Geography Reference
In-Depth Information
-
There is a slight tendency for the performance to
decrease with increasing air temperature. The pattern
of elevation dependence is more complex and may
relate to the regional patterns of aridity and snow
processes.
indices and topographic elevation (in mountainous
regions) are the most frequently used predictors of the
slope/shape of FDCs, in addition to the aridity index,
which governs the mean of the FDC.
Process-based methods are not widely used for predict-
ing FDCs in ungauged basins; however, there is great
potential for an increased use of process-based methods,
especially as we gain improved understanding of the
underlying process controls.
-
The performances increase with catchment size for
the case of geostatistics. Scale dependencies are less
clear for the other methods. The process-based
methods underestimate the slope of the FDC for the
smallest and largest catchments.
Comparative assessment of several prediction methods
indicated that, at least for some of the methods (e.g.,
regressions), predictive performance decreases with
increasing aridity (Level 1 and 2 assessments). Predic-
tions based on availability or collection of short records
outperform regionalisation methods in humid regions,
where inter-annual variability is small (Level 1 assess-
ment), whereas this may not be the case in arid regions
due to the fact that inter-annual variability is much
larger, and short records are insufficient to fully capture
this variability. Performance of geostatistical methods is
good (Level 2 assessment). These methods require
stream gauges in the region of interest.
-
Geostatistics and regression methods perform better
than the other methods. This may be partly related to
the higher stream gauge density of the data sets used
for these methods.
7.6 Summary of key points
The flow duration curve (FDC) is a statistical (i.e.,
frequency domain) representation of runoff variability
at all time scales (from inter-annual variability all the
way to event-scale variability), and therefore it embeds
within it aspects of all the other signatures studied in this
book. The mean of the FDC is mean annual runoff. The
seasonal flow regime smoothens out variability at both
short (i.e., floods) and long (i.e., low flows) time scales;
consequently the middle part of the FDC reflects runoff
variability that is reflected in the seasonal flow regime.
Much more insight into the FDC can be gained if the
contributions of both annual runoff variability and also
seasonal flow regime can be separated from the FDC,
with the distribution of the residuals explored through
the use of process-based models, especially of the
derived distribution type. Approached in this way, there
is considerable scope to approach the FDCs in a com-
parative manner, through bringing out the differences of
the FDCs between different places (e.g., climates and
landscapes), and seeking explanations for their differ-
ences using understanding of the underlying process
controls.
Similarity indices for FDC therefore include the aridity
index (for annual runoff variability), geology (for low
flows), storage capacity, mountain elevation and tem-
perature (for seasonal runoff variability), and event char-
acteristics (for floods).
The FDC also reflects the multiplicity of pathways
within the catchment that water follows and the associ-
ated time scales, and hence it connects to all of the co-
evolutionary processes impacting, and impacted by,
water flow processes, such as ecological, geomorpho-
logical and pedological processes.
Considerable potential exists for a joint investigation of
spatial patterns (within regions, along a river network,
and between regions) of not only FDCs, but also associ-
ated co-evolutionary features such as hydraulic geom-
etry, sediment stratigraphy, riparian vegetation and
patterns of biodiversity of aquatic biota.
Current predictions of FDCs are heavily dominated by
statistical methods. Geology or geology/soil
related
Search WWH ::
Custom Search