Microsoft Word - 2. Kern-Popa-jav-2010-11-30-vegleges.doc Journal of Env. Geogr. Vol. II. No. 3-4. (2009) pp. 15-22 ASSESSING TEMPERATURE SIGNAL IN X-RAY DENSITOMETRIC DATA OF NORWAY SPRUCE AND THE EARLIEST INSTRUMENTAL RECORD FROM THE SOUTHERN CARPATHIANS Kern, Z. 1* - Popa, I. 2 1: Institute for Geochemical Research, Hungarian Academy of Sciences, H-1112 Budaörsi út 45, Budapest, Hungary 2: Research Station for Norway Spruce Silviculture, Forest Research and Management Institute, Câmpulung Moldovenesc, Romania *correspondence: E-mail: kern@geochem.hu, Tel: 00 36 1 3092600/1106, Fax: 00 36 1 3193137 Abstract Radiodensity data derived from Norway spruce were studied from a southern Carpathian site. Maximum density record showed significant positive relationship (r=0.59) with the growing season (April-September) air temperature and mini- mum density (MND) record showed clear and significant negative response (r=-0.41) to June-July mean air temperature. This significant MND response to climate is a novel result as traditionally this densitometric parameter was regarded not to carry any meaningful climatic signal. Derived temperature sensitive proxy records were compared to instrumental data of Sibiu (Nagyszeben/Hermannstadt) the oldest available regional station. Results of the running window correlation analysis pointed out notable inhomogeneities in the instrumental data before 1906. The Sibiu temperature series should be subjected to scrutiny revision to clean it from inhomogeneities. Key words: maximum latewood density, minimum earlywood density, Picea abies, inhomogeneous instrumental temperature, dendroclimatology, Romania INTRODUCTION Norway spruce (Picea abies (L.) Karst.) is the main species of the Carpathian coniferous belt. Its longevity is documented to reach 576 years (Schweingruber F. H. – Wirth C. 2009). Norway spruce deserves special atten- tion in the regional tree-ring research due to its impor- tance from viewpoints of silviculture, dendroclimatology and dendroarchaeology. Several studies have been al- ready conducted to decipher climatic information pre- served in ring width variability of spruce in the Carpa- thians (e.g. Bednarz Z. et al. 1998-99, Szychowska- Krapiec E. 1998, Popa I. 2003, 2004, 2005, Savva Y. et al. 2006, Kaczka R. – Büntgen U. 2007, Kern Z. – Popa I. 2007, Popa I. – Kern Z. 2007, Bouriaud O. – Popa I. 2009), while other parameters were analysed solely in the Tatras (Büntgen U. et al, 2007). However, wood density data from spruce are also available from three eastern and the southern Carpathian sites. These sites were sampled in the frame of a global dendrochronologi- cal sampling campaign. These southern Carpathian stands were used to assess spatial teleconnections in the network (Schweingruber F. H. 1985) and in a broad scale dendroclimatological comparison (Schweingruber F. H. et al. 1987). Nevertheless, the climatic signal pre- served in the radiodensity data has never ever been ex- plored separately. The fact that Schweingruber F. H. (1985) defined an individual densitometric zone in the European coniferous belt centred on the southern Carpa- thians gives even further interest for the climatic inter- pretation of the southern Carpathian archive wood den- sity data. An additional motivation of this study was to test the earliest instrumental temperature record of the region (Ro-Sibiu/H-Nagyszeben/G-Hermannstadt) in order to see if its known inhomogeneities might be tracked com- paring with climate sensitive proxy records. Curiosity emerged as proxy/data comparisons suggested biased early instrumental data for many sites in Europe (e.g. Moberg A. et al. 2003, Frank D. et al. 2007a, Winkler P. 2009), and corrections were recommended for some temperature data even before the 1900s. MATERIAL AND METHODS Radiodensity data Wood density data originated from ITRDB database (Schweingruber F. H. 2000, NOAA 2009), from which database a southern Carpathian timberline site (45.30N, 23.67E, 1650 m a.s.l.) near Novaci in the Parang Mts (Fig. 1) was selected for this study. Sampled species was Norway spruce. The archive dataset contains 30 series of maximum and minimum density from 15 trees (each tree represented by two series). The wood density data were developed by the X-ray densitometric technique (Polge H. 1970, Schweingruber F. H. et al. 1978) in the framework of a global dendrochronological sampling network of conifers (Schweingruber