Experiments based on blue intensity for reconstructing North Pacific temperatures along the Gulf of Alaska
Ring-width (RW) records from the Gulf of Alaska (GOA) have yielded a valuable long-term perspective for North Pacific changes on decadal to longer timescales in prior studies but contain a broad winter to late summer seasonal climate response. Similar to the highly climate-sensitive maximum latewo...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Copernicus Publications
2017-08-01
|
Series: | Climate of the Past |
Online Access: | https://www.clim-past.net/13/1007/2017/cp-13-1007-2017.pdf |
Summary: | Ring-width (RW) records from the Gulf of Alaska (GOA) have yielded a valuable
long-term perspective for North Pacific changes on decadal to longer timescales in prior studies but contain a broad winter to late summer seasonal
climate response. Similar to the highly climate-sensitive maximum latewood
density (MXD) proxy, the blue intensity (BI) parameter has recently been
shown to correlate well with year-to-year warm-season temperatures for a
number of sites at northern latitudes. Since BI records are much less labour
intensive and expensive to generate than MXD, such data hold great potential
value for future tree-ring studies in the GOA and other regions in
mid- to high latitudes. Here we explore the potential for improving tree-ring-based reconstructions using combinations of RW- and BI-related parameters
(latewood BI and delta BI) from an experimental subset of samples at eight
mountain hemlock (<i>Tsuga mertensiana</i>) sites along the GOA. This is
the first study for the hemlock genus using BI data. We find that using
either inverted latewood BI (LWB<sub>inv</sub>) or delta BI (DB) can improve the
amount of explained temperature variance by > 10 % compared to
RW alone, although the optimal target season shrinks to June–September, which
may have implications for studying ocean–atmosphere variability in the
region. One challenge in building these BI records is that resin extraction
did not remove colour differences between the heartwood and sapwood; thus, long
term trend biases, expressed as relatively warm temperatures in the 18th
century, were noted when using the LWB<sub>inv</sub> data. Using DB appeared to
overcome these trend biases, resulting in a reconstruction expressing
18th–19th century temperatures ca. 0.5 °C cooler than the
20th–21st centuries. This cool period agrees well with previous
dendroclimatic studies and the glacial advance record in the region.
Continuing BI measurement in the GOA region must focus on sampling and
measuring more trees per site (> 20) and compiling more sites to
overcome site-specific factors affecting climate response and using
subfossil material to extend the record. Although LWB<sub>inv</sub> captures the
inter-annual climate signal more strongly than DB, DB appears to better
capture long-term secular trends that agree with other proxy archives in the
region. Great care is needed, however, when implementing different detrending
options and more experimentation is necessary to assess the utility of DB for
different conifer species around the Northern Hemisphere. |
---|---|
ISSN: | 1814-9324 1814-9332 |