Mike she usa

Journal of Hydrology (2007) 334, 73– 87
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jhydrol
Assessment of the effects of DEM gridding on the
predictions of basin runoff using MIKE SHE
and a modelling resolution of 600 m
R.F. Va´zquez a,*, J. Feyen b,1
a Centro de Investigacioń y Tecnologıá Agroalimentaria de Aragoń, Unidad de Suelos y Riegos, Avenida de
Montanãna 930, 50059 Zaragoza, Spain
b Division Soil and Water Management, Department of Land Management and Economics, K.U.Leuven, Celestijnenlaan
200 E, 3001 Heverlee, Belgium
Received 7 February 2006; received in revised form 26 September 2006; accepted 1 October 2006
KEYWORDS
DEM;
TOPOGRID;
MIKE SHE;
Catchment distributed
modelling;
Resolution
Summary A 586-km2 catchment was modelled with the distributed hydrologic model
MIKE SHE. Coarse digital elevation models (DEMs) having a 600-m resolution and gridded
from a set of elevation points geographically distributed with a much finer resolution were
used in the modelling with the purpose of investigating potential effects of the DEM gen-eration
methods on (i) model parameter values; (ii) adequacy of model global predictions;
and (iii) the evaluation of internal state predictions. To address these aspects, this paper
describes the DEM gridding methods, assesses the accuracy of the DEMs and examines sys-tematically
the sensitivities of parameter values and predictions of the distributed model
with respect to the DEMs. Three types of gridding methods were applied. Methods type I
were based on the use of the MIKE SHE interpolation tool (Bilinear algorithm) for process-ing
input elevation data distributed about the periphery of the gridded DEM cells. Input
elevation data distributed about the centre of the gridded DEM cells were processed in
gridding methods type II. The third type was based on the use of the TOPOGRID algorithm
that considers landscape features, such as digitised streams, to improve the drainage
structure of the gridded DEMs. A multi-criteria protocol was applied for assessing the ele-vation
quality of DEMs and their suitability for hydrologic purposes. It was found that the
quality of the DEM products of the MIKE SHE interpolation tool were poorer. The indepen-dent
calibration of the assembled hydrologic models revealed (i) important variations of
* Corresponding author. Tel.: +34 976 716324; fax: +34 976 716335.
E-mail addresses: raulfvazquezz@yahoo.co.uk, rvazquezz@aragon.es (R.F. Va´zquez), jan.feyen@biw.kuleuven.be (J. Feyen).
1 Fax: +32 16 329760.
0022-1694/$ - see front matter ª 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jhydrol.2006.10.001

74 R.F. Va´zquez, J. Feyen
model predictions; and (ii) from average to important variations of effective parameter
values, as a function of the different DEMs. A multi-criteria protocol analysing discharge
time series, peak flows and piezometric levels showed that model performance is in broad
terms in agreement with the elevation and slope quality of the DEMs.
ª 2006 Elsevier B.V. All rights reserved.
Introduction
Digital Elevation Models (DEMs) are important tools in
hydrologic research and water resources management owing
to the relevance that geo-morphological features intrinsic
in the DEMs have for the simulation of important water flow
processes such as surface runoff, evaporation and infiltra-tion.
However, DEMs, as source of spatially distributed
ground elevations, are not free of errors and limitations.
DEM square-grid structures have limitations for handling dis-continuities
in elevation and representing adequately all of
the landscape features. Indeed, either triangulated irregu-lar
networks (TIN) or contour lines should be preferred for
representing a surface for hydrologic purposes (Wise,
2000; Vivoni et al., 2005). However, square-grid DEMs are
still widely used for hydrologic purposes owing mainly to
their simplicity and computational efficiency.
In this context, the referred limitations of grid DEMs for
handling discontinuities in elevations and representing
appropriately landscape features are reduced by decreasing
as much as possible their grid size (Walker and Willgoose,
1999; Wise, 2000). Particularly, in catchment distributed
modelling using grid DEMs, research has enabled to recom-mend
the use of DEM grid sizes smaller than 50 m for ade-quate
flow pathway analysis at the hillslope scale (Saulnier
et al., 1997a; Beven and Freer, 2001). Thus, using the dis-tributed
code TOPMODEL (Beven et al., 1995), Zhang and
Montgomery (1994) selected a 10-m grid size for the ade-quate
simulation of geomorphic and hydrologic processes
in two small catchments (0.3-km2 and 1.2-km2). Beldring
(2002) used a 10-m DEM for modelling a 6.2-km2 catchment.
Braud et al. (1999) modelled a 5.47-km2 mountainous catch-ment
with the ANSWER code (Beasley et al., 1980) using a
30-m grid size. Gu¨ntner et al. (1999) applied TOPMODEL
on a well-monitored 40-km2 catchment considering a 50-m
grid size.
The use of DEM grid sizes smaller than 50 m is however
not always possible in catchment distributed modelling.
This is in part due to the lack of world-wide data with the
appropriate resolution. Other important reason is related
to computational efficiency, which is sensitive to the num-ber
of horizontal and vertical (modelling) computational
units and, as such, to the size of the modelled catchment.
In this respect, Xevi et al. (1997) and Christiaens and Feyen
(2002) modelled a 1-km2 well-monitored experimental
catchment with the code MIKE SHE (Refsgaard and Storm,
1995) considering a grid size of 100 m. Refsgaard (1997)
and Madsen (2003) considered grid sizes larger or equal to
500 m for modelling a 440-km2 catchment. Refsgaard and
Knudsen (1996) modelled a 1090-km2 catchment with a
1000-m grid size using MIKE SHE and Jain et al. (1992) mod-elled
with the same hydrological code the 820 km2 Kolar
catchment in India with grid sizes ranging from 500 to
4000 m.
The use of such coarse grid sizes in catchment distrib-uted
modelling implies important spatial scale differences
among the scale to which the physical structure of the
hydrologic codes were obtained, the scales to which the dif-ferent
data are collected and the coarse scales to which the
hydrologic codes are applied (Bergstro¨m and Graham, 1998;
Va´zquez et al., 2002; Va´zquez, 2003). The following are
therefore important issues that are related to the impact
of grid scale on the predictions of catchment modelling:
(i) what is the adequate grid resolution for achieving
accurate model predictions, while keeping computa-tional
times under reasonable limits?
Prior sensitivity analyses demonstrated that using (more
or less) different data for the same modelling variable
lead to significant differences in both effective parame-ter
values and model performance (Va´zquez et al., 2002;
Va´zquez and Feyen, 2003b).
(ii) Given that geomorphologic features intrinsic in the
DEMs (i.e. elevation, slope, curvature, etc.) are impor-tant
for the simulation of flow processes such as surface
runoff, infiltration and evaporation, and provided that
different DEM accuracies are expected from the applica-tion
of different DEM gridding methods, do the effective
parameter values reflect the differences of these DEM
generation methods when using a coarse modelling
resolution?
(iii) Is the adequacy of global predictions affected by dif-ferent
DEM generation methods? and
(iv) Is the evaluation of internal state predictions
affected by the DEM generation methods?
The assessment of the first of these grid-scale issues will
demand the consideration of various aspects such as param-eter
error, model structural error and data (input and eval-uation)
measurement error. In this context, Va´zquez et al.
(2002), after using 300, 600 and 1200-m modelling grid
sizes, found that an acceptable compromise between accu-racy
of model predictions and computational (i.e. running)
time was reached when using a grid size of 600 m for the
modelling of the Gete catchment (Belgium) with the MIKE
SHE model. This study did not consider model structural er-ror
owing to the lack of access to the structure of the MIKE
SHE model (access limitations linked to the commercial nat-ure
of the software). However, the main conclusions of the
referred study were based on parameter calibration, the
evaluation of internal state predictions and a brief assess-ment
of data measurement error concerning piezometric
data (for evaluation).
With regard to the other grid-scale issues, previous stud-ies
have used topographically driven codes such as TOPOG
(Vertessy et al., 1993) and TOPMODEL for examining the ef-fects
of both the scale of the input elevation data and the
resolution of the gridded DEMs on model performance

