Tree-Ring Amplification of the Early Nineteenth-Century Summer Cooling in Central Europe

TLDR: In this paper, the suitability of tree-ring chronologies for reconstructing climate variability over centuries to millennia is discussed, and the most important proxy archives to reconstruct climate variability are presented.

Abstract: Annually resolved and absolutely dated tree-ring chronologies are the most important proxy archives to reconstruct climate variability over centuries to millennia. However, the suitability of tree- ...

Figures

FIG. 2. On the left-hand side, time series of (a) 10 instrumental (thin red lines), (b) 5 simulated (thin orange lines), and (c) 2 reconstructed (green and blue lines indicate the Alps and Tatra Mountains, respectively) CEU summer temperatures; that is, each time series was scaled against the 1971–2000 JJA reference period (see also Table 1 for details), together with their corresponding means after 20 yr low-pass filtering (black curves; cubic-spline smoothing with 50% frequency response cutoff at 20 yr). Right-hand sidemaps reveal spatial Pearson correlation patterns of the unfiltered mean of the (a) 10 instrumental, (b) 5 simulated, and (c) 2 reconstructed CEU summer temperatures records [i.e., their linear relationships with high-resolution (0.258 3 0.258) gridded JJA landmass temperatures from E-OBSv10.0 (Haylock et al. 2008, updated)]. All correlations were calculated over the 1971–2000 calibration period and only values exceeding r . 0.5 are plotted.

FIG. 5. Summer Z500 anomalies [in geopotential meters (gpm), with respect to the 1971–2000 climatology; data from Luterbacher et al. (2002)] over the North Atlantic/EU sector and the years 1815–22, with the black arrows indicating TRW increase or decrease relative to the previous year.

FIG. 4. (a) The mean of the 10 instrumental, 5 simulated, and 2 reconstructed (red, orange, and green) CEU summer (JJA) temperature anomalies (8C;w.r.t. 1971–2000), (b) the corresponding proxy–target residuals (i.e., TRW minus instruments), (c) mean cloudiness recorded in Milano and Torino (% sky coverage), (d) sum of relative solar radiation reassessed for PK (% sky coverage with respect to 1971–2000), and (e) precipitation totals (mm) averaged over the greater Alpine region. All time series are JJA anomalies of the period 1971–2000.

FIG. 1. (a) Natural Swiss stone pine (Pinus cembra) distribution across CEU (red), together with the geographical location of the two sampling regions in the western Austrian Alps (;478N and 128E; ;2300m MSL) and northern Slovakian Tatra Mountains (;498N and 208E; ;1500m MSL). (b) Wood anatomical microsection highlights the growth depression in a pine sample from the Tatra Mountains (CE32b) following the Tambora eruption [Lesser Sunda Islands, Indonesia, 10 Apr 1815, volcanic explosivity index (VEI) 7 and an estimated tephra volume of 1603 109m3]. However there are no growth responses visible following the Galunggung (Java, Indonesia, 8 Oct 1822, VEI 5 and an estimated tephra volume of 13 109m3) and the unknown eruption in 1809 (Guevara-Murua et al. 2014, and references therein). The vertical black lines refer to the three eruptions.

FIG. 3. (a) The mean of the 10 instrumental, 5 simulated, and 2 reconstructed (red, orange, and green) CEU summer (JJA) temperature anomalies (w.r.t. 1971–2000), with the gray vertical bar and values in degrees Celsius referring to the coldest decades of each of the three records (for data information and comparisons with volcanic forcing, see Table 1, Fig. 2, and Fig. S6). (b) The corresponding lag-1 autocorrelation functions of the mean instrumental, model and proxy records calculated over 31-yr moving intervals. (c) Anomalies of JJA cloudiness (percent of sky coverage with respect to 1971–2000) recorded in Milano and Torino back to 1763 and 1887 (blue and light blue), respectively. (d) Lag-1 autocorrelation structure of the cloud cover records computed as above.