F. H. 1985). Maximum and minimum density values are the characteristic parameters of the radial wood density structure. In a given tree ring the density maximum is linked to latewood and minimum is linked to earlywood (Schweingruber F. H. et al. 1978). The Novaci dataset covers a 178-years long period, 1804-1981 AD, so despite its almost tridecadal antiquity it 16 Z. Kern – I. Popa JOEG II/3-4 is still the longest wood densitometric record from the entire eastern and southern Carpathian domain. Mean tree age is 88 years and ranges from 35 to 178 years. Replica- tion decays before AD 1944 down to 4 at AD 1827 (Fig. 2). A sole series dates before AD 1821. Average maximum and minimum density of the population is 0.769 and 0.256 g/cm 3 , respectively. Average maximum density value fits nicely in the range found for Norway spruce in the Tatras (Büntgen U. et al. 2007), but lower than the value reported from a Slovenian alpine spruce stand (Levanič T. et al. 2009). Average minimum density fits similarly well in the range presented for spruce in the Alps (Schweingruber F. H. et al. 1978). Standardization, Index calculation Raw density data were subjected to standardization procedure to eliminate non-climatic trends and empha- size climatic signal (Cook E. R. et al. 1990). A relatively stiff cubic spline function (50% frequency cut-off at 300yrs) was fitted to each individual series (Cook E. R. – Peters K. 1981). Raw density data were converted into index values as ratio between measured and modelled values. Final population indices for maximum density (MXD) and minimum density (MND) were computed as biweigth robust mean (Cook E. R. 1985). Variance of derived MXD and MND chronologies was adjusted to minimize bias due to changing sample replication (Osborn T. J. et al. 1997, Frank D. et al. 2007b). Stability of climate related signal preserved in the index series was controlled by the Expressed Population Signal (EPS) statistics applying the standard acceptance thresh- old of 0.85 (Wigley T. et al. 1984). Mean interseries correlation (Rbar) and EPS were calculated for 40 yrs moving window with 20 yrs steps. Standardization and index calculation procedure was carried out using the ARSTAN software (Cook E. R. – Krusic K. J. 2006). Instrumental data For climate response analysis monthly temperature means of grid-box enclosing the Novaci tree-ring site (Fig. 1) were extracted from the CRU TS2.1 database (Mitchell T. D. – Jones P. D. 2005) from 1901 to 1981. The longest instrumental temperature record of the southern Carpathian region originates from Sibiu a prominent town of Transylvanian Saxons (Fig. 1). Ob- servations at Sibiu began in 1851. Monthly mean tem- perature data since 1851 are available via the Climate Explorer (van Oldenborg G. J. et al. 2005). Sibiu is re- garded as the most precise and most reliable temperature record in the region (Ţîştea D. at al. 1966). However, location of observation as well as termini and calculation method of daily mean temperature had changed a couple Fig. 1 Location of the sample site (filled square) and Sibiu (filled triangle) the earliest meteorological station of the region of southern Carpathians. Right corner: Location of Romania shown by black shading. Left corner: Black rectangle shows the location of the site in the southern Carpathians JOEG II/3-4 Assessing temperature signal in X-ray densitometric data of Norway spruce and the earliest instrumental record from the southern Carpathians 17 of times during the station’s history (Table 1). For in- stance, daily and inherently monthly mean temperatures are positively biased by the (T7+T14+T21)/3 calculation method compared to (T7+T14+2*T21)/4, the so-called Kaemtz-method (see Dall’Amico M. – Hornsteiner M. 2006). Nevertheless mean temperatures are obviously suffering negative bias due to the earlier morning (6 vs. 7 a.m.) and the later evening (22 vs. 21 p.m.) readings before 1880. Climatic signal analysis Pearson’s correlation coefficients have been com- puted between MXD, MND and gridded monthly mean air temperature to assess the temperature effect of each month on the annual radiodensitometric characteristics of spruce. Months from the June of the previous year to October of current year of tree-ring formation were in- volved in the analysis. After the first trial, bimonthly means of August-September (AS) for MXD and June- July (JJ) for MND were also invoked. In addition, the