature variations in interannual to decadal time scales (Robock 2000; Sigl et al. 2015). Sulfate aerosols from large volcanic eruptions spread globally within a few months and remain in the stratosphere for 1–2 years, blocking part of the solar radiation and thus causing a decrease in global surface temperature (Robock 2000). Furthermore, multiple large volcanic eruptions in a short period of time can trigger changes in ocean temperature and heat content thus leading to cold periods over decadal or even multidecadal scales (Church et al. 2005; Zanchettin et al. 2013, 2012). Both instrumental and proxy data show that large volcanic eruptions have triggered low temperatures in the TP during instrumental and historical periods (Duan et al. 2018; Hao et al. 2016; Li et al. 2017; Wang et al. 2017). However, there are limited relevant studies in the TP, and they report varying cooling times following large volcanic eruptions (Duan et al. 2018; Li and Shao 2016; Hao et al. 2016). Therefore, further research is necessary to investigate the cooling time following large eruptions and its underlying mechanisms.
Owing to the short duration of the instrumental data in the TP, studying climate variability and climate drivers on longer time scales requires the use of proxy data. The tree-ring maximum latewood density (MXD) is mainly affected by growing season or late growing season temperature of the current year, and has proven to be a reliable proxy for reconstructing temperature variations over hundreds to thousands of years (Briffa et al. 2002; Esper et al. 2012; Frank and Esper 2005). In the past several decades, a number of studies have been conducted to reconstruct the climate based on MXD on local or regional scales and have investigated the temperature variability and its drivers (include ENSO, PDO, AMO, solar activity, volcanic eruption, and anthropogenic influence) in the region on the TP (Duan and Zhang 2014; Duan et al. 2018; Fan et al. 2009; Li et al. 2017, 2018; Liang et al. 2016; Wang et al. 2009; Yin et al. 2021). Nevertheless, previous studies have shown differences in the temperature variability and its drivers due to various factors such as different reconstruction seasons and/or regions. Additionally, observations also indicated substantial spatiotemporal differences in temperature variability within the region (Wang et al. 2015; Duan et al. 2015). Another important issue is that current studies based on MXD are concentrated in the southeast TP. This urgently requires us to explore more tree-ring sampling sites over a larger area.
The purpose of this study was to reconstruct the temperature series using MXD chronology for a new study area in the southern TP. We compared this reconstruction with other temperature reconstructions in this region and the Northern Hemisphere, and to the AMO and various records of volcanic eruption. By doing so, we aimed to achieve a further understanding of the regional temperature responses to the AMO and large volcanic eruptions over the past two centuries.
2. Data and methods
a. Study area
Our study area is situated in Nyingchi City, in the southern TP (Fig. 1), with high forest cover and diversity of tree species. The Smith fir (Abies George var. smithii) is one of the dominant species, which is distributed primarily at an altitude of 2900–4300 m. The study area has a typical alpine canyon and mountain river valley landscape. The high mountains in the north block cold air from the north, while the warm and humid airflow from the Bay of Bengal in the south can move northward along the river valley. Influenced by the South Asian monsoon systems, the climate in the study area is distinguished by warm and humid summers and cool and dry winters. Meteorological records from the Nyingchi station (Fig. 2) show that the annual mean temperature was 9.0°C, with 16.0°C in July (the warmest month) and 0.7°C in January (the coldest month) during 1960–2020. Mean annual precipitation was 674.8 mm with approximately 72% occurring during June–September.
Locations of the tree-ring sampling sites.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
Temperature and precipitation data from the meteorological station in Nyingchi in the southern Tibetan Plateau from 1960 to 2020.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
b. Tree-ring samples and MXD chronology development
We conducted the field campaign in July 2021 and chose the mature Smith fir forest for sampling in the upper tree line of Milin (ML1) (28°42′N, 93°33′E; 4212 m MSL) on the slope facing north, and Langxian (LX1) (28°46′N, 93°05′E, 4310 m MSL) on the slope facing west (Fig. 1). We collected 56 cores from 33 living trees at ML1 and 54 cores from 29 living trees at LX1 using 10-mm diameter increment borers. These cores did not decay and had no broken pieces. Both sampling sites are far away from villages or residential areas and show no signs of human interference.
The tree-ring samples were visually cross-dated, and the tree-ring widths were measured using the LINTAB measurement system with a resolution of 0.01 mm. The quality of measurements and cross-dating were checked using the program COFECHA (Holmes 1983). The tree-ring samples were then processed for densitometry analysis. First, we cut the cores into 3–4-cm sections in length and mounted them on support blocks. Then the wood fiber angles of each section were measured using a dendroscope. Last, we cut the sections into about 1.0-mm slices using a twin-blade Dendrocut. All of the slices were numbered and soaked in 80°C water for 72 h to eliminate resin and sugar (Schweingruber et al. 1988). The slices were left in a room set at 20°C and 50% relative humidity for more than 12 h, X-ray photographs were taken, and grayscale variations were measured using the DENDRO-2003 densitometer. Seven parameters [earlywood and latewood widths, tree-ring width (TRW), mean earlywood and latewood densities, minimum earlywood densities, and MXD] were obtained. MXD data were then cross-dated using the tree-ring width cross-dating pattern, and then quality checked using the program COFECHA. Samples that could not be cross-dated due to poor quality were removed from further analysis. A total of 81 cores were successfully cross-dated (37 cores from Milin and 44 cores from Langxian), and used for dendroclimatology analysis.
The correlations among MXD series from the two sites were high. Therefore, we combined the MXD series from the two sites to develop the chronology using the “signal-free” technique with RCSsigFree software (https://www.geog.cam.ac.uk/research/projects/dendrosoftware/). This technique was used to remove the nonclimatic trends due to age, size, and effects of stand dynamics (Melvin and Briffa 2008, 2014; Fang et al. 2012). The age-dependent smoothing spline was chosen to generate initial chronology indices, and the biweight robust mean was used in order to reduce the effects of outliers (Cook and Kairiukstis 1990). Finally, the SF-RCS MXD chronology was produced after seven iterations. The expressed population signal (EPS) quantifies the representativeness of a sample to the population, with a value above the threshold of 0.85 indicating reliability with a sufficient number of samples (Wigley et al. 1984).
c. Climatic data, volcanic eruption data, and AMO index data
The nearest meteorological station to our sampling sites, Milin (29°08′N, 94°08′E, 2950 m MSL), started monitoring meteorological factors in 1979, while the second nearest station, Nyingchi (29°24′N, 94°12′E, 2992 m MSL), recorded climatic variables from 1960. Therefore, we chose the longer climate records from Nyingchi for reconstruction. In addition, the gridded monthly mean climate data with 0.5° spatial resolution from 1960 to 2020 were used in the analysis. The climatic variables included monthly mean temperature (Tmean), monthly mean minimum temperature (Tmin), monthly mean maximum temperature (Tmax), and monthly precipitation (PPT). All of the climatic data were obtained from the National Meteorological Information Center (https://data.cma.cn/).