Full text

Tree-Ring Amplification of the Early Nineteenth-Century Summer Cooling in
Central Europe
a
ULF BÜNTGEN,
b,c,d
MIROSLAV TRNKA,
d,e
PAUL J. KRUSIC,
f,g
TOMÁ

S KYNCL,
d,h
JOSEF KYNCL,
h
JÜRG LUTERBACHER,
i
EDUARDO ZORITA,
j
FREDRIK CHARPENTIER LJUNGQVIST,
k,l
INGEBORG AUER,
m
OLIVER KONTER,
n
LEA SCHNEIDER,
n
WILLY TEGEL,
o
PETR

S
T

EPÁNEK,
d
STEFAN BRÖNNIMANN,
c
LENA HELLMANN,
b,c
DANIEL NIEVERGELT,
b
AND JAN ESPER
n
b
Swiss Federal Research Institute WSL, Birmensdorf, Switzerland
c
Oeschger Centre for Climate Change Research, Bern, Switzerland
d
Global Change Research Centre AS CR v.v.i., Brno, Czech Republic
e
Institute of Agriculture Systems and Bioclimatology, Mendel University in Brno, Brno, Czech Republic
f
Department of Physical Geography, Stockholm University, Stockholm, Sweden
g
Navarino Environmental Observatory, Messinia, Greece
h
Moravian Dendro-Labor, Brno, Czech Republic
i
Department of Geography, Justus Liebig University, Giessen, Germany
j
Institute for Coastal Research, Helmholtz Zentrum, Geesthacht, Germany
k
Bolin Centre for Climate Research, Stockholm University, Stockholm, Sweden
l
Department of History, Stockholm University, Stockholm, Sweden
m
Central Institute for Meteorology and Geodynamics (ZAMG), Vienna, Austria
n
Department of Geography, Johannes Gutenberg University, Mainz, Germany
o
Institute for Forest Growth (IWW), University of Freiburg, Freiburg, Germany
(Manuscript received 6 October 2014, in final form 16 March 2015)
ABSTRACT
Annually resolved and absolutely dated tree-ring chronologies are the most important proxy archives to
reconstruct climate variability over centuries to millennia. However, the suitability of tree-ring chronologies to
reflect the ‘‘true’’ spectral properties of past changes in temperature and hydroclimate has recently been de-
bated. At issue is the accurate quantification of temperature differences between early nineteenth-century
cooling and recent warming. In this regard, central Europe (CEU) offers the unique opportunity to compare
evidence from instrumental measurements, paleomodel simulations, and proxy reconstructions covering both
the exceptionally hot summer of 2003 and the year without summer in 1816. This study uses 565 Swiss stone pine
(Pinus cembra) ring width samples from high-elevation sites in the Slovakian Tatra Mountains and Austrian
Alps to reconstruct CEU summer temperatures over the past three centuries. This new temperature history is
compared to different sets of instrumental measurements and state-of-the-art climate model simulations. All
records independently reveal the coolest conditions in the 1810s and warmest after 1996, but the ring width–
based reconstruction overestimates the intensity and duration of the early nineteenth-century summer cooling
by approximately 1.58C at decadal scales. This proxy-specific deviation is most likely triggered by inflated bi-
ological memory in response to reduced warm season temperature, together with changes in radiation and
precipitation following the Tambora eruption in April 1815. While suggesting there exists a specific limitation in
ring width chronologies to capture abrupt climate perturbations with increased climate system inertia, the
results underline the importance of alternative dendrochronological and wood anatomical parameters, in-
cluding stable isotopes and maximum density, to assess the frequency and severity of climatic extremes.
1. Introduction
It is well accepted that tree-ring chronologies can
provide annually resolved and absolutely dated tem-
perature and hydroclimatic reconstructions over centu-
ries to millennia (Frank et al. 2010; Masson-Delmotte
et al. 2013; and references therein). These records
a
Supplemental information related to this paper is available at the
Journals Online website: http://dx.doi.org/10.1175/JCLI-D-14-00673.s1 .
Corresponding author address: Ulf Büntgen , Swiss Federal Research
Institute WSL, Zürcherstrasse 111, CH-8903, Birmensdorf, Switzerland.
E-mail: buentgen@wsl.ch
5272 JOURNAL OF CLIMATE VOLUME 28
DOI: 10.1175/JCLI-D-14-00673.1
 2015 American Meteorological Society