April-September (AMJJAS) mean temperature was also included into the MXD comparison. Correlation coeffi- cients were compared to 95% significance level derived from Student’s t-test to evaluate their importance. Inves- tigation was restricted to the period after 1906, because the first five years were designated as biased instrumen- tal data (see later). Temporal stability of relation between MXD/MND and Sibiu record was investigated by running correlation computed in 21-year moving windows. Moving window correlation technique is a standard tool to trace temporal stability in proxy-climate correlations (Aykroyd R. G. et al. 2001). The choice of any window width is somewhat arbitrary but the applied 21-year window over the screened 130-year long period is an acceptable compro- mise between a very narrow window (with a higher sampling error) and boarder window (which approximate the behaviour of the full dataset) (Aykroyd R. G. et al. 2001) and it is in conformity with choice of similar den- droecological studies (e.g. Carrer M. et al. 2007, Frie- drichs D. A. et al. 2009). Timing of detected unusual shifts was compared to events in station history. RESULTS AND DISCUSSION Novaci spruce MXD and MND record Both density indices (Fig. 2a) preserve quite strong common signal as evidenced by the high and stable signal strength statistics. Mean Rbar (EPS) is 0.44 (0.95) and 0.27 (0.91) for MXD and MND, respectively. Same statis- tics calculated in moving windows (Fig. 2b,c) showed exceptional signal stability and present that the relatively higher values found in mean values for MXD is perma- nently valid through the entire studied period. EPS values exceed the 0.85 threshold level suggesting robust chronol- ogy for both radiodensitometric parameters. The correla- tion between MXD and MND is 0.02, indicating that independent environmental information is preserved in the proxies. Wimmer R. and Grabner M. (2000) also found that in a spruce stand the density measured in earlywood (e.g. MND) were widely independent from latewood, suggesting that they are not under the same control. Relationship between temperature and maxi- mum/minimum density of spruce wood In the case of MXD, August and September yielded the largest coefficients among monthly means 0.56 and 0.39, respectively (Fig. 3a). These findings broadly agree with the previous observations carried out at vari- ous sites in the Carpathians (Schweingruber F. H. et al. 1987). However, comparison with April, May and June showed also significant positive coefficients. The coeffi- cient of July has not reached the 95% significance level but still positive and exceeds the other monthly re- sponses. Regarding AS bimonthly mean temperature the coefficient slightly improved, and the multi-monthly AMJJAS presented even bit higher coefficient (0.59). The latter one was found as the optimal target season corresponding to the climate information of spruce MXD Table 1 Recorded changes of observation conditions of the Sibiu/Nagyszeben/Hermanstadt station (source: Ţîştea D. et al. 1966). Row-head codes: A: date of change; B: location of observation; C: Termini (time of regular thermometer readings) and calculation method of daily mean temperature. Note the width of cell is not proportional with the duration of the represented period! A 1851 January 1861 1880 1887 May 1897 August 1900 July 1905 October 1921 1940 September 1947 August B Weinanger platz 14 Sag- gasse 15 Elisabeth- gasse 9 Schiff- bäumel 3 Lang- gasse 5 Reisbach- gasse 9 Weber- gasse 6 Strada Riului 5 Dimitri Anghel 3 C 6-14-22 arithmetic mean 7-14-21 arithmetic mean 7-14-2*21 Kaemtz-method 8-14-20 Köppen- formula 18 Z. Kern – I. Popa JOEG II/3-4 F ig . 2 R a d io d e n si to m e tr ic c h ro n o lo g ie s o f sp ru c e f ro m N o v a c i a n d g ra p h ic a l il lu st ra ti o n o f si g n a l st re n g th s ta ti st ic s c a lc u la te d f o r 4 0 y rs m o v in g w in d o w w it h 2 0 y rs s te p s, w h e re b la c k ( g re y ) c u rv e s a n d s y m b o ls s h o w M X D ( M N D ) re c o rd s. a : in d ic e s o f m a x im u m ( M X D ) a n d m in im u m ( M N D ) d e n si ty ;. b : in te rs e ri e s c o rr e la ti o n ( R b a r) ; c : e x p re ss e d p o p u la ti o n s ig n a l (E P S ). D a sh e d h o ri z o n ta l li n e d e n o te s th e 0 .8 5 t h re sh o ld l e v e l; d : sa m p le d e p th in the Tatras (Büntgen U. et al. 2007) and in the Alps (Frank D. – Esper