Geoscience Reference
In-Depth Information
States, and since then numerous studies have used dendrochronological techniques
to record flood events in many regions, including northern California, the southwest-
ern United States, Colorado, South Dakota, British Columbia, and Manitoba (see
review in Yanosky and Jarrett 2002 ) . Tree-ring-based mass-balance estimates have
been generated for Peyto Glacier, Alberta, Canada (Watson and Luckman 2004 ) ,
and tree-ring proxies for winter glacial accumulation and summer ablation at Glacier
National Park have been used to assess recent glacier retreat in a 300-year context
(Pederson et al. 2004 ) .
8.3 Contributions to the Study of Water Resources
8.3.1 Extensions of Gauge Flow Records
Observed streamflow records in the western United States are seldom as long as a
century, and so cannot represent multicentury fluctuations due to climate variability,
should such fluctuations exist. Moreover, the gauge records in a sense are a snap-
shot in time of one particular part of the long-term hydroclimatic history, and the
snapshot may well be unrepresentative of extreme conditions (e.g., low-flow years)
that may have occurred. It should also be recognized that gauged flow records are
themselves imperfect measures of the volume of water passing the stream gauge,
and that the accuracy of the gauged record can change over time depending on the
type of gauge installed and the location of the gauge in the stream channel (Rantz
1982 ) .
The immediate aim of streamflow reconstruction is the temporal extension of
a time series of streamflow beyond the instrumented gauge record. Specifically,
reconstruction targets streamflow unaffected by works of humans, which include
artificial diversions, storage, modifications of the drainage network, and other fac-
tors. Streamflow defined in this way is sometimes also called 'natural flow,' 'virgin
flow,' or 'runoff' (Chow 1964 ) . The initial requirement in a streamflow reconstruc-
tion study is a time series of natural flow for statistical calibration with tree-ring
records. For basins with little impact of humans, the gauged flows may provide a
suitable calibration time series. For example, gauged flows for the Salt River near
Roosevelt, Arizona, were judged sufficiently free of anthropogenic effects for direct
use in a reconstruction model (Smith and Stockton 1981 ) . If human influence cannot
be dismissed as negligible, a modified time series of streamflow, adjusted to natural
conditions by restoring reservoir evaporation losses, artificial diversions, etc., must
be used in the reconstruction. Adjusted flow series were used; for example, in recon-
structions for the Colorado River at Lees Ferry, Arizona (Woodhouse et al. 2006 ) ,
and the Sacramento River, California (Meko et al. 2001 ) .
The extended streamflow series provided by tree-ring analysis have contributed
in many ways to a greater appreciation of the natural variability of streamflow
and the susceptibility of water resources to climatic fluctuations (Table 8.1 ) . Direct
input of time series of reconstructed flows into river management models to test
robustness of the system to extreme climatic variation is one natural application. An
Search WWH ::




Custom Search