represent the backbone of high-resolution paleoclimatol-
ogy, offering a long-term perspective of Earth’s climate.
The ability of tree-ring-based proxy records to accu-
rately capture the ‘‘true’’ spectrum of past natural and
recent anthropogenic climate change, including externally
forced and internally modulated stoch astic and quasi-
periodic climate variability (Hegerl et al. 2007; Esper
et al. 2012), has recently been challenged (Bunde et al.
2013; Franke et al. 2013; Tingley et al. 2014). Two possible
explanations for how tree-ring-based reconstructions may
under- or overestimate high- and low-frequency variations
in climate are the site-specific climate sensitivity of tree
growth (Fritts 1976) and the method of tree-ring stan-
dardization used to ‘‘detrend’’ raw measurements (Melvin
and Briffa 2008, 2014). Different detrending techniques
can dramatically affect the properties of a chronology
through their ability to preserve or remove specific fre-
quency bands from a time series (Esper et al. 2003, 2005).
At the same time, a dispute concerning a potentially
limited sensitivity to cold temperatures in trees growing
near the tree line, causing missing rings after very large
stratospheric volcanic eruptions (Anchukaitis et al. 2012;
Mann et al. 2012a,b; D’Arrigo et al. 2013; Esper et al.
2013b; Mann et al. 2013), has raised questions not only
about the reliability of tree-ring chronologies as high-
resolution climate proxy archives but more generally their
credibility as precise dating tools (Büntgen et al. 2014).
Strong tropical volcanic eruptions can trigger abrupt
perturbations in Earth’s climate system with subsequent
effects on human societies (Stothers 1999, 2000; Robock
2000; de Boer and Sanders 2002; Cole-Dai 2010). Such
eruptions represent a unique opportunity to evaluate
the climatological fingerprint of a rapid climate change
in proxy-based reconstructions (Hegerl et al. 2003, 2011;
Fischer et al. 2007; Wahl et al. 2014). In turn, so-called
detection and attribution studies aim to provide a better
understanding of climate models’ sensitivity to external
forcings and/or internal modulations (Barnett et al. 1999;
Hegerl et al. 1996; Merlis et al. 2014). Explosive volcanism
injects sulfate aerosols into the stratosphere, scattering
incoming solar radiation and absorbing outgoing infrared
radiation (Cole-Dai 2010). As a consequence, Earth’s
surface is cooled while the lower stratosphere is warmed
(Robock 2000). Precise estimates of the climate response
to a given volume, height, and chemical composition of an
eruption cloud depend on our degree of understanding the
physical connection between volcanic eruptions and at-
mospheric processes (Cole-Dai 2010; Esper et al. 2013a,b).
The relatively short residence time of volcanic aerosols in
the stratosphere limits the duration of direct volcanic im-
pacts on regional- and large-scale temperature and pre-
cipitation dynamics, lasting only a few years succeeding
an eruption (Robock 2000; Stenchikov et al. 2002;
Fischer et al. 2007; Cole-Dai 2010; Merlis et al. 2014).
However, recent results from coupled ocean–
atmosphere model simulations indicate that strong
tropical volcanic eruptions can initiate decadal-scale
dynamical responses in the climate system, thereby ex-
tending climate recovery beyond the short-lived radia-
tive forcing effect (Miller et al. 2012; Zanchettin et al.
2013a,b). The impacts of volcanic aerosols generated in
the stratosphere on tropospheric cloud formation, pre-
cipitation, and the diurnal temperature range are still poorly
documented (Auchmann et al. 2012, 2013; Wegmann
et al. 2014; Brugnara et al. 2015), neither from obser-
vations nor from the transient climate sensitivity in
forced models (Shindell 2014). Little information is
available on the influence of volcanic eruptions on regional
modifications of the global water cycle and atmospheric
circulation patterns (Fischer et al. 2007; Anchukaitis et al.
2010; Joseph and Zeng 2011; Timmreck 2012). This deficit
emerges from the fact that complex, and localized, sea-
sonal responses to volcanism may exist (Shindell et al.
2004; Trigo et al. 2009; Wahl et al. 2014), such as modu-
lations of the monsoon system and other large-scale tele-
connections (Wahl et al. 2014; Wegmann et al. 2014),
which are not yet fully quantified. Hydroclimatic behavior
following large volcanic eruptions may abate or exacerbate
the negative effects of reduced summer temperatures on
ecosystem functioning and productivity (Briffa et al. 1998;
Anchukaitis et al. 2012).
At the same time, it is still debated if enhanced pho-
tosynthetic activity under increased diffuse sunlight may
compensate for some of the cooling-related growth re-
ductions (Farquhar and Roderick 2003; Gu et al. 2003;
Krakauer and Randerson 2003). For example, the slight
drop in the growth rate of atmospheric CO
2
following
the 1991 Mount Pinatubo (Luzon) eruption could be
explained by an increase in forest net primary pro-
duction (NPP) if indeed forests use diffuse light more
efficiently than direct light for photosynthesis. Thus, the
fraction of diffuse sunlight in the years following an
eruption, because of scattered volcanic sulfur aerosols,
can have a dynamical effect on terrestrial ecosystem
productivity and the global carbon cycle (Gu et al. 2003).
Given our limited understanding of plant physiological
behavior under (rapid) climate change (Körner 2006), the
pulselike nature of posteruptive summer cooling in tan-
dem with possibly augmented photosynthetic activity due
to diffuse light provides optimal conditions for assessing
the growth response of high- and midlatitude forest trees
to abrupt negative summer temperature deviations
(Fischer et al. 2007; Esper et al. 2013a,b). The so-called
biological memory, inherent to all trees and partly re-
flecting their complex plant physiology (and possibly
ecosystem biogeochemistry), describes the dependency of
1JULY 2015 B Ü NTGEN ET AL. 5273

