Study On The Response Of Tree Ring Growth To Climate Changes In Scandinavia

In the background of global climate change, knowledge of past climate variability and large-scale atmospheric circulation helps to accurately assess the current climate status in a historical context and the human impacts on climate change. In addition, it also provides paleoclimatic data to evaluate the ability of climate models to simulate realistic climate changes. Understanding temporal and spatial characteristics of climate in Scandinavia and its relationship with summer North Atlantic Oscillation is helpful for us to get the knowledge of past climate variability and large-scale atmospheric circulation in high-latitude region. In this thesis, tree ring width proxies of Scots pine were used to investigate the response of tree ring growth to temperature and precipitation along west-east transect in west-central Scandinavia, and used to reconstruct the average summer temperature of Storlien that enriched the temperature reconstruction networks based on tree ring width proxies in Scandinavia. The tree ring width chronology networks including multiple tree species such as Scots pine, Norway spruce and European birch in Scandinavia were used to investigate the temporal and spatial distribution of tree ring growth mode in this region and it relationship with summer North Alantic Oscillation.Five tree ring width series derived from five sites along west-east transect in west-central Scandinavia were used to do the response analysis against temperature and precipitation data collected from five meteorological station along the transect. The results showed that the Atlantic Ocean had a big impact on climate in west-central Scandinavia, while temperature and precipitation along west-east transect showed a significant gradient distribution. Tree ring width series in these five sites showed high series intercorrelations, high average mean sensitivities, and high value on first order autocorrelations which emphasized the effect of precedent year’s climate on the current year’s tree ring growth. The statistical characteristics of these tree ring width series such as average ring widths, first order autocorrelations, and series intercorrelations showed a gradually increasing trend along the west-east direction which underlined the Atlantic Ocean effect on tree ring growth characteristics. Growing season temperature was a primary limiting factor of tree ring growth in this area. Tree ring width series in these five sites could well reflect local July average temperature variation. Tree ring width indexes and July average temperature showed a stronger response relationship in inland area compare with that in oceanic area. In the past250years, tree ring growth variation in the four inland sites showed a high homogeneity, while tree ring growth variation in the most oceanic site of Nonshaugen showed a significant changing correlation with tree ring growth in other sites along the time. Significant decreasing correlations between five tree ring width chronologies and July average temperature appeared in the second half20th century comparing with those in the first half20th century. At the same time, the area of tree ring chronology-July average temperature spatial correlation extent decreased, and correlation center had moved from north to south.Tree ring width proxies of Scots pine were used to reconstruct the average summer temperature of Storlien. The temperature reconstructions based on ARSTAN chronology and residual chronology separately explained30%and28%variation of Storlien average summer temperature. The reconstructions well correlated with Jamtland Maximum latewood density (MXD) warm season temperature reconstructions before20th century. In20th century, the Jamtland MXD reconstructions showed a distinct increase of warm season temperature which didn’t appear in the Storlien reconstruction. This was probably due to the adjustment of Atlantic Ocean on climate which induced less distinct increase of tree ring growth in Storlien than of that in inland sites. The results showed that the inland reconstruction contained more periodic signals including tow long cycles than that of Storlien reconstruction which primarily contained two cycles of3-5years and11-12years. We inferred that the stable climate in Storlien caused by the ocean impact leaded to less but stronger periodic signals in tree ring series in Storlien than that in inland sites.Based on multiple-species tree ring width chronology networks, rotated principal component analysis was used to divide tree ring growth mode into tow types in Scandinavia. One type contained the tree ring series from the sites mainly located in north-west area, the other type contained the series from the sites mainly locate in south-east area. The results showed that in the period of AD1860-AD1978, the tree ring growth modes were clustered according to the approximation degree of the locations of sampling sites. The factor of tree species didn’t influence the cluster results which indicated climate factors were the most significant factors on determining the tree ring growth modes in Scandinavia. Temperature and precipitation were separately the primary limiting factors of tree ring growth in north-west mode and south-east mode. Tree ring growth modes and average mean sensitivity in Scandinavia changed along the time. We inferred that summer North Atlantic Oscillation and the20th century global warming had a big influence on tree ring growth mode in Scandinavia. The classification of tree ring growth modes could reflect changes of climate and large-scale atmospheric circulation. The west-central Scandinavia was the most sensitive area of Scandinavia concerning the response of tree ring growth modes variations to climate changes.