J. 2005). The weakened mid-summer temperature response seems to be a characteristic pattern of the spruce MXD response. Both for the Alps and Tatras June coefficient dropped below 95% significance level (Frank D. – Esper J. 2005, Büntgen U. et al. 2007). In the case of the southern Carpathian spruce MXD chronology a bit different pattern was found. June coef- ficient is significant while for July it showed low values. JOEG II/3-4 Assessing temperature signal in X-ray densitometric data of Norway spruce and the earliest instrumental record from the southern Carpathians 19 Interesting to note that in a recent study on two Slovenian spruce stands reported September mean tem- perature as the main climatic driver of cell wall thicken- ing (Levanič T. et al. 2009). This discrepancy between the Slovenian and south Carpathian spruce MXD re- sponses, however, can be easily explained as the grow- ing season at the more temperate Slovenian sites is longer compared to the more continental southern Carpa- thian timberline. The shorter vegetation period means earlier cell maturation as well. None months from the year before tree-ring growth showed significant effect on cell wall thickening during the subsequent year. This lack of responses from the year preceding the growth is normal for the MXD (Schwein- gruber F. H. et al. 1987). In the case of MND June and July showed the strongest results -0.33 and -0.34, respectively (Fig. 3b). Regarding JJ bimonthly mean temperature the correla- tion further improved (-0.41). May mean temperature yielded similarly negative coefficient albeit it slightly lags behind the 95% significance level. This response is especially exciting as the pioneer milestone study of radiodensitometric dendroclimatology conducted in the Alps (Schweingruber F. H. et al. 1978) experienced no any clear relationship between this densitometric pa- rameter of spruce and climate variables, consequently MND was usually neglected in later researches. A sole exception is the study of Wimmer R. and Grabner M. (2000) from the following 30 years. They analysed 16 anatomical variables (including also MND and MXD) averaged from 20 Norway spruce trees over a relatively short, 40-year long, tree-ring sequence. They found poor climatic response in MND. However the neglected potential of MND as climate archive can be emphasized referring to a very recent study. Grabner M. et al. (2009) investigated MND record Fig. 4 Fluctuation of coefficients of 21yrs moving window correlation computed between radiodensitometric indices of Novaci spruce samples and their optimal bimonthly temperature target from the Sibiu record. a: maximum density vs. AS, b: minimum density vs. JJ. Dashed horizontal lines denote the 95% significance levels Fig. 3 Columns indicate the Pearson’s correlation coefficients computed between monthly and multi-monthly mean air tem- peratures and maximum (a) and minimum (b) density indices over the 1906-1981 interval. Few coefficients are given near to the corresponding column. Horizontal dashed lines denote the 95% significance level 20 Z. Kern – I. Popa JOEG II/3-4 in a dataset of juvenile spruce stands. Surprisingly they found stronger correlation between MND and climate variables than with MXD. Statistical results suggest that dominantly June-July thermal conditions determine the minimum earlywood density of spruce tree ring in the southern Carpathians. Higher June-July air temperature seems to retain cell wall thickening and/or to produce larger lumen diameter in the earlywood. May also has some minor role in this process. We propose that only the late half of May pre- ceding June affects the process, and the corresponding coefficient become non-significant as the monthly means integrates also the daily temperatures from the early half of the month. The relatively high positive coefficient of October lacks explanation as cell maturation in the earlywood section is probably ceased for mid-autumn. Even more, it might be a fake response due to biased instrumental reference (see later). Comparison with the earliest local instrumental temperature record For the sake of brevity out of the many prepared moving window correlation (hereafter MWC) analyses only a few arbitrary selected graphs are presented. They were chosen to best illustrate our conclusion. Fig. 4 shows the temporal fluctuation of the correla- tion coefficient calculated with the optimal bimonthly mean