annual ring formation on previous year circumstances
(Frank et al. 2007). Although lagged responses are par-
ticularly strong during periods of feeble growth conditions
(Büntgen et al. 2006), it remains somewhat unclear if and
how tree-ring-based temperature reconstructions over-
estimate the amplitude and duration of postvolcanic
cooling (Esper et al. 2013a,b; Tingley et al. 2014). A de-
tailed examination of possible reconstruction e rror may
depend on having sufficient overlap between annually re-
solved tree-ring records and a credible set of instrumental
measurements during episodes of large volcanic eruptions
(Frank et al. 2007). Additional independent mechanistic
understanding can emerge from climate model simula-
tions (Gómez-Navarro et al. 2012, 2013, 2014, 2015;
Schimanke et al. 2012; Gutiérrez et al. 2013), which
should subsequently be considered in high-resolution
paleoclimatology ( PAGES 2k Consortium 2014). The
availability of extant instrumental, proxy, and model
data from central Europe (CEU), in the first half of the
nineteenth century, satisfies all these conditions.
To assess the level of coherency between tree-ring de-
viations and climate swings following large (mainly tropi-
cal) volcanic eruptions, we developed two independent
Swiss stone pine (Pinus cembra) ring width chronologies
from high-elevation, near–tree line sites in the Slovakian
Tatra Mountains and Austrian Alps. After combining
these datasets, the new compilation is used to reconstruct
interannual-to-centennial-long changes in CEU summer
temperature. Our proxy-based temperature history is
compared with paleoclimatic evidence of externally forced
model simulations from phase 5 of CMIP (CMIP5)/PMIP
phase 3 (PMIP3) experiments (Taylor et al. 2012)aswell
as with long instrumental measurements of temperature,
precipitation, cloud cover, and solar radiation. Attention is
paid to the detection of potentially differing responses to
increased volcanic activity in the early nineteenth century.
The early nineteenth century is of particular interest
as it includes the largest annually dated geographically
assigned eruption of the past millennium, Tambora in
April 1815 (Stothers 1984; Oppenheimer 2003). The
Tambora eruption produced the strongest simulated
summer cooling in the CMIP5/PMIP3 models back to
AD 850 (Masson-Delmotte et al. 2013) and probably
altered the hydrological cycle over parts of the North
Atlantic/European (EU) sector (Fischer et al. 2007;
Luterbacher and Pfister 2015), particularly over the Ibe-
rian Peninsula (Trigo et al. 2009). Further intensification
of the early nineteenth-century volcanic response
emerges from exceptionally low solar activity during the
Dalton Minimum between about 1790 and 1830 (Lean
et al. 1995; Wagner and Zorita 2005; Steinhilber et al.
2012). While focusing on the various limitations of tree-
ring width chronologies to properly capture the effects
of abrupt climate perturbations, our discussion also
emphasizes the potential of wood anatomical charac-
teristics, such as cell dimension and lignin concentration,
to more accurately capture changes in frequency and
severity of temperature extremes.
2. Data and methods
A total of 565 tree-ring width (TRW) samples (5-mm
cores) were collected in the Austrian Alps and the Slovakian
Tatra Mountains. All samples are from Swiss stone pines
(Pinus cembra) growing in recently protected, and thus
with significant likelihood to be relatively undisturbed,
tree line ecotones in western-central Austria (.2100 m
MSL; 410 samples) and northern Slovakia (.1500 m
MSL; 155 samples). The level of disturbance during
historical times, however, remains unknown. Site se-
lection resulted in a near-optimal spatial representation
of the species’ natural distribution across CEU (Fig. 1a).
Standard wood anatomical techniques were applied to
subsamples from both regions to quantify anomalous
TRW depression in the early nineteenth century (Fig. 1b
and Fig. S1a in the supplemental material). After cutting
the sample surfaces with a core microtome (Gärtner
et al. 2014), the resulting microslides were double
stained with safranin and astra blue (Büntgen et al.
2014), rinsed with ethanol, and finally embedded in
Canadian balsam. Digital images were taken with an
Olympus ColorView IIIu camera fitted to a Leica MZ12
microscope at 0.8 times magnification.
To guarantee the development of two independent
TRW chronologies, all measured series from Austria
and Slovakia were cross dated and standardized sepa-
rately at the country level. Data from living trees in the
Alps and Tatra Mountains span the periods 1417–2008
and 1687–2012, respectively (Fig. S1a). Their expressed
population signals (EPS; Wigley et al. 1984) are above the
0.85 quality threshold back to 1700 (Fig. S1b). The EPS
statistic, computed over 30-yr windows, lagged by 15 yr,
represents a summary measure of TRW coherency de-
scribing how well a chronology, based on a finite number
of samples, estimates the theoretical population from
which it was drawn (Briffa et al. 2013). The mean segment
length of the Alps and Tatra Mountains samples is 191 and
154 yr, respectively (Fig. S2 in the supplemental material),
and their average growth rates (AGR) are almost iden-
tical at 1.13 and 1.14 mm yr
21
.
Various detrending techniques were applied to remove
the nonclimatic, so-called biological age, trend from the
raw TRW measurement series (Fritts 1976; Linderholm
et al. 2014), while preserving high-to-low-frequency
temperature variability that occurred during the last
centuries, when CEU summer temperatures started to
5274 JOURNAL OF CLIMATE VOLUME 28

