Monday, April 9, 2012


date: Fri, 31 May 1996 09:16:30 +1200
from: Limin Xiong <>
subject: Proposed Programme of Research
to: Keith Briffa <>

NZ Science & Technology Post-Doctoral Fellowships Application Form


6.1 Programme Title:

Reconstruction of the climate since AD 1500 using tree-tings

6.2 Science Area (refer to section 6.2 of the Guidelines) H6

6.3 General Abstract:

At present, the possibility that the climate is changing and that this
may lead to unprecedented warming, is at the top of political agendas.
Greater understanding of climatic fluctuations is necessary, not
only for the last century, but also of earlier times. The longer
the record of climate is, the more confidence we will have in
determining how unusual recent events have been. The aim of this
proposal is to provide the last five centuries of climate data for
New Zealand derived from tree-rings. The reconstructed climate
would be compared with patterns seen in other parts of the
Southern and particularly the Northern Hemisphere (such as the
Little Ice Age). Although there is voluminous indirect evidence
that climatic conditions in the past 500 years were often
different from our contemporary experience, the precise nature of
these differences, and what caused them, remains elusive.
Tree-rings provide the best hope for an answer.

Keywords: Tree-rings; climate change; palaeoclimatology;

6.4 Programme Goal:

The goal of the proposed programme is to develop a network of
existing and new tree-ring sites which best enable the
reconstruction of New Zealand's past climate for the last 500
years. The programme is directly targeted to meet a priority
research need identified in the September 1994 Report of the
National Science Strategy Committee for Climate Change (p26). It
will also contribute to the Past Global Changes project of the
IGBP. This is important because international climatic data used
in general circulation modelling is dominated by Northern
Hemisphere information and dendroclimatology has the potential of
significantly extending the climate record in New Zealand.

6.5 Ethical or Regulatory Obligations:

Sampling new locations or updating some existing locations will
require approval from the Department of Conservation (DOC).
Permits have been obtained in the past and we would expect no
problems for future sampling.

6.6 Programme Significance and Justification:

New Zealand is of global significance due to the role of the
Southern Oscillation in the climate of this region, and the
opportunities the mid- and high Southern latitudes provide for
understanding the role of the oceans and the continents in climate
variability. Instrumental climate data has only been recorded for
about 150 years and this is widely acknowledged as too short to
detect the full range of natural variability likely to be present.
Dendroclimatology has the potential for extending New Zealand
quantitative climate records back several centuries (there are
only two other Southern Hemisphere regions where this potential
exists: southern South America and Tasmania). A workshop convened
by the Royal Society of New Zealand on 'Palaeoclimates and Climate
Modelling' (May, 1994) revealed that, while considerable knowledge
exists on palaeoclimates in New Zealand, very little is in the
quantitative form that is needed for understanding the present
limits of climate variability and for validating hindcast regional
climatic models. This research programme will address this by: (1)
providing high quality, quantitative, palaeoclimate information
for New Zealand, so that an improved balance of Southern
Hemisphere data is achievable in global climate prediction models;
(2) helping to ensure that the best possible quantitative data are
available to define the limits of climatic variability and for
calibration of climatic models; and (3) ensuring that such data
are available to test the hindcast output from regional climatic
models, thereby enabling greatest confidence to be placed in their
capacity to forecast within acceptable levels of accuracy.

All 84 tree-ring chronologies in New Zealand, produced by a range
people and methods from several different species, will be
screened and reanalysed (as a quality control measure). New
Zealand tree-ring width chronology has been established by Ahmed &
Ogden (1985), Fowler (1988), LaMarche et al. (1979), Norton
(1983a,b,c), Palmer (1989), Xiong (1995) and others. The
reanalysed information will then be lodged with the International
Tree-Ring Data Bank (ITRDB; Colorado, USA) and be publicly
available for the first time. In the data sparse Southern
Hemisphere, such a contribution will be highly significant for
global climate modelling and prediction. Some sites will also be
extended and updated and some new sites developed. New tree-ring
parameters - such as maximum and minimum wood intensity and
earlywood/latewood ratios - will be derived and analysed from each
species in an attempt to extract the maximum climatic information
possible. No New Zealand species has had all these parameters
investigated. The climate reconstructions will be compared with
earlier published work (Ogden & Ahmed (1989), Norton et al.
(1989), Palmer (1989), Salinger et al. (1994), Xiong (1995),
Murphy & Palmer (1992)) for comparison of results and gauging the
level of enhancement achieved.

The proposed research builds on the experience of the applicant
while at the same time taking full account of the research of

6.7 Programme Objectives and Method(s):

Tree-rings are widely recognised as one of the most important
proxy-climate records for analysing recent climate change, because
they are annual, widespread and easy to sample. The research
proposed here will follow well established methodology in
dendroclimatology (Fritts 1976; Cook & Kairiukstis 1990) and adopt
techniques and analytical approaches that are now largely
standardised at the international level. This will ensure
consistency of results with international research and permit
global comparisons to be made. The goal of this project is to
produce climate reconstructions from high-quality tree-ring
chronologies. Therefore, as much climate detail as possible will
be extracted from selected chronologies of long-lived species.
This will provide data to compare with other indicators of climate
(e.g. glacier advance and retreat, bore hole temperature,
speleothems) and other international studies.