The volcanic eruption data were obtained from the Smithsonian Institution’s Global Volcanism Program (GVP) (https://volcano.si.edu/), in which 22 volcanic eruptions with volcanic explosivity index (VEI) ≥ 5 since 1792 were chosen for the analysis. We also obtained the ice-core volcanic index (IVI) data (Gao et al. 2008) and aerosol optical depth (AOD) data (Crowley and Unterman 2013) produced based on ice cores. Furthermore, the monthly AMO SST index data were obtained from the NOAA ESRL Physical Sciences Laboratory (https://psl.noaa.gov/data/timeseries/AMO/). Two warm phases (1927–57 and 1991–2014) and two cold phases (1893–1926 and 1958–90) in the multidecadal period were derived from AMO SST index data for further analysis. In addition, decadal AMO SST data reconstructed by Mann et al. (2009) were also used in this study.
d. Reanalysis data
The gridded geopotential height reanalysis data were obtained from NOAA-CIRES-DOE Twentieth Century Reanalysis version 3 (20CRv3) (https://psl.noaa.gov/data/gridded/data.20thC_ReanV3.html) and ECMWF ERA5 (https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5). Composite analysis of the anomalous geopotential height (AGH) and flow field at 500 and 200 hPa was performed to reveal the large-scale atmospheric circulations during the two warm phases and two cold phases of the AMO for the period 1893–2015. To exclude the effect of warming trends, the AGH fields during the cold phase 1893–1926 and the warm phase 1927–57 were subtracted from the average of the period 1893–1957, while the AGH fields during the cold phase 1958–90 and the warm phase 1991–2015 were subtracted from the average of the period 1958–2014.
The gridded heat flux and radiant flux reanalysis data obtained from the NOAA-CIRES-DOE Twentieth Century Reanalysis version 3 (20CRv3) (https://psl.noaa.gov/data/gridded/data.20thC_ReanV3.html) and ECMWF ERA5 (https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5) were used to investigate the influence of volcanic eruption. The anomalous heat flux and radiant flux field in the years of the large volcanic eruptions and the following two years were calculated using composite analysis. The low-frequency variations in each grid were removed with a loess filter before we performed the composite analysis. Because the ECMWF convention for vertical fluxes is positive downward, the sensible heat flux and latent heat flux are calculated after taking the opposite number, and the upward longwave radiation is calculated by using the net longwave radiation flux minus downward longwave radiation flux.
e. Tree growth–climate relationships, calibration, and verification method
The tree growth–climate relationships and potential target for reconstruction were investigated using Pearson’s correlation analysis. The Pearson’s correlation coefficients between the MXD chronology and climatic variables (Tmean, Tmin, Tmax, and PPT) were calculated over the common period 1960–2020. Because the climate could have a lagging effect on the tree growth in the following year (Fritts 1976), the analysis was conducted over a 14-month climate window extending from the previous September to the current October. To evaluate the tree growth–climate relationships in the high-frequency domain, we also calculated the correlation coefficients using the first differences of the MXD chronology and the climate variables.
Based on the tree growth–climate relationships, a transfer function was derived using the linear regression model in which the MXD was the independent variable and the selected climatic variable was the dependent variable. The split-period independent validation (Fritts 1976) and leave-one-out cross-validation method (Michaelsen 1987) were used to validate the transfer function. The validation statistics included Pearson’s correlation coefficient (R), the sign test (ST), the first-difference sign test (ST1), product means test (PMT), reduction of error (RE), and coefficient of efficiency (CE). The ST, ST1, and PMT are measures of how well the estimated series follows the direction of variation in the actual series (Fritts 1976). Both RE and CE quantify shared variance between actual and estimated series, with positive values indicating that the transfer function is valid (Cook et al. 1994).
To examine the spatial representativeness of the reconstructed series, we calculated the spatial correlation between the reconstructed series and the CRU gridded data over the period 1960–2020 using KNMI Climate Explorer (https://climexp.knmi.nl/).
f. Climatic periodicity and possible climate drivers
Multitaper method (MTM) spectral analysis, continuous wavelet transform, and wavelet synchrosqueezed transform (Thakur et al. 2013) were used to identify the climatic cycle in the reconstruction series. The significance of the MTM analysis was tested by AR1 confidence level estimates and harmonic F test for significance estimates. The candidate cycles are identified via isolation of those frequencies that achieve the 90% “red noise” confidence level and MTM harmonic F test confidence level (Meyers 2012). The significance of continuous wavelet transform was tested using a Monte Carlo resampling procedure. Composite analysis of the geopotential height fields and wind field at 500 and 200 hPa were performed to reveal the large-scale atmospheric circulations related to AMO different periods. The heat flux and radiant flux in the first and second year after large volcanic eruptions were also calculated using composite analysis. The warming rate is derived by calculating the linear regression coefficient of temperature and time, and the significance of the trend is checked with the Mann–Kendall trend test (Libiseller and Grimvall 2002).
3. Results
a. Regional MXD chronology
Among the different detrending methods used to remove the nonclimatic trends and develop the chronology, the signal-free method produced good statistical indicators (Fig. 3) and best captured the low-frequency trend of the temperature series during the period 1960–2020 (Fig. A1 in the appendix). Therefore, we chose this method for detrending and establishing the MXD chronology.
(a) Composite maximum latewood density (MXD) chronology (thin black line) (1792–2020) with an 11-yr spline smoothing line (red heavy line) and the sample depth (gray line). (b) The running expressed population signal (EPS) calculated over 51 years with a 1-yr lag. The dotted line indicates the 0.85 criterion. (c) Rbar statistics calculated over 51 years with a 1-yr lag in the chronology.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
The regional MXD standard chronology (Fig. 3) covered the period 1764–2020. Beginning in 1792, the MXD chronology is considered reliable with sufficient samples as the EPS reached 0.85 with nine cores. The variations in the sample depth, running EPS, and running averaged correlation coefficient among series (Rbar) for the MXD chronology are shown in Fig. 3.
b. Tree growth–climate relationships
For the original data, significant positive correlations were found between MXD chronology and temperature from the previous September to the current October at the 0.01 level, except for the previous October and November with Tmean; previous October and November and current January, February, and October with Tmin; and previous September and October and current January, April–July, and October with Tmax (Fig. 4a). No significant correlation was found to exist between MXD chronology and PPT. For the first-difference data, significant positive correlations were found in the current August and September with Tmean, current May and August with Tmin, and current September with Tmax, and a significant but low negative correlation was found with July Tmin (Fig. 4b). The correlations between MXD chronology and PPT were much higher than in the original data, but still insignificant. The differences between the original and first-difference data indicated that the positive correlations between MXD chronology and temperature primarily resulted from similar variations in the low-frequency domain except for the current August and September temperatures. The strongest correlation was found between the August–September mean temperature and the MXD chronology for the original (r = 0.814, p < 0.0001) and first difference (r = 0.601, p < 0.01) data. A similar correlation between the gridded August–September mean temperature and the MXD chronology for the original data (r = 0.813, p < 0.0001), and a higher correlation for the first-difference data (r = 0.671, p < 0.0001) were found (not shown in the figure). Therefore, the monthly mean temperature in late summer is the major limiting factor of tree growth at our sampling sites in both low- and high-frequency domains.