increase from exceptionally cold conditions in the early
nineteenth century to the most recent warming at the
onset of the third millennium (Luterbacher et al. 2004;
Büntgen et al. 2006, 2011; PAGES 2k Consortium 2013).
Cubic-spline smoothing with 50% frequency response
cutoff at 150 and 300yr (SPL; Cook and Peters 1981),
negative exponential and straight line functions (Neg1
and Neg2), as well as the regional curve standardization
(RCS; Esper et al. 2003) were applied using the most
recent version of the ARSTAN software (Cook and
Krusic 2005). Consideration of all five detrending
methods supports the assessment of interannual-to-
multidecadal, and possibly even lower, frequency in-
formation on centennial time scales. The corresponding
index values were calculated either as ratios or residuals
after power transformation (PT; Cook and Peters 1997)
between the nontransformed or transformed measure-
ments and their corresponding curve fits. The final TRW
chronologies for each country (Austria and Slovakia)
were produced using biweight robust means where tem-
poral variance changes in the chronologies were further
stabilized with respect to fluctuations in sample size
(Osborn et al. 1997).
The 10 Austrian TRW chronologies share a significant
fraction of common high-to-low-frequency variability
(Fig. S3 in the supplemental material). Their interseries
correlation (Rbar) is 0.86 (1723–2008). A statistically
similar coherency is found between all Slovakian chro-
nologies, reaching an Rbar of 0.93 over the period dur-
ing which records are replicated by at least 10 series.
Although temporally varying (Fig. S3c), the agreement
between all TRW chronologies from the Alps and Tatra
Mountains (Rbar 5 0.45), together with their shared
sensitivity to June–August (JJA) temperatures (Büntgen
et al. 2007, 2011), permits combining the two regional
mean chronologies into a single CEU Pinus cembra
chronology (details on the site-specific TRW behavior
and the effect of different calibration periods are pro-
vided in Figs. S3 and S5 in the supplemental material,
with statistics being summarized in Table 1). TRW data
from the Alps and Tatra Mountains were also separately
utilized to reconstruct JJA temperatures to assess their
coherency (Table 1).
The 10 slightly different, but not fully independent,
instrumental datasets of monthly and spatially resolved
gridded summer temperatures, including the Historical
FIG. 1. (a) Natural Swiss stone pine (Pinus cembra) distribution across CEU (red), together with the geographical
location of the two sampling regions in the western Austrian Alps (;478N and 128E; ;2300 m MSL) and northern
Slovakian Tatra Mountains (;498N and 208E; ;1500 m MSL). (b) Wood anatomical microsection highlights the
growth depression in a pine sample from the Tatra Mountains (CE32b) following the Tambora eruption [Lesser
Sunda Islands, Indonesia, 10 Apr 1815, volcanic explosivity index (VEI) 7 and an estimated tephra volume of 160 3
10
9
m
3
]. However there are no growth responses visible following the Galunggung (Java, Indonesia, 8 Oct 1822, VEI
5 and an estimated tephra volume of 1 3 10
9
m
3
) and the unknown eruption in 1809 (Guevara-Murua et al. 2014, and
references therein). The vertical black lines refer to the three eruptions.
1J
ULY 2015 B Ü NTGEN ET AL. 5275