temperature target of both radiodensitometric pa- rameters (i.e. AS vs. MXD; JJ vs. MND). Although MND lost significance in the last decade, coefficients generally exceed the significance levels over the last seven decades. In contrast, coefficients of both densi- tometric parameters drop below significance level at 1905-1906. The coefficients reach significance level again before 1896 and 1880 for MXD and MND, respec- tively. Each date can be found in the station movement history. This suggests that between 1880 and 1905, when the location of observation frequently changed settling in one place, as short as 3 to 9 years homogeneity of obser- vation conditions were failed to maintain. As gridded temperature data largely relies on Sibiu record before 1920 in the region, this is the reason why the proxy/climate relationship was finally analysed on the truncated record from 1906. Coefficients of MWC calculated between MXD and April monthly mean temperature are presented in Fig. 5a. Coefficients are well above significance level during the early decades, they start to decline from 1880, and sink permanently below significance level from 1886. They fluctuate around zero level over few decades, afterwards abruptly increase from 1922, and exceed the significance level from 1924. The period of weakened even more non-significant coefficients coincide with the era of T07-T12-T21 readings, but also include the hectic station movements. The date of sudden recovery of posi- tive correlation relationship coincides with the introduc- tion of Kaemtz-method. The last sudden drop from 1943 to 1944 is a sole event which is not mirrored in station history. However, its time is equally near to a station movement (i.e. 1940) and a methodological change (in- troduction of Köppen-formula) which was accompanied with station movement, as well. MWCs calculated between MND and May and Oc- tober monthly mean temperatures are presented in Fig. 5b. Temporal evolution of MWC coefficients show strange shifts again. May coefficients fluctuate slightly below the 95% significance level through the main part of the studied period. Some decline can be seen during the last decade but a very unusual drift appears at the earliest time. Coefficients steadily increase before 1883, Fig. 5 Fluctuation of coefficients of 21yrs moving window correlation computed between radiodensitometric indices of Novaci spruce samples, and a few arbitrary chosen monthly mean temperature targets from the Sibiu record. a: maximum density vs. April, b: mini- mum density vs. May (open squares) and October (filled squares). Dashed horizontal lines denote the 95% significance levels JOEG II/3-4 Assessing temperature signal in X-ray densitometric data of Norway spruce and the earliest instrumental record from the southern Carpathians 21 change sign at 1881, and become significant before 1878. The period of unusual positive coefficients broadly coincides with the era of T06-T12-T22 readings. October coefficients show more sudden changes. The first from 1886 to 1887 when coefficients change from negative to positive, and the second after 1906 when swap sign again. Both date link to station movement. The coeffi- cients are close to zero for a decade, then start a gradu- ally rise from 1921, and abruptly jump to significance level after 1925. Coefficients fluctuate around signifi- cance level until 1946, after drop suddenly to zero level. The period of positive, occasionally even significant, MWC coefficients of October coincides with the utiliza- tion of the aforementioned Kaemtz-method. We suspect that the found positive response to October mean tem- perature might be the effect of the Kaemtz-method. CONCLUSIONS Climatic information was evaluated for archive Norway spruce maximum and minimum radiodensity data from a southern Carpathian location. Spruce MXD record showed significant positive relationship with the growing season (May-September) air temperature. Simi- lar response was found for the same species in the Tatras and the Alps. However, some characteristic discrepancy was also experienced (i.e. July vs. June signal weaken- ing). The southern Carpathian MXD response showed distinct difference compared to the Slovenian one, too (i.e. strongest reaction for August vs. September). These small differences compared to surrounding areas could plausibly explain the existence of the individual densi- tometric provenance defined by