Correlation coefficients between tree-ring MXD chronology and climatic variables (monthly mean temperature, monthly mean maximum/minimum temperature, and monthly total precipitation) for the (a) original and (b) first-difference data.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
c. August–September mean temperature reconstruction
(a) Scatterplot and (b) graph of the observed and reconstructed August–September temperature for the full calibration period 1960–2020. (c) Reconstructed August–September mean temperatures (black thin line) and 11-yr spline smoothing (red thick line) both based on maximum latewood density (MXD) data from 1792 to 2020; the gray area denotes the confidence interval at 95%, the red dashed lines denote the mean ±1σ, and the black dashed lines denote the mean ±2σ.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
Statistics of calibration and validation. SE is the standard error, ST the sign test, ST1 the first difference sign test, PMT the product mean test, RE the reduction of error, and CE the coefficient of efficiency; also, one asterisk (*) indicates p < 0.05 and two (**) indicate p < 0.01.
d. Temporal variation and climatic cycle
The reconstructed series began in 1792 when the EPS value goes above 0.85 (Fig. 3). The reconstructed temperature exhibited intense interannual- to decadal-scale variations (Fig. 5c), providing a valuable long series for evaluating local climate variability. To remove the effect of warming in recent decades, mean temperature was calculated using values for the period 1792–1990 and extreme temperatures were defined using a range of ±2σ beyond the mean, with σ calculated also using values for 1792–1990. The extreme cold years in the reconstructed series include 1817, 1857, 1885, 1920, 1933, 1962, and 1972, with the coldest year in 1817. The extreme warm years include 1869, 1992, and 17 years during 2001–20 excluding 2002, 2006, and 2013. Furthermore, the warm and cold periods are defined using a range of ±1σ beyond the mean. According to this classification, the reconstruction series shows cold periods during 1797–1803, 1809–18, 1856–62, 1902–07, and 1915–24, with the coldest decade in the past 229 years being the 1810s. The warm epochs occurred during 1822–26, 1846–51, 1872–79, 1895–97, 1931–43, 1980–84, 1989–2020. A rapid warming trend can be observed over the past 30 years, and the past two decades are the warmest in the past 229 years.
Spatial correlation was calculated between the observed and reconstructed (Fig. 6) August–September mean temperature with the gridded Tmean8–9 (CRU TS 4.05) for the period 1960–2020. For the original and first-difference data, the reconstructed August–September mean temperature correlated significantly with gridded temperature over a region covering approximately 25°–35°N and 85°–100°E (r > 0.5, p < 0.1), both showing a similar spatial structure of the correlation between the observed and the gridded temperature data.
Spatial correlation fields of the (a),(b) observed and (c),(d) reconstructed Tmean8–9 with the gridded average August–September mean temperature. Spatial correlation for the (left) original and (right) first-difference data for the period 1960–2020. The green dots show the location of sampling sites, and the black flag shows the location of meteorological station.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
The reconstructed series exhibits cycles of approximately 2–3, 20–30, and 40–60 years. Nevertheless, only the 2–3-yr cycle reaches the 95% significance level, suggesting that the periodic signal of the reconstructed series is weak. However, when we removed the last 20 years of the series, these climatic cycles became significant. We assumed that the warming in the last 20 years was mainly caused by human activities (Zhou and Zhang 2021), so removing this period allowed us to better understand the influence of natural factors on the climate cycles. The MTM analysis, continuous wavelet transform, and wavelet synchrosqueezed transform (Fig. 7 and Fig. A2) indicate the existence of 2–3-, 20–30-, and 40–60-yr cycles in the reconstruction series. Notably, the 20–30-yr climatic cycle is significant before 1900 but disappears after 1900. It is important to note that the periodicity is sensitive to including the last 20 years, which means that the derived climate cycles are not very rigorous. Nevertheless, the results of the periodicity analysis can provide a reference for subsequent analysis of climate drivers.
(a) Multitaper method (MTM) spectral analysis, (b) continuous wavelet transform, and (c) wavelet synchrosqueezed transform of the reconstructed August–September temperature. The black line in (b) denotes the 95% confidence level.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
e. Possible driving factors for temperature variability
The 2–3-yr cycles in the August–September temperature reconstruction series coincided with the cyclicity (2–8 years) of El Niño–Southern Oscillation (ENSO), and our reconstructed series and SSTNiño-3.4 are inversely correlated in first-difference data (R = −0.178, p < 0.05) during 1850–2020. Approximate 2–8-yr cycles have been found in a number of tree-ring studies in the TP (Li et al. 2018; Yin et al. 2021; Liang et al. 2008). Additionally, significant correlations were found between instrumental and reconstructed temperature series in TP and ENSO index (Yin et al. 2021; Zhang et al. 2014; Yin et al. 2000), indicating that ENSO may exert a notable influence on TP temperature. The influence of ENSO on the climate of the TP has been demonstrated and its mechanisms discussed using observation data, reanalysis data, and a climate model (Hu et al. 2022, 2021).
The 10–30-yr and approximately 50-yr cycles in our reconstruction coincided with the cyclicity (10–30 and 50–70 years) of the AMO, which, as reported previously, may have an impact on the study area (Yin et al. 2021; Liang et al. 2016; Wang et al. 2014; Shi et al. 2019). The result of cross-wavelet analysis and wavelet coherence of our reconstruction with reconstructed AMO series by Mann et al. (2009) indicated a strong correlation between two time series for periodicities of 20–30 years in the period preceding the 1900s, and strong correlation for periodicities of 50–70 years during the period 1792–2020 (Fig. A3). A high positive correlation (r = 0.44, df = 213, edf = 140, p < 0.001 for original data and r = 0.79, df = 13, p < 0.001 for 30-yr loess smoothing series) existed between the reconstructed August–September temperature and AMO index reconstructed by Mann et al. (2009) for the overlapping period in 1792–2006. In addition, a high positive correlation (r = 0.45, p < 0.001 for original data and r = 0.64, edf = 5, p < 0.1 for the 30-yr smoothing series) also existed between our reconstruction and August–September AMO during 1856–2020. The warm and cold periods in our reconstruction generally coincided with those of the AMO (Fig. 8).
Comparisons of our reconstruction with the AMO series reconstructed by Mann et al. (2009) and August–September AMO SST index from the NOAA ESRL Physical Sciences Laboratory. The thick lines result from 30-yr loess smoothing.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
The teleconnection between the AMO and the late summer temperature in our study area was shown on the anomalous geopotential height (AGH) field at 500 and 200 hPa (Fig. 9 and Fig. A4). There were differences in the AGH field and the wind field at 500 and 200 hPa between positive and negative AMO periods. During the warm phase of the AMO, positive anomalous centers and cyclonic systems occurred in the midlatitude North Atlantic, northern Europe, and Pacific Northwest, promoting warmer summers in these regions. Negative AGH centers and anticyclonic systems occurred in central Siberia, causing colder summers in this region. In the TP, there were higher geopotential height and divergent horizontal wind field at 500 and 200 hPa, which were conducive to air mass vertical sinking, causing more sunny days and warmer summers. Additionally, the stronger southerly winds transported more heat from lower latitudes to the TP, also leading to warmer summers. The opposite AGH spatial distribution and wind field yield occurred during the cold phase of the AMO, inducing more rainy days and cooler summers. It is worth noting that the AGH field in each warm phase and each cold phase of the AMO since 1893 shows generally consistent spatial patterns with those in Fig. 9 (not shown in the figure), which also further verifies our statement.