TABLE 1. Extreme year variability and different summer temperature lag responses following the Tambora eruption (T-Res) as reflected by the various instrumental targets fHISTALP
in Auer et al. (2007) and Dobrovolný et al. (2010) [Coarse Resolution Subregional Means (CRSM) for low-level (low CRSM), northwest subregion (NW CRSM), and northeast subregion
(NE CRSM)]; Berkeley Earth in Muller et al. (2013) and Rohde et al. (2013)g, model simulations, and proxy reconstructions used in this study. All extremes were calculated for the JJA
seasonal mean and over the 1774–2005 common period after scaling each time series against the meteorological reference period 1971–2000. The 20 different TRW chronologies from the
Alps and Tatra Mountains that are herein summarized in minimum and maximum values with further information being provided in Fig. S3. Mean and standard deviation of the individual
years can diverge from the grand average values that were differently calculated and may refer to specific years, whereas the former statistics (mean and std dev) were always aggregated
over different years. The following parameter choices were applied for the Student’s t test (T-Test): 1) single-tailed, because the hypothesis evaluated is that the reconstruction results
have greater amplitude than the instrumental and model data, and 2) two-sample, unequal variance.
Data source Warmest year Coldest year
1-yr T-Res
1816
2-yr T-Res
1816–17
3-yr T-Res
1816–18
4-yr T-Res
1816–19
5-yr T-Res
1816–20
6-yr T-Res
1816–20
Instrumental
measurements
Berkeley (508–558N, 158–208E) 1992 (2.38C) 1821 (22.78C) 21.88C 21.18C 21.18C 20.78C 20.88C 21.18C
Berkeley (458–558N, 108–208E) 2003 (2.78C) 1821 (22.58C) 22.28C 2
1.48C 21.28C 20.88C 20.88C 21.18C
Berkeley (458–558N, 108–258E) 2003 (2.48C) 1821 (22.58C) 22.08C 21.38C 21.18C 20.88C 20.88C 21.18C
Berkeley (468–478N, 108–118E) 2003 (3.78C) 1816 (23.08C) 23.08C 21.98C 21.58C 21.38C 21.18C 21.38C
Berkeley (488–498
N, 198–208E) 1992 (2.58C) 1821 (22.68C) 21.98C 21.280 21.28C 20.88C 20.88C 21.18C
Berkeley (498–508N, 208–218E) 1992 (2.58C) 1821 (22.78C) 21.88C 21.18C 21.18C 20.88C 20.88C 21.18C
