Post-glacial climatic change and variability - The Global Climate System: Patterns, Processes, and Teleconnections - page 180

the best data sets and strongest research effort have occurred. The time periods
7800-8200 BP, and AD 800-1850 (containing the Medieval Warm Period (MWP)
and Little Ice Age (LIA)) are a focus for discussion. The chapter concludes with a
description of climate change in the Southern Hemisphere (SH), and comment on
comparison between the hemispheres. The evaluation of the post-1850 period is
left to Chapter 9 , where the impacts of global warming are assessed using global
circulation models. Links to human behavior are explored in Chapter 8 .
6.2 Determining past climate through the use of proxies
Detailed measurements of weather and climate are the best way to provide
accurate interpretations of climate variability and change. However, measure-
ments are rare earlier than AD 1850, and, those that do exist further back in time
are not representative of data from today's methodologies. Past climates, there-
fore, must then be determined indirectly, through the use of proxy indicators that
react or change as climate changes (Jones et al. 1998 ; IPCC 2001 ; Jones and
Mann 2004 ). There are two main types of paleoclimate research: annual to
decadal time series, which can be correlated against measured data to determine
strength and accuracy; and assessments at poorer resolutions, over periods of
100 or more years, which provide a much coarser picture. More detailed time
series provide the best information, with the coarser resolution proxy data
providing background support.
Paleoclimatic reconstructions can occur for a number of atmospheric para-
meters. Temperature proxies are most often used, however, because the range of
data is the largest, and temperature is possibly the most important indicator of
climate change. Moisture is the next best choice, but data are more limited, and
local and regional influences are strong confounding factors. Whatever the
proxy or the parameter, researchers must recognize the limitations in the use
of proxy data for past climate assessment. Some of the major proxy methods and
their limitations are listed in Table 6.1 . Trenberth and Otto-Bliesner ( 2003 ) list
several complexities that must be considered. These include seasonality, oscilla-
tions (Chapter 2 ), environmental interferences, quality and scale of dating,
potential biological factors, climate system reactions, and accurately establish-
ing what the proxy actually represents.
Several authors emphasize the importance of a multi-proxy integrated
approach which will allow the most accurate climate interpretations possible.
Calibration and standardization are critical, allowing the quality of correlation
and accuracy to be determined (Diaz and Bradley 1995 ). This can be difficult to
accomplish with data that are highly spatially diverse, and that represent differ-
ing time scale accuracies. Given the quality of most proxy results, resultant
multi-proxy temperatures are often plotted in annual or decadal averages, with a
30 to 50 year filter used to show major variations and trends.