(a),(b) 500- and (c),(d) 200-hPa anomalous geopotential height field (shade) and anomalous flow field (velocity) during the AMO (left) cold and (right) warm period in the 20CRv3 dataset.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
The 1815 Tambora eruption may have been the largest volcanic eruption in the past 650 years (Gao et al. 2017; Sigl et al. 2015). Low temperature occurred in 1816 and 1817 in our reconstructed temperature series, and 1817 was one of the coldest years in the past 208 years. Low temperature in 1817 has been reported in previous studies of the southeast TP (Wang et al. 2009; Li et al. 2017). Volcanic eruption data from different sources were also used to perform superposed epoch analysis (SEA). The results showed the postvolcanic August–September cooling is strongest in the second year after an eruption, for eruptions with an IVI > 20 based on data from Gao et al. (2016) or VEI ≥ 5 based on data from GVP (Figs. 10a–d). Volcanic eruptions with VEI ≤ 4 had no significant impact on the temperature of the study area (not shown in the figure). A volcanic eruption in a tropical area could cause a greater temperature decrease than a volcanic eruption in an extratropical area (Figs. 10e,f).
Superposed epoch analysis (SEA) testing the impact of large volcanic eruptions on the August–September temperature in the southern TP. SEA based on (a) 10 volcanic eruptions with ice-core volcanic index (IVI) > 10, (b) 5 volcanic eruptions with ice-core volcanic index (IVI) > 20, (c) 5 volcanic eruptions with VEI > 5, (d) 22 volcanic eruptions with VEI ≥ 5, (e) 10 tropical volcanic eruptions with VEI ≥ 5, and (f) 12 extratropical volcanic eruptions with VEI ≥ 5. Solid (dashed) lines represent the 99% (95%) confidence limits obtained by performing a Monte Carlo significance test.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
Large volcanic eruptions may cause a cold period at decadal and multidecadal time scales (Sigl et al. 2015; D’Arrigo and Jacoby 1999; Esper et al. 2013). Multiple eruptions in combination with a lower solar minimum phase may have caused the 1810s to be the coldest decade. The cold periods in the reconstructed series coincide with episodes of high volcanic forcing values, and the warm periods occur during gaps in large volcanic activity (Fig. 11), which has also been reported in a previous study (Li et al. 2017) of the southeast TP. However, the relationship between our reconstructed temperature and volcanic eruptions weakens in the last 100 years as shown in Fig. 11. Several large volcanic eruptions occurred in the twentieth century, and no corresponding cold periods were found in the reconstructed series. Correlation analysis between the reconstructed August–September temperature and AOD (Crowley and Unterman 2013) series showed that the reconstructed August–September temperature was significantly correlated (r = −0.60, edf = 5, p < 0.1) with the 10-yr moving average AOD series during 1792–1900, while insignificantly correlated during 1901–2020.
Comparisons of our reconstruction with four volcanic eruption chronologies. The red lines result from 10-yr loess smoothing.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
To understand the mechanism through which large volcanic eruptions influence temperature in our study area, the heat flux and radiant flux in the first and second year after large volcanic eruptions were calculated using composite analysis. The results of SEA show that the postvolcanic August–September cooling is strongest in the second year after an eruption. The variation of the temperature after large volcanic eruptions is consistent with the variation of the heat fluxes and radiation fluxes (Fig. 12). During the period 1836–2015, the composite analysis results showed that the heat fluxes and radiation fluxes in the first year after large volcanic eruptions were slightly higher than average, whereas in the second year after large volcanic eruptions, the latent heat flux and sensible heat flux were 0–5 W m−2 lower than average in the southern TP; meanwhile the upwelling longwave radiation was 5–10 W m−2 lower than average, and downward solar radiation was 10–15 W m−2 lower than average (Figs. 12a–d). The composite analysis results of the instrument period (1959–2020) are consistent with the period 1836–2015, and the decrease of heat flux and radiation flux after the eruption is more prominent (Figs. A5 and A6). This indicates that in the late summer of the second year after large volcanic eruptions, the solar shortwave radiation reaching the surface of the southeastern TP is significantly reduced, while the latent heat flux and the sensible heat flux are both reduced, thus possibly leading to an extreme low temperature.
Anomalous sensible heat flux (shtfl), latent heat flux (lhtfl), upward longwave radiation flux (ulwrf), and downward shortwave radiation flux (dswrf) field in 20CRv3 dataset after volcanic eruptions with an ice-core volcanic index (IVI) > 10 when using the volcanic eruption data from Gao et al. (2008) during the period 1836–2015. The green triangle indicates the study site.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
4. Discussion
a. The relationship between MXD and climate
The MXD chronology correlated significantly with the August–September mean temperature in our study. The information seems physiologically meaningful, as secondary cell wall formation and lignification in the latewood of conifers is reportedly controlled by late summer temperature in many sites in the TP (Duan et al. 2010; Li et al. 2018; Li et al. 2015; Xing et al. 2014). The latewood density is directly affected by cell wall dimensions (Björklund et al. 2017), which are closely related to late summer temperatures (Franceschini et al. 2013; Moser et al. 2010); that is, warm summers/late summers can promote the formation of thicker cell walls and vice versa. In China, especially in the southeast TP, tree-ring densities have been used to reconstruct the warm season temperature or late summer temperature (Bräuning and Mantwill 2004; Duan and Zhang 2014; Fan et al. 2009; Li et al. 2018; Liang et al. 2016; Wang et al. 2009; Yin et al. 2021). Li et al. (2018) found that mean latewood density (LWD) of Smith fir is significantly correlated with August–September temperature, which is consistent with our study, but the proxy used was LWD and not MXD. Our MXD shows higher correlation coefficients with the August–September minimum temperature than the maximum temperature, which was also found in a previous study conducted in the neighboring region (Liang et al. 2016). This is probably because the Tmin is more indicative of nighttime temperature, and both xylem cell thickening and lignification occur at night (Hosoo et al. 2002). The negative correlation between MXD chronology and precipitation in September for the first-difference data is because with more precipitation the temperature is lower. The correlation coefficient between MXD chronology and precipitation is weak in all months, which was also found in a few other studies of the neighboring region (Duan and Zhang 2014), because our sampling sites are located at high altitude and there is adequate moisture for tree growth in the study area.
b. Comparison with other temperature reconstructions
To further evaluate the temperature variation history in our reconstructed series, we compared our reconstructed series with two MXD- and one LWD-based temperature reconstruction for the eastern TP (TP1; TP2) (Duan et al. 2019; Xing et al. 2014), and Sygera Mountain in southeastern TP (TP3) (Li et al. 2018), and two Northern Hemisphere temperature series, one from NH land CRU (NH1) (https://crudata.uea.ac.uk/cru/data/temperature) and one reconstructed by Mann et al. (2009) (NH2) (Fig. 13). On the interannual scale, the correlation coefficients between our August–September temperature reconstruction and TP1, TP2, TP3, and NH1 were 0.68, 0.59, 0.58, and 0.69, respectively. On the decadal scale, our reconstructed temperature significantly correlated with NH1 and NH2 with coefficients of 0.92 and 0.7, and several warm and cold periods of our reconstructed series corresponded well with five other temperature series (Fig. 13). In addition, the common modern warming in the past few decades (1980–2020) existed in our reconstruction, TP1, NH1, and NH2. However, there were differences between our reconstruction with other temperature series on the interannual and decadal scale; our reconstruction showed more pronounced decadal variability in the nineteenth century than other reconstructions in the southeast TP, and with a more pronounced warming trend in the mid-twentieth century. These discrepancies were induced by the use of different proxies, local climatic variation, and different detrending techniques applied to the tree-ring data. In addition, our reconstruction is more consistent with Northern Hemisphere temperature variability on a decadal time scale. It is worth noting that since our sampling point is more westward, our reconstructed series are farther west in the region of significant spatial correlation compared to previous MXD-based temperature reconstruction studies.