Assessment of the effects of DEM gridding on the predictions of basin runoff using MIKE SHE 75
(Zhang and Montgomery, 1994; Wolock and Price, 1994;
Lane et al., 2004) and on the effective values of saturated
hydraulic conductivity (Saulnier et al., 1997a,b). However,
published modelling studies using MIKE SHE with coarse
DEMs have not addressed explicitly either of these topics
(Refsgaard and Storm, 1995; Refsgaard and Knudsen, 1996;
Xevi et al., 1997; Refsgaard, 1997; Feyen et al., 2000; Chris-tiaens
and Feyen, 2002; Madsen, 2003). Furthermore, dis-cussion
about the incidence of different gridding methods
on either the quality of the DEM products, the model global
predictions, the evaluation of internal state predictions or
the effective parameter values of the hydrologic model is
not common in previous publications about distributed mod-elling
of catchments.
Therefore, in contrast to previous published work, this
article presents an assessment of different methods for
gridding coarse-DEMs (up-scale gridding) and the potential
effects of these methods on the referred grid-scale issues,
namely, adequacy of global predictions, effective parame-ter
values and the evaluation of internal state predictions.
In line with the conclusions of previous work (Va´zquez
et al., 2002), the current assessment involves the use of a
600-m grid size. The assessment is furthermore based on
the application of the MIKE SHE model on the data of the
Gete catchment (Belgium).
Materials
The study site
The study site, the Gete catchment (586 km2), located to
the east of Brussels-Belgium (Fig. 1), comprises the sub-catchments
of the Grote Gete (326 km2) and the Kleine Gete
(260 km2). The elevation of the area varies from approxi-mately
27 m in the northern part to 174 m in the southern
part. Land use is mainly agricultural with some local for-ested
areas. The local weather is characterised by moderate
humid conditions. Nine soil units can be distinguished
according to the legend of the Belgian soil map (Vander
Poorten and Deckers, 1994; Va´zquez, 2003), e.g. loamy soils
(Aba, Ada and Adc), sand–loamy soils (Lca, Lda and Ldc),
clay soils (Eep and Uep) and soils with stony mixtures
(Gbb). The dominant soil type in the catchment is the Aba
soil unit. The reader is referred to Va´zquez et al. (2002,
2003) and Va´zquez and Feyen (2003b) for additional details
about the description of the catchment.
The hydrologic code
The MIKE SHE code (Refsgaard and Storm, 1995) was consid-ered
for the integral modelling of the study site. MIKE SHE is
a well-known deterministic-distributed code that has been
used and described in a wide range of applications (e.g.,
Refsgaard, 1997; Jayatilaka et al., 1998; Feyen et al.,
2000; Christiaens and Feyen, 2002; Va´zquez and Feyen,
2003a). MIKE SHE integrates the entire land phase of the
hydrologic cycle and can model interception, actual evapo-transpiration
(ETact), overland flow, channel flow, flow in
the unsaturated zone, flow in the saturated zone and ex-change
between aquifers and rivers. MIKE SHE applied at a
catchment scale implies the assumption that smaller scale
equations are valid also at the larger scale; thus, it performs
an upscaling operation using effective parameter values.
The MIKE SHE model uses a square-grid modelling structure.
Consequently, square-grid DEMs were used in this research.
The input elevation data
The DEMs were gridded from a set of elevation spot heights
(Zsource), available from the Flemish Spatial Data Infrastruc-ture
(OC-GIS Vlaanderen-Belgium). There are no published
details about the methods used to create these elevation
spot heights. However, according to staff at the National
Geographical Institute of Belgium (NGIB), these data were
derived originally from digitised contour lines from a
1:50000 topographic map and finally arranged in a non-orthogonal
but regular grid mesh with an approximate reso-lution
of 40 m in the X-direction and 30 m in the Y-direction.
Q-Gete
Q-Kleine Gete Catchment
outlet
Main flow
direction
North sea The
Gete
Multi-Site observation well
Multi-Site stream station
Split-Sample observation well
Split-Sample stream station
Coordinates: simulation grid
5 10 15 20 25 30 35 40 45 50 55 (600x600 m² model)
55
50
45
40
35
30
25
20
15
10
5
Q-Grote Gete
Legend:
N
Flemish region
Walloon region
France
Netherlands
catchment
Figure 1 Location of the study site and distribution of the calibration and evaluation wells and stream stations (after Va´zquez and
Feyen, 2003a).

Inspection of the numerical characteristics of Zsource re-vealed
that the information was recorded to the nearest
meter. This constitutes by itself an important factor of
uncertainty associated with Zsource that should be taken into
account when assessing the elevation quality of the gridded
DEMs, as owing to this uncertainty, Zsource is not a true rep-resentation
of the topography of the study site. Further-more,
digitised streams and the topographical catchment
divide, as derived from the 1:50000 scale maps, were
available.
Methods
DEM gridding methods
The selection of a gridding method depends principally on
the spatial distribution of the input data and resolution of
the output grid. In common hydrologic-model applications,
DEMs are normally gridded by means of interpolation algo-rithms
for estimating elevations, so that they have a resolu-tion
that is finer than or similar to the average resolution of
the input data (i.e. down-scaling operation). The use of
coarse grid DEMs for hydrologic purpose demands, on the
contrary, an up-scaling gridding operation to generate DEMs
with a resolution that is much coarser than the average res-olution
of the input data.
In this work the spot elevation data were gridded accord-ing
to five up-scaling methods. The first method ((A)) uses
the spot-based Bilinear (Bi) interpolation algorithm, avail-able
as a pre-processing MIKE SHE tool (DHI, 1998). This
interpolation tool uses up to a maximum of four points
(one per quadrant) to estimate the elevation at every grid-ded
cell corner. The points are the nearest to the cell cor-ner
and are selected on the basis of (i) their distance to the
corner and (ii) a user-defined searching radius around the
corner. The method then estimates, the centre value for
every cell as the average of the four corner values. There-fore,
depending on both the density of the elevation data
and the modelling resolution, up to a maximum of 16 data
points may contribute to the estimation of the cell centre
value. Considering the approximate resolution of Zsource
(i.e. 40 m · 30 m), the searching radius was given a value
(much) greater than 50 m ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð40Þ2 þ ð30Þ2
q
m, that is, long
enough to use 16 elevation points. For DEM coarse resolu-tions,
this up-scaling method uses information from points
located around the gridded cell corners, that is, far away
from the cell centre where the elevation is finally derived.
The other four approaches, methods (B), (C), (D) and (E),
were implemented with the aid of algorithms that are avail-able
for gridding DEMs in common geographical information
system (GIS) packages, namely, ARC/INFO (ESRI, 1996) and
IDRISI (Clark Labs, 1998). These methods are commonly
used for downscaling operations, but they are not necessar-ily
the most appropriate to carry out the up-scaling gridding
of coarse DEMs. Therefore a systematic evaluation of their
DEM-products is needed. For methods (B) and (C) a TIN sur-face
was produced with ARC/INFO on the basis of a linear
interpolation process. In method (B), the TIN surface was
further interpolated into a regular lattice (TIN to lattice
transformation) with a 600-m resolution. In general, this
transformation involved a linear interpolation along the
edges of the TIN triangles (positioned around the cell cen-tre)
to determine the elevation of the lattice pixels. Thus,
unlike method (A), method (B) uses input elevation data
concentrated around the cell centre where the elevation
is estimated. In method (C), contour lines were derived with
ARC/INFO from the TIN surface. These contours were then
utilised as elevation input for the MIKE SHE Bi interpolation
utility. Thus, alike method (A), method (C) uses information
from elevation spot heights located at the periphery of the
gridded cell.
For methods (D) and (E), the ARC/INFO module TOPO-GRID
(ESRI, 1996) was used. TOPOGRID is a finite difference
interpolation technique that is based on the specialised
interpolation approach ANUDEM (Hutchinson, 1989) that is
characterised by its computational efficiency, so that it
can handle large data sets, and by a drainage enforcement
algorithm that uses landscape features, such as digitised
streams, to improve the structure of DEM products for
hydrologic purposes. TOPOGRID imposes interpolation con-straints
to remove spurious sinks that result in a more cor-rect
drainage structure and representation of ridges and
streams. The use of TOPOGRID is therefore considered as
one of the best current practices in DEM gridding, as it gen-erates
a hydrologically consistent DEM (ESRI, 1996; Hutchin-son,
1989). The readers are referred to Hutchinson (1989),
Hutchinson and Dowling (1991) and ESRI (1996) for a com-plete
description of ANUDEM and TOPOGRID.
In method (D) a 20-m DEM was initially obtained through
a common (i.e. downscaling) TOPOGRID application. Re-sampling
methods were then applied on the 20-m DEM for
obtaining coarse 600-m DEMs, namely, the Nearest Neigh-bour
(Nn), the Bilinear (Bi) and the Cubic Convolution (Cu)
methods (ESRI, 1996; Hanselman and Littlefield, 1998).
The Nn algorithm assigns the value associated with the clos-est
cell centre on the input grid to the re-sampled cell. The
Bi interpolation algorithm uses a weighted average deter-mined
by the values of the input cells at the four nearest
cell-centres and their weighted distance to the centre of
the re-sampled cell. Cu is a surface-fitting algorithm that
uses the 16 nearest input cell centres and their values to
determine the re-sampled cell value. Depending on the res-olution
of the input grid and similarly to method (B), these
re-sampling methods use elevation data concentrated
around the coarse cell centre. The three DEM products of
the re-sampling methods were compared with each other
for selecting a DEM product representative of method (D)
to be used in the hydrologic modelling. This assessment
was based on the analysis of the cumulative distributions
of DEM elevations, slopes, etc, as explained later in the
text. The assessment of DEM elevation quality indicated a
marginal difference in detriment of the product of the Nn
re-sampling technique (DEM(D_Nn)) and in favour of the
product of the Cu re-sampling technique (DEM(D_Cu)).
Therefore, DEM(D_Cu) was selected as representative of
method (D) for the hydrologic modelling. This selection
was also supported by the fact that more input data (16 in-put
cell centres) are included in the Cu method than in the
other re-sampling methods for determining every re-sam-pled
cell centre.
For method (E), the coarse 600-m DEM was obtained di-rectly
from Zsource without the need of a re-sampling meth-od
through the application of TOPOGRID, within an up-

