- Bri�a, K. (2000). Annual climate variability in the Holocene: interpreting the message 16 from ancient trees. Quat. Sci. Rev. 19, 87-105.- Brohan, P., Kennedy, J. J., Harris, I., Tett, S. F. B. and Jones, P. D. (2006). Uncertainty estimates in regional and global observed temperature changes: A new data set from 1850. J. Geophys. Res. 111, D12106, doi:10.1029/2005JD006548- D'Arrigo, R., Wilson, R. and Jacoby, G. ( 2006). On the long-term context for late twentieth century warming. J. Geophys. Res. 111, D03103, doi:10.1029/2005JD006352- D'Arrigo, R., Wilson, R., Liepert, B. and Cherubini, P. (2008). On the 'Divergence Problem' in Northern Forests: A review of the tree-ring evidence and possible causes. Glob. Planet. Change, 60(3-4), 289-305.- Esper, J., Cook, E.R. and Schweingruber, F.H. (2002). Low-frequency signals in long tree-ring chronologies for reconstructing past temperature variability. Science 295, 2250-2253.- Goetz, S.J., Mack, M., Gurney, K., Randerson, J. and Houghton, R.A. (2007). Ecosystem responses to recent climate change and �re disturbance at northern high latitudes Observations and model results contrasting northern Eurasia and North America. Environ. Res. Lett. (2), 045031.- Mitchell T.D. and Jones P.D. (2005). An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Climatol., 25, 693-712.

Laia Andreu¹, Rosanne D’Arrigo¹, Kevin Anchukaitis¹, Scott Goetz² and Pieter Beck²1Tree-Ring Laboratory, Lamont-Doherty Earth Observatory, 61 Route 9W, Palisades NY 10964 USA

²Woods Hole Research Center, Falmouth, MA 02540 USA

Varying Northern Forest Response to Arctic Environmental Change

The sensitivity of northern forests to Arctic warming and related seasonality e�ects is likely to be highly complex, involving shifts in the timing and dynamics of multiple factors that can signi�cantly impact tree growth.

The “divergence” problem

@D

'Arr

igo

et a

l. (2

008)

In recent decades, tree growth-climate relationships at some high northern latitude sites appear to have weak-ened. This phenomenon has become known as the “di-vergence” problem. If occurring on a large scale, it can have implications for the interpretation of dendrocli-matic reconstructions.

This plot compares several tree-ring based Northern Hemisphere temperature reconstructions (Bri�a 2000, Esper et al. 2002, D’Arrigo et al. 2006) that extend into the 1990s with land-based mean annual extra-tropical temperatures (20-90°N, Brohan et al. 2006). Reconstruc-tions scaled to instrumental data over the common 1856-1992 period and the linear increase per decade calculated over the same period.

Arctic vegetation trends

Trends in satellite observations of vegetation photo-synthetic activity derived from a 1982–2005 time series of GIMMS-G AVHRR vegetation indices, with sig-ni�cant positive trends show in light green and nega-tive trends in red (Goetz et al. 2007). These high-resolution satellite-derived vegetation observations indicate growing season “browning” in interior Alaska, transitioning to “greening” in the Alaskan tundra over recent decades, with an intervening region of rela-tively little trend in photosynthetic activity.

@G

oetz

et a

l. (2

007)

We investigate the response of northern boreal forests to changes in Arctic climate and seasonality, fo-cusing herein on tree-ring data from a site in northern Alaska, one of the regions of greatest warming on the globe today.

Summer temperature trends

Signi�cant trend in summer temperature anomalies for northern Alaska overlapping the Firth River tree-ring site (averaged over June, July and August) for the 20th century. The meteorological data set used was from CRU TS 2.1 grid point 0.5º (Mitchel and Jones 2005), ob-tained from Climate explorer (http://www.knmi.nl/).

This is one of the regions of most rapid warming on the globe today.

The Firth River site

The Firth River tree-ring width chronology is based on 233 samples from 111 trees (31 living trees and 84 cross-sections from subfossil wood). It is a millennial-length record that spans from AD 1067 to 2002 (936 years).

A tree ring-width chro-nology was established at the Firth River area (northern Alaska) to identify the nature of forest response to cli-matic and environmen-tal changes for white spruce (Picea glauca), a dominant Arctic tree-line species.

DAWSON

Tree Ring-Width chronology (RW)

Correlations residual RW vs local meteorological data

Our results indicate a signi�cant de�nite divergence between ring-width and summer temperature at the 2nd half of the 20th century. This divergence has serious implications for paleoclimatic reconstructions based on tree ring-width data.

Same correlation patterns as Dawson, but di�erent level of signi�-cance (note: only 20 years available for the 1st period). Although June and Summer correlation values are not signi�cant, their de-creases from the 1st to the 2nd period are statistically signi�cant.

Signi�cant correlations between residual RW and summer tem-perature months from 1901-1950, but none signi�cant from 1951-2001. Decreases in June, August and Summer correlations from the 1st to the 2nd half of the 20th are statistically signi�cant.

Signi�cant correlations (p<0.05) are found be-tween residual RW and summer temperatures in the 1st half of the 20th century, with the broadest spatial extent in June. In contrast, almost no signi�cant correlations are found in the 2nd half of the 20th century for the same months.

