Erika Knutsson

An MXD Chronology of picea abies from Jämtland - A millennial long summer temperature reconstruction

Date, time and venue: 19th of February, at 16:05, in U37
Supervisor: Björn Gunnarsson
Examiner: Steffen Holzkämper
Master's Programme in Quaternary Science and Climate Development, 120 credits
Degree Project in Physical Geography and Quaternary Geology, 45 credits

Presentation language: English


Paleoclimatology is an important tool to understand the natural variations of past climate. An understanding of past conditions can help us make projections of future variations, and understanding the effects of past climate fluctuations can give us means to better adjust to the changes that will occur in the future. Past climate can be reconstructed by analysis of natural archives that has recorded what climate was like during their lifetime. One of the strongest climate proxies are tree rings, which due to its high continuity and precision as well as replicability can be used for high-resolution paleoclimatological reconstructions.

Here an MXD (Maximum Latewood Density) chronology attained from Norway Spruce from Jämtland in north-western Fennoscandia is presented. Past dendrochronological studies in the area have mainly focused on Scots pine, of which both a TRW (Tree Ring Width) and an MXD chronology have been made. The first study on Norway spruce was executed in 2014 and resulted in a TRW chronology (Rocha, 2014). The work presented in this thesis, producing an MXD chronology of Norway spruce, will therefore fill in a gap, as no previous MXD spruce chronology exists at the location, and contribute to the understanding of regional climate variability. The produced chronology consists of material from living and subfossil trees ranging from 926 A.D. to 2008 A.D.

The MXD chronology provide a strong climate signal for recorded summer temperatures and has a correlation with meteorological data of 72.4% for May-August. Through linear regression a temperature reconstruction model was made, which explains for 52.5% of the variance. The 5 coldest reconstructed years were 1207, 1254, 1332, 1899, 1902 and the 5 warmest years were 1251, 1255, 1305, 1306, and 1311. Noteworthy is that the reconstruction model could not fully recreate the warm temperatures that was registered by the meteorological data during the late 1900s and early 2000s. This is a phenomenon that have been seen in other dendrochronological studies as well and has been accredited to increased stress during the dormant season, affecting tree growth.

The spruce chronology was compared to a pine MXD chronology from the same area, which was produced by Gunnarson et al. (2011). A significant covariance in the climate signal from the two species was found (over 0.7 during the last centuries). The two chronologies were combined trough averaging of the MXD data. The strongest climate signal of the combined chronology was achieved for the season April-September, where the explained variance was 53.2% - surpassing both the pine and the spruce chronology.