ИСТИНА

Tree-ring collections from the northern part of European Russia were mostly collected in the 1970s and 1980s and were never updated since that time. In this study we present the results of the spatio-temporal analysis of the climatic signal of spruce (Picea abies (L.) Karst and Picea obovata Lebed.) growing in the Solovki Islands in different ecological conditions. Solovki is one of the most promising region from the point of view of dendrochronology: the archipelago is located in the vicinity of the northern tree limit, is reach in the spruce of old age and wooden architectural monuments dating back to 12th century when the Solovetsky monastery was established. As a result more than 140 tree-ring width series were successfully cross-dated and used for developing 12 local chronologies. All local chronologies were compared with temperature and precipitation records. All local chronologies showed statistically significant common sensitivity to June air temperature variation. Pearson’s coefficients of correlation vary from 0.26 to 0.56 (p<0.05). Final composite spruce tree-ring width chronology covers the period 1626-2012 ss. and consists of 134 samples. EPS value exceeds the threshold value 0.85 after 1676 when at least 6 samples are present in the chronology. The bootstrapped response climate analysis revealed a positive relation of spruce growth to June-July air temperatures of the current year and a negative one to temperature in February. There is also a positive response of spruce growth to July precipitation of the current year. June air temperature is the most distinct climatic parameter controlling spruce growth which is reflected in both static and moving response function analysis. Computed moving correlation between the spruce chronology and the climatic factors shows that only the correlation with June temperature is stable over time. The relation is weaker in the beginning of the instrumental records (R= 0.3 – 0.4, p<0.05) but it is getting stronger since 1920s (R = 0.5-0.6, p < 0.05). July air temperature signal disappears in the beginning and in the end of the instrumental period. Statistically significant and temporally stable relation between the composite ring width spruce chronology and June air temperature allowed to reconstruct this parameter back in time. For the reconstruction the composite chronology was scaled against June air temperatures. Comparison between predicted and actual June temperatures revealed a good agreement over the period 1901-2012. Positive values of RE and CE statistics indicate predictive skills of the applied model. This finding allowed to reconstruct the June air temperature using spruce tree-ring width chronology since 1676. According to smoothed by 30-year spline reconstruction the cold anomalies date back to 1676–1680, 1761-1823,1836-1899, 1935-1952, and 1960–1979, and the warmings occurred in 1681-1760, 1824-1835, 1900-1934, 1953-1959, and 1980–2012. The coolest reconstructed June air temperature occurred in 1836 and was 2.9 ⁰C cooler than the reference period (1901- 2012). Other strong cold anomalies in June were identified in 1976 (−2.4 ⁰C), 1982 (−2.4⁰C), 1820 (−2.4 ⁰C), 1790 (−2.3 ⁰C), 1817 (−2.2 ⁰C), 1879 (−2.1 ⁰C), and 1810 (−2.1 ⁰C).According to the reconstruction the warmest June temperature occurred in 1685 (+3.0 ⁰C).The comparison of June temperature reconstruction with the dates of major climatically effective explosive volcanic eruptions (Sigl et al. 2015) showed that many reconstructed coldest years, such as 1836, 1982, 1817, 1810, 1661 and 1674 CE occurred after major tropical volcanic eruptions (Briffa et al. 1998; Sigl et al. 2015) or followed within several years (1-2 years). The coldest June temperature anomaly in the reconstruction occurred in 1836 CE after the eruption of Cosiguina (Nicaragua) in June 1835. This research is funded by the Russian Scientific Foundation № 17-77-20123.