Comparisons of the August–September temperature reconstruction with temperature series in the Tibetan Plateau (TP) and two Northern Hemisphere (NH) temperature series. (a) This reconstruction; (b) NH land surface temperature from CRU (https://crudata.uea.ac.uk/cru/data/temperature/); (c) NH temperature reconstructed using different proxy records (Mann et al. 2009); (d) April–September mean temperature reconstructed using tree-ring MXD network in the southeastern TP (Duan et al. 2019); (e) April–September mean temperature reconstructed using tree-ring MXD series in the eastern TP (Xing et al. 2014); (f) April–September mean temperature reconstructed using tree-ring LWD series in the Sygera Mountains (Li et al. 2018).
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
c. Climate warming
A significant warming trend since the 1960s was found in our reconstruction, with a trend of 0.21°C decade−1 (p < 0.01). The rapid warming during this period was also found in the instrumental data (Li et al. 2010; Zhou and Zhang 2021). Both the observed and reconstructed series show rapid warming during the period 1960–2000, while a warming hiatus existed during 2001–12, then another rapid warming period followed. In our August–September temperature reconstruction, the 20-yr average August–September temperature for 2001–20 increased by 1.10°C relative to 1850–1900, while the 10-yr average August–September temperature for 2011–20 increased by about 1.23°C, slightly higher than the global land surface average (0.99°C for 2001–20 and 1.09°C for 2010–20). Temperature in each decade of the last 30 years has been successively warmer than any of the previous decades since 1792. Some temperature reconstructions based on TRW or MXD from both the southeastern and central TP also reveal warming during this period (Li et al. 2018; Liang et al. 2009; Liu et al. 2009; Yin et al. 2015). It is noteworthy that there are spatial differences in the climate warming of the TP, and there is no obvious warming trend in some regions (e.g., summer temperature around Qamdo and Dingqing) in both observed and reconstructed data (Bräuning and Mantwill 2004; Xing et al. 2014).
In addition, our reconstructed series showed no significant warming in the southern TP until the mid-twentieth century. However, some studies have shown that warming may have occurred earlier than the mid-twentieth century. Duan et al. (2020) reconstructed the late summer temperature of the southeast TP using MXD and found an increasing trend since the late nineteenth century. Yin et al. (2022) reconstructed the late summer temperature of the eastern TP using MXD and found a warming trend from 1867. Shi et al. (2015) reconstructed the summer minimum temperature, which revealed a long-term persistent warming beginning in the 1820s. Because our reconstruction used the same detrending method and reconstructed the same seasons as Duan et al. (2020) and Yin et al. (2022), the difference in the timing of the appearance of the warming may indicate regional differences in climate warming. The difference in the timing of the occurrence of warming with that of Shi et al. (2015) may be related to factors such as different climatic factors, different seasons, and different tree-ring indicators.
d. Impact of AMO and volcanic forcing
Our results demonstrate that during the warm phase of the AMO, higher pressure and divergent horizontal wind over the TP promote air mass vertical sinking, causing more sunny days and warmer summers. The stronger southerly winds also transport more heat from lower latitudes to the TP, contributing to warmer summers. This finding could improve our understanding of the mechanisms linking AMO to summer temperatures on the TP. Additionally, the simulated summer rainfall also exhibited dry anomalies on the TP during the warm phase of the AMO (Shi et al. 2019). The higher pressure and divergent horizontal wind over the TP during the warm phase of the AMO may be linked to the propagation of Rossby waves in the middle and upper troposphere over Eurasian continent (Li et al. 2008; Luo et al. 2011; Wang et al. 2015).
The influence of large volcanic eruptions on decadal temperature change in some regions may be related to the influence of large volcanic eruptions on ocean temperature and heat content (Church et al. 2005; Zanchettin et al. 2013, 2012). Historically, the negative phase of AMO has been triggered by large volcanic eruptions and lasted for several years or longer (Brönnimann et al. 2019; Mann et al. 2021). The negative phase of AMO leads to cold periods in the TP and other parts of East Asia. The correlation between large volcanic eruptions and temperature weakened in the twentieth century, which may be related to climate warming or to the lower volcanic forcing in the twentieth century.
Our research shows that, contrary to the southeastern (Duan et al. 2018) and northeastern TP (Wang et al. 2021; Wang et al. 2017; Chen et al. 2017), the southern TP experienced the lowest temperatures in the second year following large volcanic eruptions (Fig. 10). A study based on MXD data similarly found that the minimum temperature in northern Europe occurred in the second year after a large volcanic eruption, whereas in central Europe it occurred one year after the eruption (Esper et al. 2013). In addition, our results also show that near-surface downward solar radiation reaches its minimum in the second year after a large volcanic eruption in the southern TP (Fig. 12). This finding is crucial for demonstrating the cooling mechanism in the southern TP after a large volcanic eruption. Our findings offer long-term evidence on the timing and mechanisms of regional cooling following a large volcanic eruption.
5. Conclusions
In this study, we developed a 257-yr Smith fir MXD chronology based on two new sampling sites in the southern TP. The sampling sites were more westerly than those of previous studies. Based on the tree growth–climate relationship, the late summer mean temperature was reconstructed for the period 1792–2020, in which the transfer function explained 66.2% of the variance in the observed temperature for the calibration period 1960–2020. The temperature reconstruction showed a warming trend since the 1960s in the southern TP, and the temperature in each decade of the last 30 years has seen successively warmer than any of the previous decades since 1792. Higher pressure and divergent horizontal winds over the TP contributing to warmer summers in the region during the warm phase of the AMO. Additionally, the southern TP experienced the lowest temperature and downward solar radiation in the second year following large volcanic eruptions.
Acknowledgments.
This research was supported by the Second Tibetan Plateau Scientific Expedition and Research Program (Grant 2019QZKK0301), and National Natural Science Foundation of China (Grants 41977391 and 41571194).
Data availability statement.
The data that support the findings of this study have been published in Zenodo (https://zenodo.org/record/8337358).
APPENDIX
Supporting Information
This supporting information provides additional figures (Figs. A1–A6) to support the main manuscript.
Comparison of difference detrending methods.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
Multitaper method (MTM) spectral analysis result. (a) The MTM power and confidence levels. (b) The AR1 confidence level estimates result. (c) The harmonic F test confidence level estimates result.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
(a) Cross-wavelet (XWT) and (b) wavelet coherence (WTC) of reconstructed August–September temperature in southern TP with reconstructed AMO series by Mann et al. (2009).