Spatial correlations residual RW vs CRU TS 2.1

Firth River site Fairbanks, USA Dawson, Canada

1100 1200 1300 1400 1500 1600 1700 1800 1900 2000-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

YEARS

ANO

MAL

IES

°C

0

0.31

0.62

0.94

1.25

1.56

1.87

2.19

2.5

IND

ICES

NH TEMPERATURE RECONSTRUCTIONALASKAN RCS CHRONOLOGY The tree-ring width Regional Curve Standardized chronology (RCS)

shows evidence for inferred low-frequency temperature variability for the past millennium that is consistent with a large-scale tem-perature reconstructions for the Northern Hemisphere (NH) extra-tropics (D’Arrigo et al. 2006).

Low frequency trends

The quality of the RW chronology is reliable (Expressed Population Signal or EPS > 0.85) since the middle of the 14th century due to the low sample depth prior to AD 1354.

r = 0.47, p < 0.01 (1067-1995)r = 0.71, p < 0.01 (1354-1995)r = 0.44, p < 0.01 (1950-1995)

residual RW vs Dawson temperature

-0.4

-0.2

0

0.2

0.4

0.6Period 1901-2001

-0.4

-0.2

0

0.2

0.4

0.6Period 1901-1950

Jan Feb Mar Apr May Jun Jul Ago Sep Oct Nov Des Ann Win Spr Sum Aut

-0.4

-0.2

0

0.2

0.4

0.6Period 1951-2001

No Significant correlations (p ≥ 0.10) Significant correlations at 95% (p < 0.05)

residual RW vs Fairbanks temperature

-0.4

-0.2

0

0.2

0.4

0.6Period 1930-2002

-0.4

-0.2

0

0.2

0.4

0.6Period 1930-1950

Jan Feb Mar Apr May Jun Jul Ago Sep Oct Nov Des Ann Win Spr Sum Aut

-0.4

-0.2

0

0.2

0.4

0.6Period 1951-2002

No Significant correlations (p ≥ 0.10) Significant correlations at 90% Significant correlations at 95% (p < 0.05)

Maximum Latewood Density tree-ring chronology (MXD)

The high frequency summer temperature signal registered in Maximum Latewood Density record seems to be reasonably stable through the 20th century, in contrast to the instability found in the ring-width proxy.

Correlations residual MXD vs local meteorological data

Same correlation patterns as Dawson in the 2nd period with di�er-ent signi�cance levels, but di�erent pattern in the 1st period (20 years of data available). The statistically signi�cant increase in August correlation reinforces the seasonality shift hypothesis.

Signi�cant correlations between residual MXD and summer tem-perature months for the both halves of the 20th century. A season-ality shift is suggested by a correlation decrease in July versus an increase in August. These changes are not statistically signi�cant.

Signi�cant correlations (p<0.05) are found be-tween residual MXD and summer tempera-tures in the 1st half of the 20th century, with a broader spatial extent for July and August. In the 2nd half of the 20th century the spatial cor-relation extent became wider and the correla-tion values increase. The highest correla-tions for August cover-ing almost all of Alaska are noteworthy.

Spatial correlations residual MXD vs CRU TS 2.1

Firth River site Fairbanks, USA Dawson, Canada

Low frequency trends

The MXD record is a preliminary chronology based on 20 samples from 16 trees that spans from AD 1691 to 2002 (311 years). The re-sidual MXD is reliable (EPS > 0.85) for the 20th century since prior to AD 1900 the sample depth is lower than 12 samples.

The Maximum Latewood Density (MXD) raw chronology shows evi-dence for inferred low-frequency temperature variablity for the past millennium that is consistent with a large-scale temperature recon-struction for the NH extra-tropics (D’Arrigo et al. 2006) until the 20th century. However, since that time the two series appear to diverge.

1100 1200 1300 1400 1500 1600 1700 1800 1900 2000-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

YEARSAN

OM

ALIE

S °C

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1.3

IND

ICES

NH TEMPERATURE RECONSTRUCTIONALASKAN MXD CHRONOLOGY

r = 0.21, p < 0.01 (1691-1900)r = 0.12, p = 0.25 (1900-1995)

residual MXD vs Dawson temperature

-0.4

-0.2

0

0.2

0.4

0.6Period 1901-2001

-0.4

-0.2

0

0.2

0.4

0.6Period 1901-1950

Jan Feb Mar Apr May Jun Jul Ago Sep Oct Nov Des Ann Win Spr Sum Aut

-0.4

-0.2

0

0.2

0.4

0.6Period 1951-2001

No Significant correlations (p ≥ 0.10) Significant correlations at 90% Significant correlations at 95% (p < 0.05)

residual MXD vs Fairbanks temperature

-0.4

-0.2

0

0.2

0.4

0.6Period 1930-2002

-0.4

-0.2

0

0.2

0.4

0.6Period 1930-1950

Jan Feb Mar Apr May Jun Jul Ago Sep Oct Nov Des Ann Win Spr Sum Aut

-0.4

-0.2

0

0.2

0.4

0.6Period 1951-2002

No Significant correlations (p ≥ 0.10) Significant correlations at 90% Significant correlations at 95% (p < 0.05)

AcknowledgementsProject entitled "Shifting Seasonality of Northern Forest Response to Arctic Environmental Change" (PIs R. D'Arrigo, K. Anchukaitis, S. Goetz and P. Beck) funded by the National Science Foundation's Arctic Seasonality Program under Grant ARC-0902051. We also would like to thank Brendan Buckley for the great �eldwork and Jan Wunder for interesting discussions.