Schweingruber F. H. (1985) for the southern Carpathians. Spruce MND record showed clear and significant negative response to June-July mean air temperature. This is a novel result as traditionally this densitometric parameter was regarded not to carry any meaningful temperature signal. Hitherto dendroclimatological research focused ex- clusively on radial growth properties in the southern Carpathians (e.g. Soran V. et al. 1981, Popa I. – Cheval S. 2007, Kaczka R. – Büntgen U. 2007). In the present study, however, we point out that densitometric proper- ties have also great potential both in ecological and pa- laeoclimate reconstructions, and future researches are recommended in this field. Derived temperature sensitive proxy records were compared to instrumental data of the oldest available regional station. Results of the moving window correla- tion analysis showed strange shifts coinciding with changes in station history. Considerable inhomogeneities can be suspected in the instrumental data before 1906. The Sibiu temperature record would have eminent im- portance in the regional historical climatology due to its completeness and exceptional length. Nevertheless, this prominent secular record is hardly usable as temperature target in regional proxy paleoclimatological research until a scrutiny revision is done. Acknowledgements The research was funded by OTKA K67583. Popa I. is supported by IDEI ID65 project. This is Budapest Tree-Ring Laboratory contribution No 19. References Aykroyd R. G. – Lucy D. – Pollard A. M. – Carter A. H. C. – Robertson I. 2001. Temporal variability in the strength of proxy-climate correlations. Geophysical Research Letters 28(8): 1559-1562. Bednarz Z. – Jaroszewicz B. – Ptak J. – Szwagrzyk J. 1998- 1999. Dendrochronology of Norway spruce (Picea abies (L.) Karst.) in the Babia Góra National Park, Poland. Dendro- chronologia 16-17: 45-55 Bouriaud O. – Popa I. 2009. Comparative dendroclimatic study of Scots pine, Norway spruce and silver fir in the Vrancea Range, Eastern Carpathian Mountains. Trees-Structure and Function 23: 95-106 Büntgen U. – Frank D. C. – Kaczka R. J. – Verstege A. – Zwijacz-Kozica T. – Esper J. 2007. Growth responses to cli- mate in a multi-species tree-ring network in the Western Carpathian Tatra Mountains, Poland and Slovakia. Tree Physiology 27: 689-702 Carrer M. – Nola P. – Eduard J. L. – Motta R. – Urbinati C. 2007. Regional variability of climate-growth relationships in Pinus cembra high elevation forests in the Alps. Journal of Ecology 95: 1072-1083. Cook E. R. – Peters K. 1981. The smoothing spline: a new approach to standardizing forest interior tree-ring width se- ries for dendroclimatic studies. Tree-Ring Bulletin 41: 45-53 Cook E. R. 1985. A time series analysis approach to tree-ring standardization. Ph.D. dissertation, The University of Ari- zona, Tucson pp.171. Cook E. R. – Briffa K. – Shiyatov S. – Mazepa V. 1990. Tree– ring standardization and growth–trend estimation. In: Cook E. R. – Kairiukstis L. (eds.) Methods of Dendrochronology: Applications in the Environmental Sciences Kluwer Aca- demic Pub. 104–162 Cook E. R. – Krusic P. J. 2006. ARSTAN4.1b_XP. http://www.ldeo.columbia.edu. Dall’Amico M. – Hornsteiner M. 2006. A simple method for estimating daily and monthly mean temperatures from daily minima and maxima. International Journal of Climatology 26: 1929-1936 Frank D. – Esper J. 2005. Characterization and climate re- sponse patterns of a high-elevation multi-species tree-ring network in the European Alps. Dendrochronologia 22: 107- 121 Frank D. – Büntgen U. – Böhm R. – Maugeri M. – Esper J. 2007a. Warmer early instrumental measurements versus 22 Z. Kern – I. Popa JOEG II/3-4 colder reconstructed temperatures: shooting at a moving tar- get. Quaternary Science Reviews 26: 3298-3310 Frank D. – Esper J. – Cook E. R. 2007b. Adjustment for proxy number and coherence in a large-scale temperature recon- struction. Geophysical Research Letters 34: L16709. doi: 10.1029/2007GL030571 Friedrichs D. A. – Büntgen U. – Frank D. – Esper J. – Neu- wirth B. – Löffler J. 2008. Complex climate controls on 20 th century oak growth in central-West Germany. Tree Physiolo- gy 29: 39-51 Grabner M. – Schueler S. – Konrad H. – Karanitsch-Ackerl S. 2009. The analysis of drought events with the help of x-ray densitometry. EuroDendro2009: 59. Kaczka R. – Büntgen U. 2007. Spatial autocorrelation and growth/climate response of