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
(a),(b) 500- and (c),(d) 200-hPa anomalous geopotential height field (shade) and anomalous flow field (velocity) during the AMO (left) cold and (right) warm period in the ERA5 dataset.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
Anomalous sensible heat flux (shtfl), latent heat flux (lhtfl), upward long-wave radiation flux (ulwrf), and downward short-wave radiation flux (dswrf) field in 20CRv3 dataset after volcanic eruptions with an ice-core volcanic index (IVI) > 10 when using the volcanic eruption data from Gao et al. (2008) during the period 1959–2015. The green triangle indicates the study site.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
Anomalous sensible heat flux (shtfl), latent heat flux (lhtfl), upward longwave radiation flux (ulwrf), and downward shortwave radiation flux (dswrf) field in ERA5 dataset after volcanic eruptions with an ice-core volcanic index (IVI) > 10 when using the volcanic eruption data from Gao et al. (2008) during the period 1836–2015. The green triangle indicates the study site.
Citation: Journal of Climate 36, 20; 10.1175/JCLI-D-22-0624.1
REFERENCES
An, W., S. Hou, W. Zhang, S. Wu, H. Xu, H. Pang, Y. Wang, and Y. Liu, 2016: Possible recent warming hiatus on the northwestern Tibetan Plateau derived from ice core records. Sci. Rep., 6, 32813, https://doi.org/10.1038/srep32813.
Björklund, J., and Coauthors, 2017: Cell size and wall dimensions drive distinct variability of earlywood and latewood density in Northern Hemisphere conifers. New Phytol., 216, 728–740, https://doi.org/10.1111/nph.14639.
Bräuning, A., and B. Mantwill, 2004: Summer temperature and summer monsoon history on the Tibetan Plateau during the last 400 years recorded by tree rings. Geophys. Res. Lett., 31, L24205, https://doi.org/10.1029/2004GL020793.
Briffa, K. R., T. J. Osborn, F. H. Schweingruber, P. D. Jones, S. G. Shiyatov, and E. A. Vaganov, 2002: Tree-ring width and density data around the Northern Hemisphere: Part 1, local and regional climate signals. Holocene, 12, 737–757, https://doi.org/10.1191/0959683602hl587rp.
Brönnimann, S., and Coauthors, 2019: Last phase of the Little Ice Age forced by volcanic eruptions. Nat. Geosci., 12, 650–656, https://doi.org/10.1038/s41561-019-0402-y.
Chen, F., H. Wang, and Y. Yuan, 2017: Two centuries of temperature variation and volcanic forcing reconstructed for the northern Tibetan Plateau. Phys. Geogr., 38, 248–262, https://doi.org/10.1080/02723646.2017.1293484.
Church, J. A., N. J. White, and J. M. Arblaster, 2005: Significant decadal-scale impact of volcanic eruptions on sea level and ocean heat content. Nature, 438, 74–77, https://doi.org/10.1038/nature04237.
Cook, E. R., and L. A. Kairiukstis, 1990: Methods of Dendrochronology: Applications in the Environmental Sciences. Kluwer Academic, 394 pp.
Cook, E. R., K. R. Briffa, and P. D. Jones, 1994: Spatial regression methods in dendroclimatology: A review and comparison of two techniques. Int. J. Climatol., 14, 379–402, https://doi.org/10.1002/joc.3370140404.
Crowley, T. J., and M. B. Unterman, 2013: Technical details concerning development of a 1200 yr proxy index for global volcanism. Earth Syst. Sci. Data, 5, 187–197, https://doi.org/10.5194/essd-5-187-2013.
D’Arrigo, R. D., and G. C. Jacoby, 1999: Northern North American tree-ring evidence for regional temperature changes after major volcanic events. Climatic Change, 41 (1), 1–15, https://doi.org/10.1023/A:1005370210796.
Dima, M., and G. Lohmann, 2007: A hemispheric mechanism for the Atlantic multidecadal oscillation. J. Climate, 20, 2706–2719, https://doi.org/10.1175/JCLI4174.1.
Dong, X., G. Zeng, G. Zhang, and X. Yang, 2023: Current AMO mitigating extreme high temperatures in Central Asia under global warming. Int. J. Climatol., 43, 3947–3962, https://doi.org/10.1002/joc.8066.
Duan, J., and Q. Zhang, 2014: A 449 year warm season temperature reconstruction in the southeastern Tibetan Plateau and its relation to solar activity. J. Geophys. Res. Atmos., 119, 11 578–11 592, https://doi.org/10.1002/2014JD022422.
Duan, J., L. Wang, L. Li, and K. L. Chen, 2010: Temperature variability since A.D. 1837 inferred from tree-ring maximum density of Abies fabri in Gongga Mountains, China. Chin. Sci. Bull., 55, 3015–3022, https://doi.org/10.1007/s11434-010-3182-8.
Duan, J., L. Li, and Y. Fang, 2015: Seasonal spatial heterogeneity of warming rates on the Tibetan Plateau over the past 30 years. Sci. Rep., 5, 11725, https://doi.org/10.1038/srep11725.
Duan, J., and Coauthors, 2018: Summer cooling driven by large volcanic eruptions over the Tibetan Plateau. J. Climate, 31, 9869–9879, https://doi.org/10.1175/JCLI-D-17-0664.1.
Duan, J., Z. Ma, L. Li, and Z. Zheng, 2019: August–September temperature variability on the Tibetan Plateau: Past, present, and future. J. Geophys. Res. Atmos., 124, 6057–6068, https://doi.org/10.1029/2019JD030444.
Duan, J., L. Li, Z. Ma, and L. Chen, 2020: Post‐industrial late summer warming recorded in tree‐ring density in the eastern Tibetan Plateau. Int. J. Climatol., 40, 795–804, https://doi.org/10.1002/joc.6239.
Esper, J., and Coauthors, 2012: Orbital forcing of tree-ring data. Nat. Climate Change, 2, 862–866, https://doi.org/10.1038/nclimate1589.
Esper, J., L. Schneider, P. J. Krusic, J. Luterbacher, U. Büntgen, M. Timonen, F. Sirocko, and E. Zorita, 2013: European summer temperature response to annually dated volcanic eruptions over the past nine centuries. Bull. Volcanol., 75, 736, https://doi.org/10.1007/s00445-013-0736-z.
Fan, Z.-X., A. Bräuning, B. Yang, and K.-F. Cao, 2009: Tree ring density-based summer temperature reconstruction for the central Hengduan Mountains in southern China. Global Planet. Change, 65 (1–2), 1–11, https://doi.org/10.1016/j.gloplacha.2008.10.001.
Fang, K., X. Gou, F. Chen, C. Liu, N. Davi, J. Li, Z. Zhao, and Y. Li, 2012: Tree-ring based reconstruction of drought variability (1615–2009) in the Kongtong Mountain area, northern China. Global Planet. Change, 80–81, 190–197, https://doi.org/10.1016/j.gloplacha.2011.10.009.
Feng, S., and Q. Hu, 2008: How the North Atlantic multidecadal oscillation may have influenced the Indian summer monsoon during the past two millennia. Geophys. Res. Lett., 35, L01707, https://doi.org/10.1029/2007GL032484.
Franceschini, T., J.-D. Bontemps, V. Perez, and J.-M. Leban, 2013: Divergence in latewood density response of Norway spruce to temperature is not resolved by enlarged sets of climatic predictors and their non-linearities. Agric. For. Meteor., 180, 132–141, https://doi.org/10.1016/j.agrformet.2013.05.011.
Frank, D., and J. Esper, 2005: Temperature reconstructions and comparisons with instrumental data from a tree-ring network for the European Alps. Int. J. Climatol., 25, 1437–1454, https://doi.org/10.1002/joc.1210.
Fritts, H. C., 1976: Tree Rings and Climate. Academic Press, 567 pp.