scaling context. In contrast to the other methods, TOPO-GRID
does not use input elevation data exclusively concen-trated
either about the centre or the periphery of the
gridded cell, but instead, it uses simultaneously distributed
elevation data and stream information to enforce a more
natural drainage structure in the DEM products.
Thus, depending on whether the gridding method uses in-put
elevation data distributed about the gridded cell centre
or about the gridded cell periphery, the gridding methods
are further classified in this work as belonging to class type
I (i.e. methods (A) and (C)), class type II (i.e. methods (B)
and (D)) and class type III (method (E)). In the forthcoming
sections of this paper, the quotation ‘‘DEM(method/type)’’
stands for the DEM product of a particular gridding method
or type.
DEM quality assessment
The assessment of the quality for hydrologic use of the
coarse DEM products was based on (i) the comparison of
the DEMs and Zsource, (ii) the comparison of the DEMs with
each other through their hydro-geomorphic properties,
and (iii) the analysis of plots such as spatial distribution of
pits, drainage patterns, derived contours, etc. The set of
hydro-geomorphic properties comprised: drainage patterns,
catchment areas, slopes and hillslope shading and the com-putation
of the topographic index (k). These analyses were
done considering the domain defined by the digitised catch-ment
boundary.
The discrepancies between the coarse DEM elevations
(ZDEM) and the input elevation data (Zsource) were condensed
into summary statistics, such as minimum, mean and maxi-mum
values, standard deviation and the root mean squared
error (RMSE)
RMSE ¼
ffiPffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi n
i¼1ðresÞ2
i
n
s
ð1Þ
where (res)i, is the residual, the arithmetic difference be-tween
a coarse DEM elevation (ZDEM) and Zsource at the i-th
test location of interest; and n is the number of test loca-tions.
The RMSE analysis included 150 test-spots that were
defined pseudo-randomly, that is, 120 spots were selected
randomly whilst the remaining 30 spots were chosen subjec-tively
to include terrain zones of hydrologic and geomorpho-logic
interest, such as depressions, spots near water courses
and uplands. The comparison of the coarse DEMs and Zsource
included also the analysis of the cumulative distributions of
residuals computed at the same test locations as in the
RMSE evaluations. Furthermore a cumulative distribution
of DEM elevations was calculated for every DEM product.
The derivation of drainage patterns and the automatic
delineation of catchment (i.e. drainage) areas are affected
by spurious sinks (i.e. local minima), as sinks interfere with
the (natural) connection of drainage patterns. Thus, sink re-moval
was carried out prior to the assessment of these geo-morphic
relationships. Because zones of natural depression
storage are very difficult to account while using coarse res-olutions,
such as the one considered in this study, all the
identified sinks were treated as artificial depressions result-ing
from random errors induced by the gridding interpola-tion.
Thus, all of the identified sinks were removed. In
doing so, a method based on the drainage enforcement
algorithm by Jenson and Domingue (1988) was applied. With
this algorithm, the downslope flow direction for each cell is
defined after inspecting the eight cardinal directions and
identifying the neighbouring cell of lowest elevation. The
algorithm fills the sinks by assigning to them the elevation
of their lowest neighbour and then assigns flow directions
to these flat cells towards a neighbouring cell that has a pre-viously
assigned flow direction.
The Jenson and Domingue (1988) algorithm was also used
to derive the drainage patterns, accounting for the flow that
accumulates at each cell of the DEM for a uniform and unit-value
rainfall. A threshold value (10 flow units) was then ap-plied
to produce the drainage networks through a reclassifi-cation
process that accounted for cells with higher
accumulated flow. The catchment areas were automatically
delineated also through the Jenson and Domingue (1988)
algorithm by using the information on the flow direction
for each cell and the grouping of cells draining to a single
outlet or seed cell. This seed pixel was defined explicitly
matching the coordinates of the Gete station (Fig. 1). The
derived catchment areas were compared to the digitised
catchment boundary.
Although some of the assumptions on which the interpre-tation
of the topographic index (k) is based are not fulfilled
(mainly because of the coarse resolution used in this
research), it was calculated to briefly assess the potential
runoff sources (Beven et al., 1995; Quinn et al., 1995;
Ambroise et al., 1996). For a particular grid cell, it is calcu-lated
as
ki ¼ ln

ai
tanðbiÞ
ð2Þ
where ai is the area draining through the cell per unit length
of contour [L]; and tan(bi) is the local surface slope [–] of
the cell. The multiple direction flow sharing algorithm by
Quinn et al. (1991) was used to determine the downslope
flow pathways and for distributing to the downslope cells
a proportion of the accumulated contributing area. This
algorithm is more suitable than that of Jenson and Domin-gue
(1988) for representing flow on divergent hillslopes
and for larger grid resolutions (Quinn et al., 1991, 1995).
The calculation of k enabled to compare the drainage
patterns derived with both methods, namely Quinn et al.
(1991) and Jenson and Domingue (1988).
Sensitivity analyses on the coarse DEMs
The grid-scale issues of interest defined in the objectives of
this manuscript were examined through a Multi-Calibration
(MCal) test, in which several models differing only in the
DEM input data were subjected to identical calibration
and evaluation. After model calibration, the differences in
the sets of effective parameter values were assessed.
During model evaluation, the differences in model perfor-mance
were characterised through the analysis of the model
residuals (i.e. differences between predictions and
observations).
In addition, an analysis of the effects of the sink removal
operation was performed considering the DEM(A) represent-ing
gridding type I and DEM(B) representing gridding type II.

