Instructional Presentation:
From Tree to Trace: How tree-ring reconstructions of streamflow
are generated
Part 1: How
Tree Rings Work (PDF)
Part 2: Building
a Tree-Ring Chronology (PDF)
Part 3: Generating
Streamflow Reconstructions (PDF)
Part 4: Validation,
Skill Assessment, and Uncertainty (PDF)
Part 5: Analysis
and Applications (PDF)
What is a tree-ring reconstruction of
streamflow?
A tree-ring reconstruction is a best-estimation
of streamflow for some past period using trees that have been
proven to be good estimators of streamflow over a more recent
period. A statistical model is developed which captures the relationship
between tree growth and the gage record during their period of
overlap. Then, this model is applied to the tree-ring data for
the period prior to the gage record.
to top
What is the physical
basis for tree-ring reconstructions?
One might think that trees growing right along
a river would act as natural stream gages, providing the best
information about streamflow. In fact, the growth of these riparian
trees is relatively insensitive to variation in streamflow, since
soil there is usually saturated, even during drought, and the
trees get all the moisture they need for growth.
The trees that do provide good proxy streamflow
data are typically found on slopes well above the river channel.
The relationship between streamflow and the growth of these trees
is indirect yet strong. The same climate factors, primarily precipitation
and evapotranspiration (loss of moisture from plants and soils),
control both the growth of moisture-limited trees and the amount
of water that reaches the stream. One can think of the tree as
a "dipstick" recording the overall moisture level in
a river basin--which is eventually expressed as streamflow.
Trees that provide the best information about
streamflow variability are those particularly sensitive to variations
in moisture. These include species such as ponderosa pine, pinyon
pine, and Douglas-fir, growing in open stands on dry and rocky
sites where soil moisture storage is minimal. Trees growing in
these types of sites are also less likely to be subject to non-climatic
disturbances (such as fires and insect infestation) and the effects
of competition from nearby trees. In addition, the oldest trees
(up to 800-1000 years old) of these species tend to be found on
these sites.
Trees used in hydroclimatic reconstructions are
not necessarily located in the same watershed as the instumental
or gage records, since tree growth and streamflow can be linked
by regional climate. The atmospheric flows of moisture which influence
both tree growth and streamflow cross watershed divides, so trees
in one basin may capture a significant portion of the variability
in streamflow in another basin. For example, reconstructions of
streamflow for the Colorado Front Range are improved when tree-ring
chronologies from western Colorado are added to the pool of model
predictors.
to top
How are tree-ring
reconstructions developed?
The reconstruction, of a streamflow record using
tree rings begins with the collection and development of tree-ring
chronologies. A chronology is a series of annual values derived
from the ring-width measurements of 10 or more trees of the same
species at a single site. To create a tree-ring chronology, cores
from the sampled trees at each site are crossdated (that is, patterns
of narrow and wide rings are matched from tree to tree) to account
for missing or false rings, so that every annual ring is absolutely
dated to the correct year. Then all rings are measured using a
computer-assisted measuring device. After growth-related (i.e.,
unrelated to climate) trends are statistically removed, the ring-width
values from all sampled trees for each year are averaged to create
a time series of annual ring-width indices.
Once a gaged flow record is selected for reconstruction,
a set of tree-ring chronologies from the region near the gage
is calibrated with the gage record to form a reconstruction model.
A statistical technique called multiple linear regression is commonly
used. The reconstruction model is evaluated by assessing how well
the reconstructed values replicate the observed values. The reconstruction
model is then validated by either testing it on a portion of the
gage data that was withheld from the calibration process, or testing
the ability of the chronologies used in the model to estimate
streamflow in different subsets of the data.
For a much more detailed description of the reconstruction
process and of how reconstructions are evaluated, see the Blue
River Case Study and the Instructional
Presentation.
to top
Why use tree-ring reconstructions
of streamflow in water management?
Water managers have long used instrumental records
of climate and gaged records of streamflow to assess the natural
variability of the system they are managing, determining the long-term
mean flow, frequency of drought events and establishing a drought
of record to use as the worst-case scenario in contingency planning.
However, instrumental and gaged records are usually only 30 to
100 years long and are unlikely to capture the full range of potential
natural variability. Tree-ring reconstructions, by providing a
much longer window into the past (300-1000+ years), more completely
describe the potential natural variability of the system, including
severe drought events. Nearly all tree-ring reconstructions have
indicated that droughts longer and more intense than those in
the instrumental record have occurred in past centuries. This
additional information on long-term hydroclimatic variability
can guide water resource planning to better meet the challenges
of potential future conditions. Water managers using tree-ring
reconstructions will not be surprised by events, like the 2002
drought, that exceed the bounds of the operational experience
of their system.
to top
How are tree-ring reconstructions
being applied to water management?
There are three basic ways the tree-ring data
are being used in water management:
1) As informal, qualitative guidance for managers,
stakeholders and decisionmakers. For example, a simple graphical
presentation of a reconstruction can be a powerful tool for increasing
awareness of the potential occurrence of droughts more severe
than those experienced during the gaged record.
2) For quantitative assessments of long-term hydrologic
variability. For example, assessing the frequency of reconstructed
droughts of a given duration and/or severity.
3) As direct inputs into hydrologic models of
a water system. This allows water managers to model system performance
under the tree-ring reconstructed hydrology, as they would do
with the gaged hydrology. This typically requires additional processing
of the reconstruction (annual values) to ingest it into the system
model, which may have monthly, weekly, or daily time steps.
Here are several examples of specific objectives
being addressed through the input of streamflow reconstructions
in system models:
- To test the effectiveness of drought response
actions in reducing demand to match supply in the ranfe of severe
drought years contained in the reconstruction
- For scenarios, using sequences of flow from reconstructions,
to evaluate the results of management decisions
- To evaluate alternatives for coordinated reservoir operations
under drought conditions
- To determine whether a given level of demand could be met in
all years of the reconstruction
to top
Is the science behind
tree-ring reconstructions new?
The first studies quantitatively relating tree-growth
to streamflow in the western US were done in 1930s. The first
modern tree-ring reconstructions of climate and streamflow (using
computers and multiple linear regression techniques) were developed
in the 1960s and 1970s. Among these was the reconstruction of
annual flow for the Colorado River at Lees Ferry by Stockton
and Jacoby (1976). Techniques for calibrating and validating
reconstruction models have been progressively refined since then.
to top
How accurate are
the reconstructions?
While the trees aren't stream gages, they do a
very credible job. Streamflow reconstructions in most areas of
the western US replicate most of the variability in flow seen
in the gaged records, and usually do particularly well capturing
the low flows. Typically, the "explained variance" (or
R2) ranges from 50% to 80%. As a specific example, a new reconstruction
of annual flow for the Colorado
River at Cisco, Utah explains 77% of the variance in the gaged
(natural) flow record (1906-1995). To put this in perspective,
water year precipitation for western Colorado (Division 2) explains
64% of the same gaged flow record. The error in the reconstruction
model can be estimated from the differences between the gaged
and reconstructed flows during the calibration or overlap period,
and model-based confidence intervals can then be applied to each
reconstructed value.
to top