Gao, C., A. Robock, and C. Ammann, 2008: Volcanic forcing of climate over the past 1500 years: An improved ice core-based index for climate models. J. Geophys. Res., 113, D23111, https://doi.org/10.1029/2008JD010239.
Gao, C., F. Ludlow, O. Amir, and C. Kostick, 2016: Reconciling multiple ice-core volcanic histories: The potential of tree-ring and documentary evidence, 670–730 CE. Quat. Int., 394, 180–193, https://doi.org/10.1016/j.quaint.2015.11.098.
Gao, C., Y. Gao, Q. Zhang, and C. Shi, 2017: Climatic aftermath of the 1815 Tambora eruption in China. J. Meteor. Res., 31, 28–38, https://doi.org/10.1007/s13351-017-6091-9.
Hao, Z., D. Sun, X. Z. Zhang, and J. Y. Zheng, 2016: Regional differences in temperature response in China to the large volcanic eruptions since the 20th century. Prog. Geogr., 35, 331–338, https://doi.org/10.18306/dlkxjz.2016.03.007.
Holmes, R. L., 1983: Computer-assisted quality control in tree-ring dating and measurement. Tree-Ring Bull., 43, 51–67.
Hosoo, Y., M. Yoshida, T. Imai, and T. Okuyama, 2002: Diurnal difference in the amount of immunogold-labeled glucomannans detected with field emission scanning electron microscopy at the innermost surface of developing secondary walls of differentiating conifer tracheids. Planta, 215, 1006–1012, https://doi.org/10.1007/s00425-002-0824-3.
Hu, S., T. Zhou, and B. Wu, 2021: Impact of developing ENSO on Tibetan Plateau summer rainfall. J. Climate, 34, 3385–3400, https://doi.org/10.1175/JCLI-D-20-0612.1.
Hu, S., B. Wu, T. Zhou, and Y. Yu, 2022: Dominant anomalous circulation patterns of Tibetan Plateau summer climate generated by ENSO-forced and ENSO-independent teleconnections. J. Climate, 35, 1679–1694, https://doi.org/10.1175/JCLI-D-21-0207.1.
Immerzeel, W. W., L. P. H. Van Beek, and M. F. Bierkens, 2010: Climate change will affect the Asian water towers. Science, 328, 1382–1385, https://doi.org/10.1126/science.1183188.
Li, L., S. Yang, Z. Wang, X. Zhu, and H. Tang, 2010: Evidence of warming and wetting climate over the Qinghai-Tibet Plateau. Arct. Antarct. Alp. Res., 42, 449–457, https://doi.org/10.1657/1938-4246-42.4.449.
Li, M., and X. Shao, 2016: Study on the relationship between large volcanic eruptions and temperature variation based on tree-ring data in the eastern Tibetan Plateau during the past millennium. Adv. Earth Sci., 31, 634–642, https://doi.org/10.11867/j.issn.1001-8166.2016.06.0634.
Li, M., L. Huang, Z.-Y. Yin, and X. M. Shao, 2017: Temperature reconstruction and volcanic eruption signal from tree-ring width and maximum latewood density over the past 304 years in the southeastern Tibetan Plateau. Int. J. Biometeor., 61, 2021–2032, https://doi.org/10.1007/s00484-017-1395-0.
Li, M., J. Duan, L. Wang, and H. Zhu, 2018: Late summer temperature reconstruction based on tree-ring density for Sygera Mountain, southeastern Tibetan Plateau. Global Planet. Change, 163, 10–17, https://doi.org/10.1016/j.gloplacha.2018.02.005.
Li, M.-Y., L. Wang, Z.-X. Fan, and C.-C. Shen, 2015: Tree-ring density inferred late summer temperature variability over the past three centuries in the Gaoligong Mountains, southeastern Tibetan Plateau. Palaeogeogr. Palaeoclimatol. Palaeoecol., 422, 57–64, https://doi.org/10.1016/j.palaeo.2015.01.003.
Li, S., J. Perlwitz, X. Quan, and M. P. Hoerling, 2008: Modelling the influence of North Atlantic multidecadal warmth on the Indian summer rainfall. Geophys. Res. Lett., 35, L05804, https://doi.org/10.1029/2007GL032901.
Liang, E., X. Shao, and N. Qin, 2008: Tree-ring based summer temperature reconstruction for the source region of the Yangtze River on the Tibetan Plateau. Global Planet. Change, 61, 313–320, https://doi.org/10.1016/j.gloplacha.2007.10.008.
Liang, E., X. Shao, and Y. Xu, 2009: Tree-ring evidence of recent abnormal warming on the Southeast Tibetan Plateau. Theor. Appl. Climatol., 98, 9–18, https://doi.org/10.1007/s00704-008-0085-6.
Liang, H., L. Lyu, and M. Wahab, 2016: A 382-year reconstruction of August mean minimum temperature from tree-ring maximum latewood density on the southeastern Tibetan Plateau, China. Dendrochronologia, 37, 1–8, https://doi.org/10.1016/j.dendro.2015.11.001.
Libiseller, C., and A. Grimvall, 2002: Performance of partial Mann–Kendall tests for trend detection in the presence of covariates. Environmetrics, 13, 71–84, https://doi.org/10.1002/env.507.
Liu, Y., Z. An, H. W. Linderholm, D. Chen, H. Song, Q. Cai, J. Sun, and H. Tian, 2009: Annual temperatures during the last 2485 years in the mid-eastern Tibetan Plateau inferred from tree rings. Sci. China, 52D, 348–359, https://doi.org/10.1007/s11430-009-0025-z.
Luo, F., S. Li, and T. Furevik, 2011: The connection between the Atlantic multidecadal oscillation and the Indian summer monsoon in Bergen climate model version 2.0. J. Geophys. Res., 116, D19117, https://doi.org/10.1029/2011JD015848.
Mann, M. E., and Coauthors, 2009: Global signatures and dynamical origins of the little ice age and medieval climate anomaly. Science, 326, 1256–1260, https://doi.org/10.1126/science.1177303.
Mann, M. E., B. A. Steinman, D. J. Brouillette, and S. K. Miller, 2021: Multidecadal climate oscillations during the past millennium driven by volcanic forcing. Science, 371, 1014–1019, https://doi.org/10.1126/science.abc5810.
Melvin, T. M., and K. R. Briffa, 2008: A “signal-free” approach to dendroclimatic standardisation. Dendrochronologia, 26, 71–86, https://doi.org/10.1016/j.dendro.2007.12.001.
Melvin, T. M., and K. R. Briffa, 2014: CRUST: Software for the implementation of regional chronology standardisation: Part 1. Signal-free RCS. Dendrochronologia, 32, 7–20, https://doi.org/10.1016/j.dendro.2013.06.002.
Meyers, S. R., 2012: Seeing red in cyclic stratigraphy: Spectral noise estimation for astrochronology. Paleoceanography, 27, PA3228, https://doi.org/10.1029/2012PA002307.
Michaelsen, J., 1987: Cross-validation in statistical climate forecast models. J. Climate Appl. Meteor., 26, 1589–1600, https://doi.org/10.1175/1520-0450(1987)026<1589:CVISCF>2.0.CO;2.