Additionally, this analysis was extended to include DEM(C)
for investigating the modelling consequences of the mis-match
that was observed between the derived catchment
outlet and the location of the Gete station, which is de-picted
later in the text.
Hydrologic model of the study site
The profile definition of the river tributaries was based on
interpolation/extrapolation of a few measured profiles.
Drains were specified in the model set-up to improve the
simulated hydrograph shape and to account for the small ca-nals
and ditches present on a scale smaller than the model-ling
resolution. The drainage depth (zdr) and the reciprocal
time constant (Tdr), a sort of drainage coefficient, were cal-ibrated
because they influence mainly the velocity of the
drainage and the peak and recession of the hydrograph.
The spatial extent of the soil units and their vertical proper-ties
could be assessed considering two soil databases (Feyen
et al., 2000). Parameters for describing the flow through the
soil system were calculated with pedo-transfer functions
(PTFs). Despite the uncertainties associated with the PTFs
and the soil databases, the soil parameters were not in-cluded
within the calibration process to avoid considering
too many parameters (for 9 soil units) during model
calibration.
The complex geology of the studied system comprises
nine geological units (Va´zquez and Feyen, 2003a), of which
only two are underlying completely the area of the catch-ment.
The geology was incorporated in the three-dimen-sional
groundwater model of MIKE SHE. A sensitivity
analysis demonstrated that the model could be simplified
further to six geological units without influencing the global
results noticeably (Va´zquez et al., 2002). The model in-cludes
five upper geological units on top of the low-perme-able
Palaeozoic rocky basement. The hydrogeologic
parameters of these units were tuned during the calibration
process. Owing to the lack of appropriate measurements,
the groundwater divide was assumed coincident with the
topographical divide and the aquifers were given no-flow
boundary conditions.
The input time step, during which no change in boundary
conditions occurs (stress period), was taken as one day, in
recognition of the lack of more precise meteorological data.
MIKE SHE requires crop potential evapotranspiration (ETp)
data for modelling ETact. Modelling ETact was done by means
of the Kristensen and Jensen (1975) approach. ETp data
were estimated in turn by means of the Kc–ET0 method that
uses crop coefficients (Kc) and crop reference evapotranspi-ration
(ET0). ET0 time series were estimated with the Food
and Agriculture Organisation (FAO) Penman-method 24
(FAO-24). Locally-applicable parameter values were avail-able
for the components of the FAO-24 method (Va´zquez
and Feyen, 2004). The estimates produced by the FAO-24
method, using these locally-applicable parameter values
for the various components of the method, were equivalent
to the estimates of the FAO-56 Penman-Monteith method
using standard parameter values recommended by the
FAO-56 report for the different components of the method
(Allen et al., 1998). No locally-applicable parameter values
were available for the components of the FAO-56 method
(Va´zquez and Feyen, 2004). The effective values for the
parameters that MIKE SHE uses to estimate ETact (Kristensen
and Jensen, 1975) were taken from a previous sensitivity
analysis (DHI, 1998; Va´zquez and Feyen, 2003b). These
parameter values were kept constant throughout the mod-elling
analysis. The reader is referred to Doorenbos and Pru-itt
(1977), Va´zquez and Feyen (2003b, 2004) for a complete
description of the Penman FAO-24 approach and the set-up
of the catchment model.
Model calibration and performance assessment
In principle, when sufficient data are available, physically
based distributed models do not need to be calibrated.
However, this type of models must be calibrated to improve
their predictions because of sub-grid variability of parame-ter
values, model structure uncertainties and data (input
and evaluation) uncertainties. A good calibration process
should aim therefore to reduce as much as possible the
model error, that is, parameter uncertainties (i.e. obtaining
appropriate grid-scale effective parameter values) and
model structure uncertainties (i.e. developing the most
accurate catchment model) so that the total modelling
error is mainly composed by the data measurement error
(which is usually unknown and scale dependant). With this
purpose in mind, and in line with their nature, distributed
models must be evaluated against distributed
measurements.
In this study, care was taken to avoid violating physical
constraints during the model calibration (Va´zquez and
Feyen, 2003a). Particularly, the availability of piezometric
data defined the calibration period as from the 1st of Janu-ary
1985 until the 31st of December 1986 and the main eval-uation
period from the 1st of January 1987 until the 31st of
December 1988 (Split-Sample test). Additional evaluation
periods of variable length were however considered in the
scope of a Multi-Window (MW) test (Va´zquez, 2003; Va´zquez
and Feyen, 2003b), as depicted later in the text.
Owing mainly to the high computational requirements
and the large number of model parameters, the calibration
of distributed hydrologic models is not a trivial activity. In
this research, a considerable computational work was re-quired
since every simulation lasted about one hour and
an average of 400 simulations were needed to reach every
model calibration in the framework of the Multi-Calibration
(MCal) test. A conventional calibration by fit process was ap-plied.
It was based on a ‘‘trial and error’’ procedure, in
which the influence of the various model parameters was
examined step by step through a multi-criteria performance
protocol. In general, for each DEM product the model was
calibrated and evaluated using a Split-Sample (SS) proce-dure
against basin-wide daily discharge measurements and
water levels for 12 observation wells, with screens in differ-ent
geological layers (Fig. 1). To investigate how well the
calibrated model was able to simulate internal variables, a
Multi-Site (MS) evaluation test was also performed for two
internal discharge stations and 6 observation wells that
were not considered during the calibration process (Fig. 1).
The general protocol used for assessing the model per-formance
consisted of two complementary components:
(i) the analysis (statistical and graphical) of different time
series properties; and (ii) the evaluation of a set of multi-objective
statistics. The inspected time series simulation

properties included: (i) cumulative flow volumes; (ii) mod-elled
versus observed discharge maxima; and (iii) high flow
extreme value statistics.
Unrelated or slightly correlated statistics were preferred
for the set of multi-objective statistics. However, as an
exception, two statistics that are correlated were consid-ered
in this research for getting a quick overview of the
model performance simulating peak flows. These statistics
are the coefficient of efficiency (EF2) (Legates and McCabe,
1999), that is frequently used for an estimation of the total
(combined systematic and random) average error and the
coefficient of determination (CD) (Loague and Green,
1991) that is related to the EF2 but is particularly useful
to assess the simulation of peak values (Va´zquez, 2003).
These statistics are defined as follows:
EF2 ¼ 1
Pn
i¼1ðOi PiÞ2
Pn
i¼1ðOi OÞ2 ð3Þ
CD ¼
Pn
i¼1ðOi OÞ2
Pn
i¼1ðPi OÞ2 ð4Þ
where, Pi is the i-th simulated value, Oi is the i-th observed
value, O is the average of the observed values and n is the
number of observations. The optimal value of EF2 is 1.0
and the feasible range of variation is 1 EF2 6 1.0, while
for CD, these parameters are 1.0 and 0.0 CD +1, respec-tively.
For a more even assessment of the simulation of both
high and low flows, the multi-objective set of statistics was
also applied on the logarithmic transformation of the ob-served
and the predicted variables (Va´zquez, 2003).
Primarily, the drainage depth (zdr) and the reciprocal
time constant (Tdr) were calibrated against the overall dis-charge
of the catchment. Next, the values of the hydrogeo-logic
parameters of the five upper most geological units
were tuned against the outflow discharge of the catchment.
When the overall discharge was reasonably well simulated,
the hydrogeologic parameters of each unit were tuned fur-ther
to improve the agreement between the predicted and
observed piezometric levels in the calibration wells. Previ-ous
modelling experiences (Feyen et al., 2000; Va´zquez
et al., 2002; Va´zquez and Feyen, 2003b) showed that after
tuning the hydrogeologic parameters for improving the pie-zometric
predictions, the model performance simulating
overall discharges was diminished and, as a consequence,
an additional tuning of zdr, Tdr and the horizontal conductiv-ity
of the loamy Quaternarian (Kw) and the clayey sand
Landeniaan (Ln) units was necessary for improving the sim-ulation
of the overall discharge. Thus, in this research a fur-ther
tuning of these parameters was carried out to improve
as much as possible the prediction of the overall discharge
but trying at the same time to reach a modelling compro-mise
to avoid affecting significantly the prediction of piezo-metric
levels.
For the extreme value analysis (EVA), independent Peak-
Over-Threshold (POT) values were extracted from the total
discharge (Qt) using independency criteria based on the dif-ferences
among the recession constants of the hydrologic
subflows: overland flow (Qov), interflow (Qin) and baseflow
(Qbs) (Willems, 2000; Va´zquez and Feyen, 2003a). The EVA
was performed from the 1st of January 1984 until the 31st
of December 1995. An exponential distribution fitted rea-sonably
well the observed data beyond an optimal threshold
Qt equal to 6.4 m3 s1 (Va´zquez and Feyen, 2003a).
Results
DEM quality assessment
The analysis of the cumulative distributions of DEM eleva-tions
and the elevation statistics did not reveal a clear dif-ference
among all of the inspected DEMs. Fig. 2a shows the
cumulative distributions of slope (tan(b)) as a fraction of
the catchment area and as a function of the DEMs after sink
removal. The plot also includes the corresponding distribu-tion
for the 20-m product of the TOPOGRID algorithm,
considered in this research as hydrologically consistent
(Hutchinson, 1989; ESRI, 1996) and, as such, as a reference.
Fig. 2b illustrates the cumulative distributions (Fr(res 6
RES)) of residuals for a particular value of interest (RES),
calculated at the 150 pseudo-random test locations as a
function of the DEMs after sink removal.
Fig. 2a shows the considerable smoothing of the DEMs
that took place as a consequence of increasing the resolu-tion
of the elevation data (from about 20 m to 600 m) and
as such indicates the generalised low elevation quality of
all of the coarse DEMs used in this research. Nevertheless,
Fig. 2a and b show that the products of methods (B) and
(D) described the catchment’s elevation slightly better than
the DEM products of methods (A), (C) and (E). Additionally,
1.0
0.8
0.6
0.4
0.2
0.0
0
DEMs after sink removal
RES [m]
Fr(resRES) [--]
(A)_rand
(B)_rand
(C)_rand
(D)_rand
(E)_rand
1.0
0.8
0.6
0.4
0.2
0.0
0
tan( ) [--]
Fraction of catchment area [--]
20-m
(A)
(B)
(C)
(D)
(E)
1 2 3 4 5 4 8 12 16
Figure 2 Cumulative frequency distributions of (a) DEM local slope; and (b) residuals among DEMs and pseudo-random elevation
spot heights (_rand).