Dobrovolný (EU scale) 2003 (3.88C) 1816 (23.18C) 23.18C 22.18C 21.78C 21.48C 21.48C 21.68C
HISTALP (low CRSM) 2003 (3.98C) 1816 (22.98C) 22.98C 21.98C 21.68C 21.48C 21.28C 21.48C
HISTALP (NW CRSM) 2003 (4.28C) 1816 (23.68C) 23.68C 22.48C 21.98C 21.68C 21.58C 21.78C
HISTALP (NE CRSM) 2003 (3.48C) 1816 (22.78C) 22.78C 21.78C 21.48C 21.18C 21.18C 21.48C
Target grand avg 2003 (3.08
C) 1821 (22.58C) 22.58C 21.68C 21.48C 21.18C 21.08C 21.38C
Target mean 3.148C 22.838C 22.508C 21.618C 21.388C 21.078C 21.038C 21.298C
Target std dev 0.738C 0.348C 0.648C 0.468C 0.298C 0.338C 0.278C 0.238C
Model simulations BCC_CSM1.1 1992 (1.28C) 1821 (22.38C) 21.88C 21.18C 21.18C 20.78C 2
0.88C 21.18C
CCSM4 2003 (2.78C) 1821 (22.58C) 22.28C 21.48C 21.28C 20.88C 20.88C 21.18C
GISS-E2-R 2003 (2.48C) 1821 (22.58C) 22.08C 2138C 21.18C 20.88C 20.88C 21.18C
IPSL-CM5A-LR 2003 (3.78C) 1816 (23.08C) 23.08C 21.98C 21.58C 21.38C 21.18C 21.38C
MPI-ESM-P 1992 (2.58C) 1821 (22.68C) 21.98C 21.28C 21.28C 20.88C 20.88C 21.18C
Model grand avg 2003 (1.28C) 1816 (22.38C) 22.38C 22.08C 21.78C 21.68C 21.68C 21.58C
Model mean 2.508C 22.588C 22.188C 21.388C 21.228C 20.888C 20.868C 21.148C
Model std dev 0.898C 0.268C 0.488C 0.318C 0.16
8C 0.248C 0.138C 0.098C
Proxy reconstructions Tatra 1939 (2.28C) 1818 (24.28C) 23.68C 23.88C 23.98C 23.88C 23.78C 23.88C
Min 1939 (1.68C) 1818 (24.98C) 24.38C 24.48C 24.68C 24.58C 24.48C 24.48C
Max 1939 (2.78C) 1818 (23.58C) 22.98C 23.08C 23.28C 23.18C 23.08C 23.08C
Alps 2003 (1.68C) 1821 (23.78C) 22.88C 22.98C 23.18C 23.18C 23.28C 23.38C
Min 2003 (1.28) 1821 (24.78) 23.58C 23.68C 23.98C 23.98C 24.08C 24.18C
Max 2003 (2.98) 1821 (23.48) 22.88C 22.88C 22.98C 23.08C 23.08C 23.08C
Proxy grand avg 1983 (1.48) 1818 (23.88) 2
3.28C 23.38C 23.58C 23.48C 23.58C 23.58C
Proxy mean 2.038C 24.078C 23.328C 23.428C 23.608C 23.578C 23.558C 23.608C
Proxy std dev 0.688C 0.638C 0.608C 0.638C 0.648C 0.608C 0.588C 0.598C
T-Test proxy–target 0.005 13 0.001 71 0.012 72 0.000 12 0.000 09 0.000 02 0.000 02 0.000 05
T-Test proxy–model 0.183 27 0.000 64 0.003 41 0.000 07 0.000 08 0.000 01 0.000 02 0.000 06
5276 JOURNAL OF CLIMATE VOLUME 28