Moser, L., P. Fonti, U. Büntgen, J. Esper, J. Luterbacher, J. Franzen, and D. Frank, 2010: Timing and duration of European larch growing season along altitudinal gradients in the Swiss Alps. Tree Physiol., 30, 225–233, https://doi.org/10.1093/treephys/tpp108.
Rangwala, I., J. R. Miller, G. L. Russell, and M. Xu, 2010: Using a global climate model to evaluate the influences of water vapor, snow cover and atmospheric aerosol on warming in the Tibetan Plateau during the twenty-first century. Climate Dyn., 34, 859–872, https://doi.org/10.1007/s00382-009-0564-1.
Ratna, S. B., T. J. Osborn, M. Joshi, and J. Luterbacher, 2020: The influence of Atlantic variability on Asian summer climate is sensitive to the pattern of the sea surface temperature anomaly. J. Climate, 33, 7567–7590, https://doi.org/10.1175/JCLI-D-20-0039.1.
Robock, A., 2000: Volcanic eruptions and climate. Rev. Geophys., 38, 191–219, https://doi.org/10.1029/1998RG000054.
Schiemann, R., D. Lüthi, and C. Schär, 2009: Seasonality and interannual variability of the westerly jet in the Tibetan Plateau region. J. Climate, 22, 2940–2957, https://doi.org/10.1175/2008JCLI2625.1.
Schweingruber, F. H., T. Bartholin, E. Schaur, and K. R. Briffa, 1988: Radiodensitometric‐dendroclimatological conifer chronologies from Lapland (Scandinavia) and the Alps (Switzerland). Boreas, 17, 559–566, https://doi.org/10.1111/j.1502-3885.1988.tb00569.x.
Shi, C., V. Masson-Delmotte, V. Daux, Z. Li, M. Carré, and J. C. Moore, 2015: Unprecedented recent warming rate and temperature variability over the east Tibetan Plateau inferred from alpine treeline dendrochronology. Climate Dyn., 45, 1367–1380, https://doi.org/10.1007/s00382-014-2386-z.
Shi, C., and Coauthors, 2019: Summer temperature over the Tibetan Plateau modulated by Atlantic multidecadal variability. J. Climate, 32, 4055–4067, https://doi.org/10.1175/JCLI-D-17-0858.1.
Sigl, M., and Coauthors, 2015: Timing and climate forcing of volcanic eruptions for the past 2,500 years. Nature, 523, 543–549, https://doi.org/10.1038/nature14565.
Thakur, G., E. Brevdo, N. S. Fučkar, and H.-T. Wu, 2013: The synchrosqueezing algorithm for time-varying spectral analysis: Robustness properties and new paleoclimate applications. Signal Process., 93, 1079–1094, https://doi.org/10.1016/j.sigpro.2012.11.029.
Wang, B., T. Chen, G. Xu, G. Wu, and C. Li, 2017: Reconstructed annual mean temperatures for the northeastern margin of the Tibetan Plateau: Associations with the East Asian monsoons and volcanic events. Int. J. Climatol., 37, 3044–3056, https://doi.org/10.1002/joc.4900.
Wang, J., B. Yang, C. Qin, S. Kang, M. He, and Z. Wang, 2014: Tree-ring inferred annual mean temperature variations on the southeastern Tibetan Plateau during the last millennium and their relationships with the Atlantic multidecadal oscillation. Climate Dyn., 43, 627–640, https://doi.org/10.1007/s00382-013-1802-0.
Wang, J., B. Yang, and F. C. Ljungqvist, 2015: A millennial summer temperature reconstruction for the eastern Tibetan Plateau from tree-ring width. J. Climate, 28, 5289–5304, https://doi.org/10.1175/JCLI-D-14-00738.1.
Wang, L., J. Duan, J. Chen, L. Huang, and X. Shao, 2009: Temperature reconstruction from tree-ring maximum density of Balfour spruce in eastern Tibet, China. Int. J. Climatol., 30, 972–979, https://doi.org/10.1002/joc.2000.
Wang, Y., X. Shao, Y. Zhang, and M. Li, 2021: The response of annual minimum temperature on the eastern central Tibetan Plateau to large volcanic eruptions for the period 1380–2014 CE. Climate Past, 17, 241–252, https://doi.org/10.5194/cp-17-241-2021.
Wigley, T. M. L., K. R. Briffa, and P. D. Jones, 1984: On the average value of correlated time series, with applications in dendroclimatology and hydrometeorology. J. Climate Appl. Meteor., 23, 201–213, https://doi.org/10.1175/1520-0450(1984)023<0201:OTAVOC>2.0.CO;2.
Wyatt, M. G., S. Kravtsov, and A. A. Tsonis, 2012: Atlantic multidecadal oscillation and Northern Hemisphere’s climate variability. Climate Dyn., 38, 929–949, https://doi.org/10.1007/s00382-011-1071-8.
Xing, P., Q.-B. Zhang, and L.-X. Lv, 2014: Absence of late-summer warming trend over the past two and half centuries on the eastern Tibetan Plateau. Global Planet. Change, 123, 27–35, https://doi.org/10.1016/j.gloplacha.2014.10.006.
Yang, X., G. Zeng, G. Zhang, J. Li, Z. Li, and Z. Hao, 2021: Interdecadal variations of different types of summer heat waves in northeast China associated with AMO and PDO. J. Climate, 34, 7783–7797, https://doi.org/10.1175/JCLI-D-20-0939.1.
Yin, H., H. Liu, H. W. Linderholm, and Y. Sun, 2015: Tree ring density-based warm-season temperature reconstruction since A.D. 1610 in the eastern Tibetan Plateau. Palaeogeogr. Palaeoclimatol. Palaeoecol., 426, 112–120, https://doi.org/10.1016/j.palaeo.2015.03.003.
Yin, H., M.-Y. Li, and L. Huang, 2021: Summer mean temperature reconstruction based on tree-ring density over the past 440 years on the eastern Tibetan Plateau. Quat. Int., 571, 81–88, https://doi.org/10.1016/j.quaint.2020.09.018.
Yin, H., Y. Sun, and M.-Y. Li, 2022: Reconstructed temperature change in late summer over the eastern Tibetan Plateau since 1867 CE and the role of anthropogenic forcing. Global Planet. Change, 208, 103715, https://doi.org/10.1016/j.gloplacha.2021.103715.
Yin, Z.-Y., Z. Lin, and X. Zhao, 2000: Temperature anomalies in central and eastern Tibetan Plateau in relation to general circulation patterns during 1951–1993. Int. J. Climatol., 20, 1431–1449, https://doi.org/10.1002/1097-0088(200010)20:12%3C1431::AID-JOC551%3E3.0.CO;2-J.
Zanchettin, D., C. Timmreck, H.-F. Graf, A. Rubino, S. Lorenz, K. Lohmann, K. Krüger, and J. H. Jungclaus, 2012: Bi-decadal variability excited in the coupled ocean–atmosphere system by strong tropical volcanic eruptions. Climate Dyn., 39, 419–444, https://doi.org/10.1007/s00382-011-1167-1.
Zanchettin, D., O. Bothe, H. F. Graf, S. J. Lorenz, J. Luterbacher, C. Timmreck, and J. H. Jungclaus, 2013: Background conditions influence the decadal climate response to strong volcanic eruptions. J. Geophys. Res. Atmos., 118, 4090–4106, https://doi.org/10.1002/jgrd.50229.