these figures clearly show three different sets of cumulative
distributions. One of the sets groups the DEM products of
methods (A) and (C). A second set groups the DEM products
of methods (B) and (D). Finally, the DEM product of method
(E) constitutes a third set. This is consistent with the three
different types (I, II and III) of DEM gridding methods that
were used in this work and indicates that these DEM gridding
methods affected the adequacy of DEM elevations.
The outcomes of methods (B) and (D) (gridding type II) in-cluded
greater numbers of sinks than the outcomes of meth-ods
(A) and (C) (gridding type I). No sink was identified in the
outcome of method (E), because sinks were automatically
removed by TOPOGRID (ESRI, 1996). Visual inspection of
the spatial distribution of sinks indicated that the coarse
DEM gridding methods (in particular methods (B) and (D))
did not rightly model the transition from the hillslopes to
the floodplains in the regions where depressions occurred.
Given the limitations for describing appropriately the
topography of the study site associated with the 600-m res-olution,
it may be concluded that the drainage networks de-rived
through the Jenson and Domingue (1988) approach
agreed acceptably well with the digitised watercourses.
However, some considerable differences of detail, such as
artificial branch discontinuities were noticed (Fig. 3a).
Fig. 3b shows that, particularly for DEM(C), the outlet of
the derived drainage network does not match the digitised
outlet (i.e. Gete station). This is likely to affect the perfor-mance
of the respective hydrologic model because the sim-ulated
discharges are evaluated at the location of the Gete
station rather than at the derived catchment outlet.
Furthermore, owing to this mismatch, the derived catch-ment
area for DEM(C) was much smaller than the digitised
boundary despite the sink removal pre-processing. As a
way of mitigating this mismatch and ensuring a drainage
area covering the extent of the digitised boundary, an addi-tional
smoothing of the coarse DEM(C) was performed with a
3 · 3 mean filter (Clark Labs, 1998). As a consequence, the
drainage area for DEM(C) was significantly improved, since
the location of the predicted catchment outlet was shifted
to the position of the Gete station but to the cost of losing
information after smoothing (Fig. 3c). Thus, the effects of
the additional smoothing of DEM(C) on the hydrologic pre-dictions
were investigated by means of a sensitivity analysis
considering the smoothed DEM into the hydrologic model
set-up. The model was parameterised using the effective
parameters that were obtained with the DEM prior the
smoothing process.
The analysis of the spatial distributions of k revealed an
acceptable agreement between the higher k values and the
digitised river network. In this respect, the k distributions
(Quinn et al., 1995) were consistent with the drainage pat-terns
derived through the Jenson and Domingue (1988) ap-proach,
even with the situation depicted in Fig. 3b for
DEM(C). Furthermore, this analysis suggested a smoother
generation of runoff for the DEM products of methods (A),
(C) and (E) with respect to the outcomes of methods (B)
and (D), which is in agreement with their slope characteris-tics
(cf. Fig. 2a). The cumulative distributions of k evolved
as a function of the three different types (I, II and III) of
gridding methods inspected in this work. This study in-spected
the cumulative distributions of the two components
of k, namely ln(1/tanb) accounting for land–surface slope
and ln(a) accounting for land–surface shape. The distribu-tions
of ln(1/tanb) were consistent with the slope distribu-tions
shown in Fig. 2a. The distributions of ln(a) indicated
no special concentrations of either convex or concave land-scape
features in the DEMs.
Hydrologic modelling
The results of the hydrologic modelling are presented with
regard to the main grid-scale issues that are addressed in
this article.
Do the sets of effective parameter values reflect the dif-ferences
of the DEM generation methods?
Table 1 lists the main effective parameter values in rela-tion
to the three types of DEM gridding methods. Further-more,
the parameters are classified in two main groups
with respect to the presence of artificial sinks in the DEMs.
The table lists only the effective values of the loamy Qua-ternarian
and the clayey sand Landeniaan layers, which
have a considerable influence on the modelling of the
groundwater flow, as well as the aquifer–river interaction.
During the model calibration, the hydrogeologic parameters
1.0
0.8
0.6
0.4
0.2
0.0
0.0 0.5 1.0 1.5 2.0 2.5
tan( ) [--]
Fraction of catchment area [--]
(C)
(C)_mf3
DEM(A) after sink removal
Legend:
N
DEM(C) after sink removal
Significant difference with respect to the digitised streams
DEM(C) after sink removal
Derived catchment Gete station
Digitised stream network
DEM-derived stream network
Catchment boundary
outlet
Figure 3 Maps showing (a) the spatial distribution of derived drainage network for DEM(A); and (b) the mismatch between the
location of the predicted catchment outlet and the position of the Gete station. (c) Cumulative frequency distributions of DEM(C)
before and after (C_mf3) smoothing through the mean filter (3 · 3 pixels smoothing mask).

Table 1 Main effective parameters in relation to the DEM generation methods
DEM properties Model parameter Geological unit Coarse-DEM gridding method
Type I Type II Type III
A C B D E
After sink removal (DEMsNOsink) zdr (m) 0.46 0.30 0.43 0.43 0.50
Tdr (s1) 6.20 · 108 1.55 · 107 6.80 · 108 8.50 · 108 5.90 · 108
Ksat (m s1) Kx (m s1) Quaternarian 7.70 · 107 8.00 · 106 2.00 · 107 2.30 · 107 2.00 · 107
Landeniaan 6.00 · 106 5.50 · 106 5.50 · 106 6.00 · 106 6.75 · 106
Kz (m s1) Quaternarian 4.24 · 107 4.80 · 106 9.00 · 108 5.75 · 108 9.40 · 108
Landeniaan 1.50 · 106 1.54 · 106 3.58 · 106 3.90 · 106 4.39 · 106
Sy (–) Quaternarian 0.17 0.21 0.20 0.22 0.23
Landeniaan 0.41 0.41 0.39 0.43 0.41
Including artificial sinks (DEMssink) zdr (m) 0.40 0.20 0.22
Tdr (s1) 7.00 · 108 1.65 · 107 1.30 · 107
Ksat (m s1) Kx (m s1) Quaternarian 2.00 · 106 4.00 · 106 1.00 · 107
Landeniaan 7.00 · 106 9.00 · 106 9.30 · 106
Kz (m s1) Quaternarian 1.74 · 106 2.40 · 106 1.00 · 107
Landeniaan 1.19 · 106 1.80 · 106 3.07 · 106
Sy (–) Quaternarian 0.20 0.20 0.20
Landeniaan 0.19 0.34 0.30
zdr = Drainage level, Tdr = Reciprocal time constant, Ksat = Saturated hydraulic conductivity, Kx = Horizontal hydraulic conductivity, Kz = Vertical hydraulic conductivity, Sy = Specific yield.