a high elevation spruce network along the Carpathian arc. In: Haneca K et al. (eds.) Tree rings in archaeology, climatology and ecology TRACE 5: 103-112 Kern Z. – Popa I. 2007. Climate-growth relationship of tree species from a mixed stand of Apuseni Mts, Romania. Den- drochronologia 24/2-3: 109-115 Levanič T. – Gričar J. – Gagen M. – Jalkanen R. – Loader N. J. – McCarroll D. – Oven P. – Robertson I. 2009. The climate sensitivity of Norway spruce (Picea abies (L.) Karst.) in the southern European Alps. Trees 23: 169-180 Mitchell T. D. – Jones P. D. 2005. An improved method of constructing a database of monthly climate observations and associated high resolution grids. International Journal of Climatology 25: 693-712 Moberg A. – Alexandersson H. – Bergström H. – Jones P. D. 2003. Were southern Swedish summer temperatures before 1860 as warm as measured? International Journal of Clima- tology 23: 1495-1521 NOAA Paleoclimatology 2009 http://www.ncdc.noaa.gov/paleo/treering.html Osborn T. J. – Briffa K. R. – Jones P. D. 1997. Adjusting variance for sample-size in tree-ring chronologies and other regional-mean time-series. Dendrochronologia 15: 89-99 Polge H. 1970. The use of X-ray densitometric methods in dendrochronology. Tree-Ring Bulletin 30: 1-10 Popa I. 2003. Analiza comparativă a răspunsului dendroclima- tologic al molidului (Picea abies (L.) Karst.) şi bradului (Abies alba Mill.) din nordul Carpaţilor Orientali. Bucovina Forestieră 11/2: 3-14 Popa I. 2004. Fundamente metodologice şi applicaţii de den- drochronologie. Câmpulung Moldovenesc: Editura Tehnică Silvică, 200 p. Popa I. 2005. Dendroclimatological research at Norway spruce (Picea abies (L.) Karst) and Swiss stone pine (Pinus cembra L.) from Ronda Mountains. Proceeding of the Romanian Academy Series B 7: 65-70 Popa I. – Kern Z. 2007. Efectul extremelor climatice asupra proceselor de creştere în pădurea de limită din Munţii Călimani. Revista Padurilor 122/2: 23-27 Popa I. – Cheval S. 2007. Early winter temperature reconstruc- tion of Sinaia area (Romania) derived from tree-rings of sil- ver fir (Abies alba Mill.). Romanian Journal of Meteorology 9: 47-54 Savva Y. – Oleksyn J. – Reich P. B. – Tjoelker M. G. – Va- ganov E. A. – Modrzynski J. 2006. Interannual growth re- sponse of Norway spruce to climate along an altitudinal gra- dient in the Tatra Mountains, Poland. Trees 20: 735-746 Schweingruber F. H. – Fritts H. C. – Bräker O. U. – Drew L. G. – Schär E. 1978. The X-ray technique as applied to den- droclimatology Tree-Ring Bulletin 38: 61-91 Schweingruber F. H. 1985. Dendroecological zones in the coniferous forests of Europe. Dendrochronologia 3: 67-75 Schweingruber F. H. – Bräker O. U. – Schär E. 1987. Temperature information from an European dendroclimatological sampling network. Dendrochronologia 5: 9-33 Schweingruber F. H. 2000. WSL-Biermensdorf Tree Ring data: http://www.ncdc.noaa.gov/paleo/treering-wsl.html Schweingruber F. H. – Wirth C. 2009. Old trees and the mean- ing of ’Old’. In: Wirth C. et al. (eds.) Old Growth Forests, Ecological Studies 207. Berlin, Heidelberg: Springer Verlag, 35-54 Soran V. – Andreica A. – Gîrlea D. – Bercea V. 1981. Studies about dendrochronology and dendroecology of Pinus cembra L. from the scientific reserve of Retezat Mountains (Southern Carpathians, Romania). (in Romanian) Ocrot. nat. med. În- conj. 25/1: 27-36 Szychowska-Krapiec E. 1998. Spruce chronology from Mt. Pilsko area (Zywiec Beskid Range) 1641-1995 AD. Bulletin of the Polish Academy of Sciences, Earth Sciences 46: 75-86 Ţîştea D. – Stonescu Ş. M. – Dissescu C. – Donciu C. – Nico- lae T. – Fetov V. 1966. Clima Republicii Socialiste Romania II. Date climatologice. Bucuresti. 277 p. van Oldenborgh G. J. – Balmaseda M. A. – Ferranti L. – Stockdale T. N. – Anderson D. L. T. 2005. Evaluation of at- mospheric fields from the ECMWF seasonal forecasts over a 15 year period. Journal of Climate 18: 2970-2989 Wigley T. M. L. – Briffa K. R. – Jones P. D. 1984. On the average value of correlated time series, with applications in dendroclimatology and hydrometeorology. Journal of Cli- mate and Applied Meteorology 23: 201-213 Wimmer R. – Grabner M. 2000. A comparison of tree-ring features in Picea abies as correlated with climate. IAWA Journal 21: 403-416 Winkler P. 2009. Revision and necessary correction of the long-term temperature series of Hohenpieissenberg, 1781- 2006. Theoretical and Applied Climatology 98: 259-268