Frequently asked questions

Q1. What have the authors contributed in "Tree-ring amplification of the early nineteenth-century summer cooling in central europe" ?

This study uses 565 Swiss stone pine ( Pinus cembra ) ring width samples from high-elevation sites in the Slovakian Tatra Mountains and Austrian Alps to reconstruct CEU summer temperatures over the past three centuries. This proxy-specific deviation is most likely triggered by inflated biological memory in response to reduced warm season temperature, together with changes in radiation and precipitation following the Tambora eruption inApril 1815. While suggesting there exists a specific limitation in ring width chronologies to capture abrupt climate perturbations with increased climate system inertia, the results underline the importance of alternative dendrochronological and wood anatomical parameters, including stable isotopes and maximum density, to assess the frequency and severity of climatic extremes. 

Q2. What is the effect of volcanic aerosols on the stratosphere?

Explosive volcanism injects sulfate aerosols into the stratosphere, scattering incoming solar radiation and absorbing outgoing infrared radiation (Cole-Dai 2010). 

Q3. What were the two methods used to estimate changes in global radiation at PK?

Combining the diurnal temperature range with daily precipitation totals (Winslow et al. 2001) and using the diurnal temperature range exclusively (Donatelli and Campbell 1998) were two methods considered to estimate changes in global radiation at PK. 

Q4. What is the effect of a large volcanic eruption on the ozone layer?

A significant thinning of the ozone layer and a consequent increase of UV-B radiation can arise from particularly rich chlorine (Cl) and bromine (Br) emissions produced by large eruptions, such as Tambora (Self 2006). 

Q5. What is the effect of volcanic aerosols on the climate system?

The relatively short residence time of volcanic aerosols in the stratosphere limits the duration of direct volcanic impacts on regional- and large-scale temperature and precipitation dynamics, lasting only a few years succeeding an eruption (Robock 2000; Stenchikov et al. 2002;Fischer et al. 

Q6. What is the effect of volcanic eruptions on the ozone layer?

It is speculated that a highly variable stratospheric winter climate over the Arctic tends to reduce the effect of volcanic eruptions on the Northern Hemispheric ozone layer. 

Q7. What is the effect of reduced summer temperatures on ecosystem functioning and productivity?

Hydroclimatic behavior following large volcanic eruptionsmay abate or exacerbate the negative effects of reduced summer temperatures on ecosystem functioning and productivity (Briffa et al. 1998; Anchukaitis et al. 2012). 

Q8. What is the purpose of detection and attribution studies?

In turn, so-called detection and attribution studies aim to provide a better understanding of climate models’ sensitivity to external forcings and/or internal modulations (Barnett et al. 

Q9. What is the way to solve the differences between reconstructed and measured temperatures?

tree-ring-based paleoclimatic research would thus benefit from incorporating aspects of plant physiology and wood anatomy because climate models alone are insufficient to solve the differences between reconstructed and measured temperatures, such as those observed after the Tambora eruption. 

Q10. What is the effect of diffuse light on terrestrial ecosystem productivity?

the fraction of diffuse sunlight in the years following an eruption, because of scattered volcanic sulfur aerosols, can have a dynamical effect on terrestrial ecosystem productivity and the global carbon cycle (Gu et al. 2003). 

Q11. What is the effect of diffuse light on the growth of high- and midlatitude forest trees?

Given their limited understanding of plant physiological behavior under (rapid) climate change (Körner 2006), the pulselike nature of posteruptive summer cooling in tandemwith possibly augmented photosynthetic activity due to diffuse light provides optimal conditions for assessing the growth response of high- and midlatitude forest trees to abrupt negative summer temperature deviations (Fischer et al. 2007; Esper et al. 2013a,b).