were allowed to vary spatially considering factors such as
the extent of the main sub-catchments, the extent of the
geologic units and the location of the abstraction wells
(Va´zquez and Feyen, 2003a). For a particular modelled geo-logic
unit, the values listed in Table 1 correspond to the
effective zone with the lowest hydrogeologic parameter
value.
In broad terms, the assessment of the sets of effective
parameter values obtained using the DEMs after sink re-moval
revealed that only an average variation of parameter
values took place as a function of the three types of DEM
gridding methods. The set associated with DEM(C) has how-ever
noticeably different values for zdr (lower absolute va-lue),
Tdr (higher value) and the saturated hydraulic
conductivity (Ksat) of the Quaternarian (Kw) layer (higher va-lue),
parameters that are related to the simulation of sur-face
processes within the modelled hydrologic system.
The reason for obtaining this different parameter set is
likely to be linked to the mismatch between the predicted
catchment outlet and the digitised catchment outlet (Gete
station, cf. Fig. 3b) where the model prediction was finally
evaluated.
When the DEMs including sinks were considered in the
MCal analysis, it was observed however a noticeable influ-ence
of the DEM gridding methods on the values adopted
by the effective parameters of the hydrologic models.
These effects were especially important with respect to
zdr, Tdr and Ksat–Kw (c.f. Table 1).
Looking at the effect of the sink removal operation (i.e.
smoothing operation) on the sets of effective parameters
values, Table 1 illustrates that, in general, higher zdr abso-lute
values and lower Tdr values (i.e. higher drainage veloc-ities)
are associated with the (smoother) DEMs after sink
removal. This is for opposing to the smoothing (i.e. flatter-ing)
of the DEMs by routing higher overland and interflow
water volumes more quickly (higher zdr absolute values)
but, at the same time, controlling the magnitude of peaks
through lower Tdr values. In some cases the acceleration
of flow routing is accentuated by higher (horizontal and ver-tical)
values of Ksat–Kw, such as for the products of the DEM-methods
(A) and (B). For the Landeniaan (Ln) layer, which is
the most influential to groundwater flow and the aquifer–
river interchange flow, the MCal analysis indicated that
comparable values of Ksat were obtained for all the models
independently of whether the DEM products included sinks.
Thus the effects of the sink removal smoothing were re-flected
principally on the variation of zdr, Tdr and Ksat–Kw.
Is the adequacy of global predictions affected by differ-ent
DEM generation methods?
Fig. 4 shows the observed and calibrated hydrographs for
the DEM products of the gridding approaches type I
(DEM(A)), type II (DEM(B)) and type III (DEM(E)). The figure
shows that, in general, the models have certain difficulties
for rightly simulating the recession limbs and the subse-quent
baseflow, especially in the periods January of 1985,
February–March of 1986 and June–October of 1986. In gen-eral,
the models tended to overestimate the peakflow
events. However, in broad terms, the analysis of the cali-brated
time series of total discharge revealed that the mod-els
related to the DEM products of gridding methods type II
((B) and (D)) and type III ((E)) predicted better discharge
series than the other models, regardless of the presence
0
10
20
30
40
50
60
70
80
90
100
21
17
13
9
5
1
Rainfall
DEM(A)
Observed
Jan-85 May-85 Sep-85 Jan-86 May-86 Sep-86 Jan-87
Rainfall [mm day -1]
Discharge [m3 s-1]
0
10
20
30
40
50
60
70
80
90
100
21
19
Jan-85 May-85 Sep-85 Jan-86 May-86 Sep-86 Jan-87 y a d m m [ l l a f n i a R -1]
17
15
13
11
9
7
5
3
1
Discharge [m3 s-1]
Rainfall
DEM(B)
Observed
0
10
20
30
40
50
60
70
80
90
100
21
19
17
15
13
11
9
7
5
3
1
Rainfall
DEM(E)
Observed
Jan-85 May-85 Sep-85 Jan-86 May-86 Sep-86 Jan-87
Date
Rainfall [mm day -1]
Discharge [m3 s-1]
Figure 4 Observed and calibrated hydrographs for the models
related to DEM(A), DEM(B) and DEM(E), after sink removal.
of artificial sinks. These results are in agreement with the
better quality, in terms of elevation, slope and land-sur-face,
of the DEM products of gridding methods types II and
III (cf. Fig. 2a).
The Multi-Window (MW) analysis included several evalua-tion
periods of different length. The MW test indicated that
the discharge performance of the models related to the DEM
products of gridding methods type II ((B) and (D)) were the
most acceptable in the different periods of analysis, regard-less
of the presence of sinks. Fig. 5 depicts yearly EF2 values
for the period (1984–1995) as a function of both DEMs
including sinks and DEMs after sink removal. Sink removal
was carried out by smoothing the DEMs. This caused a mod-ification
of the original structure of the smoothed DEMs,
particularly perceived when calculating the distribution of
slopes (smoother) and the drainage network topology. Con-sequently,
the simulation of surface water flow dynamics
was modified with a faster water routing throughout the
flatter (i.e. smoother) DEMs, characterised by a general
deterioration of the discharge performance (especially for
method (A)), as depicted in Fig. 5b. Nevertheless, for
DEM(C) the newer river network topology brought the gen-eral
enhancement of the drainage network, as compared
to the digitised drainage network, which had a positive ef-fect
on the performance of the model related to DEM(C).
In broad terms, the peak flows were reasonably well
simulated within the calibration period independently of
the DEM gridding method. However, in the main evaluation

1.0
0.8
0.6
0.4
0.2
0.0
EF2 [--]
Figure 5 Yearly MW model performances (streamflow) for the Gete station in relation to the DEM gridding methods and the
removal of artificial sinks.
38
34
30
26
22
18
14
10
DEMs including sinks
1984 1986 1988 1990 1992 1994
Evaluation window
period (1987–1988) the adequacy of the simulated peak
flows was inferior due to overestimation. The results of
the Extreme Value Analysis (EVA) in the period (1984–
1995) are illustrated in Fig. 6a for the DEMs including artifi-cial
sinks and Fig. 6b for the DEMs after sink removal. Be-sides
a generalised overestimation, Fig. 6a and b show
that removing the sinks by smoothing the DEMs produced
higher peaks, except for the product of method (C). Again,
higher peaks after sink removal were caused mainly by low-er
effective drainage levels (zdr, higher absolute values)
that accelerate the routing of water throughout the flatter
DEMs and evacuate higher overland and interflow volumes
from the catchment. Despite lower Tdr effective values
were obtained to mitigate the rising of the peaks, the pre-dicted
peaks were higher after sink removal. Fig. 6b shows
that the highest overestimation of peaks was related to
the DEM product of the MIKE SHE Bilinear interpolation algo-rithm
(method (A)). The peakflow predictions related to the
gridding methods (C), (B) and (D) were comparable and bet-ter
than the predictions related to the outcome of the
TOPOGRID algorithm (method (E)).
Concerning the DEM(C), the analysis revealed that the
improvement of the discharge performance after the addi-tional
DEM smoothing by the 3 · 3 mean filter is particularly
noticeable in the simulation of peakflows that are lower
after the additional DEM smoothing in the period (1984–
1995). This generalised improvement of the discharge per-formance
is consistent with the improved location of the
predicted catchment outlet (after the additional mean filter
smoothing) with regard to the location of the Gete station.
This illustrates the significant enhancement of the model
1984 1986 1988 1990 1992 1994
DEM(A) DEM(B) DEM(C)
38
34
30
26
22
18
14
10
1.0
0.8
0.6
0.4
0.2
0.0
Evaluation window
performance associated with the improvement of the DEM
drainage network topology despite the deterioration of
other DEM features such as the distribution of slopes (cf.
Fig. 3c). This illustrates as well the necessity of assessing
the consequences of GIS operations such as DEM smoothing
on both the structure of the DEMs and the associated model
performance before accepting the model predictions as
being valid.
Is the evaluation of internal state predictions affected
by the DEM generation methods?
Since distributed models should be evaluated against dis-tributed
measurements by considering the predictions of
internal state variables, this section illustrates the main dis-tributed
results from the hydrologic modelling.
The Multi-Site (MS) analysis of the river discharge predic-tions
for the two internal river stations that were not in-cluded
in the calibration process suggested that all of the
models have marked difficulties to predict the distributed
discharge variables with reasonable accuracy, suggesting
that some processes such as flow through saturated and
unsaturated zones may not be rightly modelled to this scale.
Besides the coarse modelling resolution, the noticeable
uncertainty attached to the input data that were used for
constructing the hydrologic model contributes probably in
a greater proportion to these low model efficiencies. In
any case, the discharge predictions for these stations (cf.
Fig. 1), related to the DEM products of the gridding methods
type II ((B) and (D)) and type III ((E)) were better.
Figs. 7 and 8 depict the piezometric level performance of
the models using DEMs after sink removal for three wells
considered in the Split-Sample (SS) test and two wells used
DEMs including sinks
6
0.1 1.0 10.0 100.0
Return period [years]
Discharge [m3 s -1]
Exp. Distr.
Observed
(A)
(B)
(C)
6
0.1 1.0 10.0 100.0
Return period [years]
Discharge [m³ s -1]
Exp. Distr.
Observed
(A)
(B)
(C)
(D)
(E)
EF2 [--]
Figure 6 EVA for the Gete station for the period (1984–1995) in relation to the DEM gridding methods and the removal of artificial
sinks.

