Geography Reference
In-Depth Information
Table 2.1. Main performance measures used to evaluate the signatures in
Chapters 5
10. For definition of Level 1 and
Level 2 assessments see
Section 2.4.3
. For description of performance measures see
Table 2.2
. Runoff signatures are as
follows: Q: runoff, q: specific runoff, Q
100
: 100 yr flood runoff, Q
95
: low flow that is exceeded 95% of time
-
Level 1
Level 2
Type of variability analysed
Chapter 5
Annual runoff
r²ofQ, q, log Q, log q
NE, ANE of mean runoff
spatial
NSE, r
2
for each month
Chapter 6
Seasonal runoff
NE, ANE of range, NSE
temporal and spatial
Chapter 7
Flow duration curves
ANE of quantiles, proportion of
NSE
<
0.75 (NSE of quantiles)
NE, ANE of slope
temporal
Chapter 8
Low flows
R², r² RRMSE of q
95
NE, ANE of q
95
spatial
Chapter 9
Floods
RMSNE of q
100
NE, ANE of q
100
spatial
Chapter 10
Runoff hydrographs
NSE
NSE
temporal
Model efficiencies are a composite measure of bias and
random error. A Nash and Sutcliffe model efficiency
(NSE) of unity implies perfect predictions, smaller
values of NSE mean less perfect predictions (Nash and
Sutcliffe,
1970
).
of runoff. This analysis was conducted for all the signatures:
annual runoff, seasonal runoff, flow duration curves, low
flows, floods, and runoff hydrographs. The Level 1 assess-
ment is a meta analysis of prior studies performed by the
hydrological community. The advantage of this type of meta-
analysis is that a wide range of environments, climates and
hydrological processes can be covered that go beyond what
can be reasonably achieved by a single study. It is a compara-
tive assessment that synthesises the results from the available
international literature. However, the level of detail of the
information provided is often limited. The results in the
literature were almost always reported in an aggregated
way, i.e. as average or median performance over the study
region or part of the study region.
Level 2 assessment: To complement the Level 1 assess-
ment, a second assessment step was performed, termed Level
2 assessment. In this step, some of the authors of the publica-
tions from Level 1 were approached with a request to provide
data on their runoff predictions for individual ungauged
basins. The data they provided included information on
the catchment and climate characteristics, on the method used,
the data availability, and predictive performance. As in Level
1, the cross-validation performance for ungauged basins
was analysed; however, information on individual catch-
ments was now available. The overall number of catchments
involved was smaller than in the Level 1 assessment, so the
spectrum of hydrological processes covered in the assessment
was narrower. However, the amount of information available
on predicting runoff signatures in particular catchments
was much higher. Level 1 and Level 2 are therefore comple-
mentary steps, as illustrated in Fig. 2.12.
Note that some of the measures are performance meas-
ures, where 1 denotes perfect performance, while others
are error measures, where 0 represents perfect perform-
ance. In the assessment plots of
Chapters 5
9, performance
measures have been plotted upwards, while error measures
have been plotted downwards on the vertical axis.
Most performance measures can be calculated either on
the basis of runoff (m³/s) or on the basis of specific runoff
((m³/s)/km²). Runoff signatures tend to produce much
higher correlations for runoff than for specific runoff
because area is always an important predictor of runoff
due to mass balance considerations.
In addition to the quantitative performance measures,
qualitative reasoning can be used to help understand how
close the runoff predictions are to the real-world system,
i.e., how realistic the model predictions are. This aspect
might have to include extensive hydrological reasoning.
One example is the interpretation of the coefficients in the
regression equations. If they match the understanding one
has of the hydrological system, they can be considered
more realistic, and one would expect that they can then
be extrapolated more reliably to ungauged basins. Another
example is the degree to which runoff models in ungauged
basins represent the flow paths with the basin of interest.
-
2.4.3 Level 1 and Level 2 assessments
In order to perform the comparative assessment of runoff
predictions in ungauged basins, a two step process has
been adopted:
Level 1 assessment:Inafirststep,aliteraturesurveywas
performed. Publications in the international refereed litera-
ture were scrutinised for results of the predictive performance
2.5 Summary of key points
Catchments are complex systems that have evolved
through a process of reciprocal evolutionary change of
soils, vegetation and topography, mediated by water
fluxes, in response to long-term climate dynamics and
Search WWH ::
Custom Search