The first methodological step involves quality control of all
available New Zealand tree-ring chronologies and the extension of
some of the key ones (Oct. 1996- July 1997). Of the 84 known
tree-ring chronologies, many are of uncertain quality. Recently
developed computer cross-dating and standardisation program
(Grissino-Mayer et al. 1992) will be used as a quality control,
screening, reanalysing where appropriate and prepared into a
consistent format. Tree-ring width chronologies using Nothofagus
menziesii, N. solandrii, Phyllocladus trichomanoides, P. glaucus,
and Libocedrus bidwillii have been extended back to 1500 AD, with
a number of chronologies back to the 13th century. Because of the
relatively small number of older samples in many of the
chronologies, older samples will be taken from some of the above
species to improve the reliability of this early data. To maximise
the climate signal, chronologies produced from areas that have
contrasting climate will be updated to allow a longer period of
calibration with the modern climate record. The focus of
additional collections will be on Libocedrus bidwillii and
Agathis australis. These are the two species in which the
chronologies could be extended back more than 1000 years. The
original collections of many of the NZ chronologies (Dunwiddie
1979) are hold by the Tree-Ring Laboratory, University of Arizona,
Tucson, Arizona, but have not been re-evaluated. As a result, it
will be a advantage to travel to USA for one month to screen these
samples during the middle of the first year (about March-April
1997). All the available chronologies will be standardised using
techniques that maintain as much of the low-frequency variance in
the tree-rings as possible (Briffa and Jones, 1990) so as to
extract climatic variability on the longest possible timescales.

The second methodological step will be the development of new
chronologies based on different aspects of the tree-rings such as
earlywood and latewood (Aug. 1997-Mar. 1998). Because none of this
kind of work has been tried in New Zealand, the initial focus will
be on the methodology using the new image analysis equipment
(Windendro scanner; Lottery Board funded to Dr Palmer 1995) at
Lincoln University. During this methodological phase emphasis will
be on Libocedrus bidwillii and Agathis australis over the period
1900-present. Once having evaluated the approach for the two
species, research will concentrate on density chronologies over
the period 1500-1900, and then extend to study a range of species
(e.g. Halocarpus biformis, Phyllocladus trichomanoides etc.).

The third methodological step will be establishing statistical
relationships between climate and tree-ring widths and wood
density for the period of instrumental record (i.e. response
function analysis) (Apr. 1998-May 1998). Response function
analysis (Fritts 1976) will be used to investigate the climatic
factors influencing tree growth at each site. This is a principal
component regression using a bootstrap technique relating monthly
climate variables to tree growth. The response function analysis
will identify the chronologies to use in the climate
reconstructions. The sequence will be: (a) Derive time series of
selected climate data (air temperature and precipitation),
including derived variables (e.g. Palmer drought index) from the
instrumental database. (b) Derive climate - tree-ring width, wood
density response functions for each tree-ring chronology site
using established response function techniques. (c) Analysis of
response functions obtained for multiple sites and multiple
species to identify spatial patterns and variations associated
with different environmental conditions.

The fourth methodological step will be the reconstructing of past
climate for earlier periods when only tree-ring data are available
(June 1998-Sept. 1998). Transfer function will be used to estimate
the climate (Fritts, 1991). Here, the principle component
regression with bootstraping techniques of Guiot (1991, 1995) will
be used to relate the selected network of tree-ring chronologies
separately to temperature, circulation indices and precipitation
data. Care will be taken to avoid artificial predictability. The
reconstructions will be calibrated and verified against relevant
high-quality data from the instrumental record. The results will
be compared with longer term reconstructions from other parts of
the globe. In the late of the second year(about June 1998), the
applicant will spent one month at the Climate Research Unit,
University of East Anglia, UK , working with Dr. Briffa to apply
the new statistical technique for standardising the chronology and
reconstructing the climate.

Information transfer: The results will be passed to the research
community by a paper at the end of the first year (annual
conference of NZ ecological society or NZ climatological society
in August 1997) and by a presentation in the second year
(international conference of tree-ring society; about May of
1998). The results will also be published in refereed journals.
The raw chronology data and reconstructed climate data will be
lodged with the International Tree-Ring Data Bank in the World
Data Centre-A, Boulder, Colorado, USA, for use by researchers.

Future plans: The initial focus will be on the period 1500-1900.
Future plans include extending the reconstruction to sub-fossil
material and to additional species. The eventual aim is to be able
to reconstruct the climate of the Holocene.

Limin Xiong, Plant Science Department, Lincoln University,
Canterbury, New Zealand. Phone: ++64-3-343,5198 (home);
Fax: ++64-3-325,3843 (dept.); Internet:
When raw wood is carved, it becomes a tool;
When a man is employed, he becomes a tool;
The perfect carpenter leaves no wood to be carved.
-- Lao-tzu, Tao Te Ching

No comments:

Post a Comment