V2TI-KU.PP2 (Landeniaan)
53
51
49
47
45
43
Jan May Sep Jan May Sep Jan May Sep Jan May Sep Jan
Head [m]
1985 1986 1987 1988
4038187 (Brusseliaan)
138
136
134
132
130
128
Head [m]
1985 1986 1987 1988
4048204 (Landeniaan)
116
114
112
110
108
106
104
Head [m]
1985 1986 1987 1988
Observed (A) (B) (C) (D) (E)
Figure 7 Predicted piezometric levels after model calibration
for some of the wells considered in the Split-sample (SS) test as a
function of the DEM gridding methods (after sink removal).
in the MS test, respectively. These figures show that, in gen-eral,
the prediction of the piezometric levels differed con-siderably
among the wells and that in some cases there
was an important variation of the performance in relation
to the DEM gridding methods. Both tests revealed that, in
general, the piezometric performances associated with
the DEM products of gridding methods type II ((B) and (D))
were poorer (cf. Figs. 7 and 8). This suggests that the mod-els
related to these DEMs were characterised by lower base-flow
and higher overland and interflow predictions, which
finally resulted in better global model performances.
These results illustrate the importance of carrying out an
evaluation of distributed models using distributed measure-ments
for the evaluation of simulated internal state vari-ables.
The variability of these results illustrates however
the inherent difficulties in doing so, including incommensu-rability
issues due to the fact that data errors (input and/or
evaluation) are usually unknown and scale dependent. It
should be noticed that Figs. 7 and 8 provide evidence for
rejection of all models, unless consideration is given to
the scale issue (i.e. incommensurability) of comparing
point-scale elevation and piezometric measurements versus
600-m grid predictions.
Conclusions
Three types of gridding methods were applied to produce
coarse DEMs (600-m resolution) for the modelling of the
Gete catchment with the MIKE SHE model. The first type
of gridding method uses input elevation data distributed
about the periphery of the gridded DEM cells (i.e. methods
(A) and (C)) and was implemented with a MIKE SHE pre-pro-cessing
tool for interpolation. The second type uses input
elevation data distributed about the centre of the gridded
cells (i.e. methods (B) and (D)). The third type is based on
the TOPOGRID (ESRI, 1996) algorithm that uses landscape
features, such as digitised streams, to improve the drainage
structure of the DEM product (i.e. method (E)).
A protocol, examining the accuracy of DEM elevations,
evaluating geomorphic relationships and predicting hydro-logic
conditions in hillslopes, was applied in this work for
characterising the quality of the coarse DEM products for
hydrologic use. The protocol revealed that, for the particu-lar
characteristics of the study site and the elevation input
data, gridding methods type II ((B) and (D)) produced coarse
DEMs with higher elevation accuracy, followed by TOPO-GRID
and finally by the MIKE SHE tool for interpolation (grid-ding
methods type I). Correspondingly, the Multi-Calibration
(MCal) analysis revealed a better performance (for outlet
discharges and peakflows) of the hydrologic models related
to gridding methods type II, regardless of the presence of
spurious sinks.
Thus, this study revealed that, in general, the DEM
products of the gridding methods type II are more appro-priate
for the current coarse modelling resolution. In this
context, the assessment of the model performance re-vealed
a congruence with the predictions of overland flow
generation from the topographic index analysis, that is,
higher runoff production induced by the DEM products of
gridding methods types II and III and smoother runoff pro-ductions
related to the DEM products of gridding methods
type I ((A) and (C)).
Some of the piezometric results suggested however a
potential underestimation of base flow associated to the
DEM products of gridding methods type II that could not
V2HG-BR.B5 (Landeniaan)
62
60
58
56
54
52
Head [m]
1985 1986 1987 1988
4038488 (Brusseliaan)
128
126
124
122
120
118
Head [m]
1985 1986 1987 1988
Observed (A) (B) (C) (D) (E)
Figure 8 Predicted piezometric levels after model calibra-tion
for some of the wells considered in the Multi-site (MS) test
as a function of the DEM gridding methods (after sink removal).

be studied further owing to the lack of baseflow measure-ments
or estimates. The present research could therefore
be extended to investigate in the future a protocol for
estimating total hydrograph subflows and assessing in this
way the performance of the hydrologic models simulating
subflows. The subflow analysis may have the potential of
improving the simulation of certain processes that other-wise
might be simulated wrongly at the current 600-m
resolution.
The Multi-Site (MS) test indicated moreover that all of
the hydrologic models predict distributed state variables
with even lower performances than the performances cor-responding
to calibrated variables. This test illustrated fur-thermore
the importance of using distributed observations
of streamflow and piezometric levels for evaluating the
model performance simulating internal state variables.
The significant variability of the model performance results
indicated however the inherent difficulties in achieving
this distributed evaluation, including incommensurability
issues, owing among other factors to data uncertainty
and scale-dependent aspects that affected the direct com-parison
of point-scale observations versus 600-m grid pre-dictions.
It is important therefore to complement in the
future the current analysis by including in the distributed
evaluation protocol estimated intervals of data uncertainty
that could enable accounting not only for data errors but
also for discrepancies between point-scale measurements
and grid-scale predictions. In this respect, the analysis of
the procedures that were followed up for deriving dis-charge
observations from the rating (i.e. level versus dis-charge)
curves and the analysis on the discrepancies
among the input elevation data (Zsource), the DEM eleva-tions
(ZDEM) and the ground levels utilised for monitoring
the observation wells (Zmonitor) are likely to play an impor-tant
role in the estimation of the referred data uncertainty
intervals.
The MCal test using DEMs without spurious sinks revealed
an average influence of the gridding methods on the effec-tive
values adopted by the parameters of the hydrologic
models, except for the DEM (C) that has a drainage outlet
located in a different position with respect to the location
of the catchment outlet.
Artificial sinks were removed from the DEM outcomes of
gridding methods types I and II. This assessment revealed
that the products of gridding methods type II include a
higher amount of artificial sinks than the outcomes of type
I. Comparing the conditions after the sink removal opera-tion
with the conditions observed prior the referred oper-ation,
the MCal analysis after sink removal showed a
tendency for obtaining zdr, Tdr and Ksat–Kw parameter val-ues
incrementing the overland and interflow volumes and
accelerating the routing of these hydrograph components,
but controlling at the same time and as much as possible
the magnitude of peaks. This general tendency is to com-pensate
the deterioration of model predictions due to the
smoothing (i.e. flattering) of DEMs caused by the sink re-moval
and explains as well the variation of baseflow pre-diction
noticed through the analysis of piezometric levels
to compensate for the changes in overland and interflow
volumes.
Despite the multi-objective and systematic approach to
multi-calibration, the trial and error methodology that
was used in this research is based on the concept of
attaining a single optimum set of parameters. However,
given the high dimensionality of the parameter space
associated with the distributed model of the study site,
it is likely that this parameter space was not adequately
sampled with the consequent risk of having identified only
a local optimum rather than a global optimum. Further-more,
the calibration of distributed models is usually fac-ing
the risk of parameter equifinality, that is, several sets
of parameter values that give acceptable fits to the cali-bration
data might be scattered widely in the parameter
space, as a result of errors in the data and model struc-ture,
besides parameter interactions (Beven and Freer,
2001).
Thus, an important future activity is to define prediction
limits for estimating the degree of confidence on the cur-rent
hydrological modelling by taking into account in the
scope of a joint deterministic-stochastic framework the
uncertainties in data, model structure and parameters. In
this context, the current research should be understood as
a preliminary sensitivity analysis aiming to reduce as much
as possible the geomorphologic data uncertainty by analy-sing
in a complementary way the accuracy of DEMs and
the associated model performances.
Finally, the reader should be aware that some of the re-sults
obtained in this research, in particular about the out-comes
of the DEM gridding methods, may be model
structure, modelling resolution and catchment specific.
The main objective of this manuscript is to communicate
to the reader the importance of assessing the quality of
the topographic input data (i.e. identifying intrinsic errors)
to avoid negative consequences on the hydrological
modelling.
Acknowledgements
This work was possible thanks to research grants from the
OSTC (Belgian Federal Office for Scientific, Technical and
Cultural Affairs, project CG/DD/08C), the Interuniversity
Programme in Water Resources Engineering (IUPWARE,
KULeuven-VUBrussel) and the Katholieke Universiteit Leu-ven
(postdoctoral project PDM/03/188, awarded to the
first author). The completion of this article was achieved
in the framework of the Research and Development con-tract
of the first author funded by the Instituto Nacional
de Investigacioń y Tecnologı á Agraria y Alimentaria (INIA,
Spain) and the Centro de Investigacioń y Tecnologı á Agro-alimentaria
de la Diputacioń General de Aragoń (CITA-DGA,
Spain). The authors would like to thank those who
have supported us and helped to clarify our way through-out
this continued research. Special thanks go to Luc
Feyen, Patrick Willems and Prof. Keith Beven for their con-structive
suggestions.
References
Allen, G.R., Pereira, L.S., Raes, D., Martin, S., 1998. Crop
evapotranspiration-guidelines for computing crop water require-ments.
FAO Irrigation and Drainage Paper, 56, Food and
Agriculture Organization, Rome, 290pp.

