Geography Reference
In-Depth Information
methodology was illustrated by a case study in the Upper
Mississippi basin using TOPEX/ Poseidon (T/P) satellite
data. Even without any runoff data, the runoff predictions
of the hydrological model were fairly reasonable. The
authors also demonstrated for their case that model param-
eter uncertainty was the main source of uncertainty, while
the contribution of remote sensing data uncertainty was
much smaller. Remotely sensed inundation patterns can
also be used profitably to improve model parameter esti-
mation (e.g., Grayson et al.,
2002
; Bauer,
2004
).
They achieved very good fits to runoff when calibrating
the parameters to runoff but other criteria, such as the
simulated new water contributions to peak runoff, were
not realistic. Constraining the model parameters by such
soft data criteria resulted in lower runoff model perform-
ance but represented the runoff mechanisms better as inter-
preted by the field hydrologists.
Winsemius et al. (2009)
integrated hard and soft hydrological information to
constrain model parameters of a conceptual runoff model
based on the generalised likelihood uncertainty estimation
(GLUE) method. The information they used was the shape
of the recession curve of the hydrograph (hard hydro-
logical information), spectral properties of daily runoff
from a period that was different from the precipitation data
period (hard statistical information), and monthly water
balance estimates based on old monthly averaged records
of precipitation and runoff (soft hydrological information).
They tested the method for the Luangwa catchment in
Zambia and found consistent parameter distributions and
a considerable reduction in the uncertainty of the param-
eters as compared to using a-priori parameters. Various
types of field-based soft data are also potentially useful for
assisting in constraining parameters. Soft data include sat-
uration areas as mapped in the early work of Dunne and
Black (
1970
) and
Dunne et al. (1975)
. There is renewed
interest in saturation patterns as illustrated by a number of
mapping projects (e.g., Kirnbauer et al.,
2005
) and their
application in constraining parameter
Tracers (possibly regionalised)
When tracer data are available, they can be used to
improve the conceptual understanding and therefore the
model structure for the catchment of interest (e.g., Fenicia
et al.,
2008a b
; Son and Sivapalan,
2007
). Tracer data may
also have potential to reduce the uncertainty of the model
parameters and improve the runoff predictions in ungauged
basins. Using tracer information with an integrated multi-
criteria calibration for a gauged catchment,
Bergström
et al. (2002)
found that, while the runoff performance
slightly decreased, the simulated tracer dynamics increased
significantly. Vaché and McDonnell (
2006
) rejected
unsuitable model parameters with the help of tracer data,
thereby increasing confidence in the flow paths conceptu-
alisation of a catchment. By adjusting the model structure
and parameterisation,
Birkel et al. (2011)
improved the
runoff model performance from a NSE of 0.71 to 0.74.
McGuire et al. (2007)
used tracers to constrain the feasible
model parameter space. In ungauged catchments there are
usually no tracer data available, but regionalisation
methods discussed in
Section 4.5
may have potential to
assist in parameter estimation. As discussed in
Chapter 4
,it
is important to recognise that tracers give information on
the movement of particles (related to the hydraulic con-
ductivity), while runoff gives information on the propaga-
tion of pressure (related to the compressibility of the
medium). These differences need to be taken into account
when using tracer data for constraining model parameters.
estimation in
rainfall
runoff modelling (e.g.,
Franks et al., 1998
). Soft
information from reading the landscape (
Chapter 3
)is
another option to be included in the parameter estimation
process. Overall, this general approach of including soft
information appears to have considerable potential for
constraining parameters in ungauged catchments and
exploiting any information one may have on the runoff
process in the catchment.
-
10.5 Comparative assessment
The aim of the comparative assessment of runoff hydro-
graph predictions in ungauged basins is to learn from the
similarities and differences between catchments in differ-
ent places, and to interpret the differences in performance
in terms of the underlying climate
Soft data and expert judgement
All the data sources discussed above can be combined as
available in order to constrain model parameters. Addition-
ally,
or qualitative information from field
surveys have been suggested in the literature to improve
model parameters beyond a-priori values (e.g., Blöschl,
2005
).
'
soft data
'
landscape controls.
Understanding these controls sheds light on the nature of
catchments as complex systems and provides guidance on
what methods to choose in a particular environment. The
assessment is performed at two levels (see
Section 2.4.3
).
The Level 1 assessment is a meta-analysis of studies
reported in the literature. The Level 2 assessment involves
a more focused and detailed analysis of individual basins
from selected studies from Level 1 in terms of how the
-
information is widely used in practical appli-
cations of catchment models where parameters are selected
based on all sources of information available to the analyst
and more formal methods for incorporating soft informa-
tion have been proposed. Seibert and McDonnell (
2002
)
used a fuzzy membership function to constrain model
parameters for the Maimai catchment in New Zealand.
'
Soft
'
Search WWH ::
Custom Search