Prediction of flow duration curves in ungauged basins - Runoff Prediction in Ungauged Basins: Synthesis across Processes, Places and Scales - page 141

and sediment-trapping dams) on the FDCs for four differ-
ent catchments located in the middle reaches of the Yellow
River in China, characterised by semi-arid continental
monsoon climate. They showed significant changes in
normalised FDCs between the baseline (1957

10 1

- 77) and

treatment (1978 2003) periods, with three of the four study

catchments showing significant reductions in runoff, espe-
cially in the range of low flows. All these studies represent
a fundamental wealth of knowledge that can assist in the
identification of the dominant catchment processes that
control the shape of the FDC under environmental change.

-

Test at Broadlands

10 0

Eden at Penshurst

10 -1

7.2.2 Similarity measures
Extrapolation and/or transfer of FDCs from gauged to
ungauged catchments is critically dependent upon the
notion of hydrological similarity, i.e., what are the relevant
physical (climatic and landscape) parameters that make
two catchments similar. Understanding hydrological simi-
larity requires knowledge and understanding of the rela-
tionships between characteristics of the FDC (magnitude,
shape etc.) and appropriate climatic and landscape
characteristics.

0 20 40 60 80 100

Percent of time exceeded

Figure 7.6. Flow duration curves normalised by the mean runoff for
the Eden at Penshurst, UK (area 224 km², mean annual precipitation
825 mm, clay-dominated in the lower parts), and the River Test at
Broadlands, UK (area 1040 km², mean annual precipitation 815 mm,
chalk-dominated catchment). Redrawn from Yadav et al.( 2007 ).

abstractions, return flows, impoundments or climate
change. In recent times, several experimental and empirical
studies have been carried out to explore the effects of
human-induced changes to the landscape, especially vege-
tation cover, on the FDCs. Brown et al.( 2005 ) presented a
review of Australian and New Zealand case studies on the
impacts of changing vegetation cover. Figure 7.7a depicts
the FDC response to conversion of deep-rooted native
forest to shallow-rooted pasture in the Wights catchment
in south-western Western Australia, a relatively dry region
where the annual actual evaporation of forests approaches
annual precipitation. In this case, the replacement of
native forest by pastures has led to a rapid rise of the
groundwater table, and associated groundwater runoff
(Schofield, 1996 ), resulting in large increases in low flows.
On the other hand, Figure 7.7b shows that reforestation has
the opposite effect on the FDC. It presents the FDCs for the
Red Hill catchment in Tumut, New South Wales, Austra-
lia, under one-year and eight-year-old pines, indicating a
50% reduction in high flows and a 100% reduction in low
flows with increasing age of the pines, with runoff in the
low flow range ceasing once the pine plantation becomes
well established (Vertessy, 2000 ).
Other kinds of anthropogenic impacts on the landscape
can also exert a strong influence on the FDCs, such as
water abstractions and construction of reservoirs (Brown
et al., 2005 ; Smakhtin, 1999 ). Mu et al. (2007) analysed
the effects of soil conservation measures (i.e., afforest-
ation, creation of stable pastures, construction of terraces

Runoff similarity
Natural indices for assessing similarity among FDCs on
the basis of runoff alone are the slope of the FDC or the
parameters of probability distributions fitted to them.
Examples of these similarity indices for a large number
of catchments in the USA are presented in Figures 7.8 and
7.9 . Figure 7.8a presents a map of the ensemble average
slope (over 50 years) of the middle limb of the FDC for
catchments in the eastern USA based on the work of
Sawicz et al.( 2011 ). The slopes were estimated on the
basis of the difference between 33% and 66% quantiles of
runoff for each year. The slope of the FDC in the middle
part of the curve, which is related to the variance of daily
runoff, is the result of the competition between the season-
ality of precipitation and that of potential evaporation, and
is mediated by subsurface drainage, thus providing a suit-
able synthetic similarity measure for the whole FDC.
On the other hand, Figure 7.8b presents a linear inter-
polation of the catchment average slopes of the FDC based
on the distances between the corresponding stream
gauging stations. Clearly, this distance or proximity-based
measure of similarity is the default case, which can be
substantially improved if the physical controls of the
FDC are well understood. Understanding of such connec-
tions between the slopes of the FDCs and climate and/or
catchment characteristics could provide a more advanced
and hydrologically sound regionalisation of the FDCs; this
is illustrated in the next example.