Ambroise, B., Beven, K.J., Freer, J., 1996. Towards a generalisation
of the TOPMODEL concepts: topographic indices of hydrologic
similarity. Water Resources Research 32 (7), 2135–2145.
Beasley, D.B., Huggins, F., Monke, E.J., 1980. ANSWERS: a model
for watershed planning. Transactions of the ASAE 23 (4), 938–
944.
Beldring, S., 2002. Multi-criteria validation of a precipitation-runoff
model. Journal of Hydrology 257, 189–211.
Bergstro¨m, S., Graham, L.P., 1998. On the scale problem in
hydrologic modelling. Journal of Hydrology 211, 253–265.
Beven, K., Freer, J., 2001. A dynamic TOPMODEL. Hydrological
Processes 15, 1993–2011.
Beven, K.J., Lamb, R., Quinn, P.F., Romanowicz, R., Freer, J.,
1995. TOPMODEL. In: Singh, V.P. (Ed.), Computer Models of
Watershed Hydrology. Water Resources Publications, USA, pp.
627–668.
Braud, I., Fernandez, P., Bouraoui, F., 1999. Study of the rainfall-runoff
process in the Andes region using a continuous distributed
model. Journal of Hydrology 216, 155–171.
Christiaens, K., Feyen, J., 2002. Constraining soil hydraulic param-eter
and output uncertainty of the distributed hydrological MIKE
SHE model using the GLUE framework. Hydrological Processes 16
(2), 373–391.
DHI, 1998. MIKE-SHE v.5.30 User Guide and Technical Reference
Manual, Danish Hydraulic Institute, Denmark, 50pp.
Doorenbos, J., Pruitt, W.O., 1977. Crop water requirements. FAO
Irrigation and Drainage Paper, 24, Food and Agriculture Organi-sation,
Rome, 156pp.
ESRI, 1996. ARC/INFO Online Documentation. Environmental Sys-tems
Research Institute Incorporated, Redlands, CA, USA.
Feyen, L., Va´zquez, R.F., Christiaens, K., Sels, O., Feyen, J., 2000.
Application of a distributed physically-based hydrologic model to
a medium size catchment. Hydrology and Earth System Sciences
4 (1), 47–63.
Gu¨ntner, A., Uhlembrook, S., Seibert, J., Leibundgut, Ch., 1999.
Multi-criterial validation of TOPMODEL in a mountainous catch-ment.
Hydrological Processes 13, 1603–1620.
Hanselman, D., Littlefield, B., 1998. Mastering MATLAB 5: A
Comprehensive Tutorial and Reference. Prentice-Hall Inc., UK,
pp. 638.
Hutchinson, M.F., 1989. A new procedure for gridding elevation and
stream line data with automatic removal of spurous pits. Journal
of Hydrology 106, 211–232.
Hutchinson, M.F., Dowling, T.I., 1991. A continental hydrological
assessment of a new grid-based Digital Elevation Model of
Australia. Hydrological Processes 5, 45–58.
Jain, S.K., Storm, B., Bathurst, J.C., Refsgaard, J.C., Singh, R.D.,
1992. Application of the SHE to catchments in India–Part 2: field
experiments and simulation studies on the Kolar Subcatchment
of the Narmada River. Journal of Hydrology 140, 25–47.
Jayatilaka, C.J., Storm, B., Mudgway, L.B., 1998. Simulation of
water flow on irrigation bay scale with MIKE SHE. Journal of
Hydrology 208, 108–130.
Jenson, S., Domingue, J., 1988. Extracting topographic structure
from digital elevation data for geographic information system
analysis. Photogrametric Engineering and Remote Sensing 54
(11), 1593–1600.
Kristensen, K.J., Jensen, S.E., 1975. A model for estimating actual
evapotranspiration from potential evapotranspiration. Nordic
Hydrology 6, 170–188.
Labs, Clark, 1998. Online Idrisi32 Help SystemClark Labs. Clark
University, Worcester, MA, USA.
Lane, S.N., Brookes, C.J., Kirkby, M.J., Holden, J., 2004. A network-index-
based version of TOPMODEL for use with high-resolution
digital topographic data. Hydrological Processes 18, 191–201.
Legates, D.R., McCabe, G.J., 1999. Evaluating the use of ‘‘good-ness-
of-fit’’ measures in hydrologic and hydroclimate model
validation. Water Resources Research 35 (1), 233–241.
Loague, K., Green, R.E., 1991. Statistical and graphical methods for
evaluating solute transport models: overview and applications.
Journal of Contaminant Hydrology 7, 51–73.
Madsen, H., 2003. Parameter estimation in distributed hydrological
catchment modelling using automatic calibration with multiple
objectives. Advances in Water Resources 26 (2), 205–216.
Quinn, P., Beven, K., Chevallier, P., Planchon, O., 1991. The
Prediction of Hillslope Flow Paths for Distributed Hydrological
Modelling Using Digital Terrain Models. Hydrological Processes 5,
59–79.
Quinn, P., Beven, K.J., Lamb, R., 1995. The ln(a/tanb) index: how
to calculate it and how to use it within the TOPMODEL
framework. Hydrological Processes 9, 161–182.
Refsgaard, J.C., 1997. Parameterisation, calibration and validation
of distributed hydrological models. Journal of Hydrology 198,
69–97.
Refsgaard, J.C., Knudsen, J., 1996. Operational validation and
intercomparison of different types of hydrological models.
Water Resources Research 32, 2189–2202.
Refsgaard, J.C., Storm, B., 1995. MIKE SHE. In: Singh, V.P. (Ed.),
Computer Models of Watershed Hydrology. Water Resources
Publications, USA, pp. 809–846.
Saulnier, G-M., Obled, Ch., Beven, K., 1997a. Analytical compen-sation
between DEM grid resolution and effective values of
saturated hydraulic conductivity within the TOPMODEL frame-work.
Hydrological Processes 11, 1331–1346.
Saulnier, G-M., Beven, K., Obled, Ch., 1997b. Digital elevation
analysis for distributed hydrological modelling: Reducing scale
dependence in effective hydraulic conductivity values. Water
Resources Research 33 (9), 2097–2101.
Vander Poorten, K., Deckers, J., 1994. A 1:250,000 Scale soil map of
Belgium. Proceedings of the Fifth European Conference and
Exhibition on Geographic Information Systems, EGIS ‘94, Utrecht
1, 1007–1015.
Va´zquez, R.F., 2003. Assessment of the performance of physically
based distributed codes simulating medium size hydrological
systems. PhD dissertation ISBN 90-5682-416-3, Department of
Civil Engineering, K.U. Leuven, Belgium, 335pp.
Va´zquez, R.F., Feyen, J., 2003a. Assessment of the performance of
a distributed code in relation to the ETp estimates. Water
Resources Management 16 (4), 329–350.
Va´zquez, R.F., Feyen, J., 2003b. Effect of potential evapotranspi-ration
estimates on effective parameters and performance of
the MIKE SHE-code applied to a medium-size catchment. Journal
of Hydrology 270 (4), 309–327.
Va´zquez, R.F., Feyen, J., 2004. Potential Evapotranspiration for the
distributed modelling of Belgian basins. Journal of Irrigation and
Drainage Engineering 130 (1), 1–8.
Va´zquez, R.F., Feyen, L., Feyen, J., Refsgaard, J.C., 2002. Effect of
grid-size on effective parameters and model performance of the
MIKE SHE code applied to a medium sized catchment. Hydro-logical
Processes 16 (2), 355–372.
Vertessy, R.A., Hatton, T.J., O’Shaughnessy, P.J., Jayasuriya,
M.D.A., 1993. Predicting water yield from a mountain ash forest
catchment using a terrain analysis based catchment model.
Journal of Hydrology 150, 665–700.
Vivoni, E.R., Ivanov, V.Y., Bras, R.L., Entekhabi, D., 2005. On the
effect of triangulated terrain resolution on distributed hydro-logic
modeling. Hydrological Processes 19 (11), 2101–2122.
Walker, J.P., Willgoose, G.R., 1999. On the effect of digital
elevation model accuracy on hydrology and geomorphology.
Water Resources Research 35 (7), 2259–2268.
Willems, P., 2000. Probabilistic immission modelling of receiving
surface waters. PhD dissertation ISBN 90-5682-271-3, Depart-ment
of Civil Engineering, K.U. Leuven, Belgium, 233pp.
Wise, S., 2000. Assessing the quality for hydrologic applications of
digital elevation models derived from contours. Hydrological
Processes 14, 1909–1929.

Wolock, D.M., Price, C.V., 1994. Effects of digital elevation model
map scale and data resolution on a topography-based
watershed model. Water Resources Research 30 (11), 3041–
3052.
Xevi, E., Christiaens, K., Espino, A., Sewnandan, W., Mallants, D.,
Sorensen, H., Feyen, J., 1997. Calibration, validation and
sensitivity analysis of the MIKE-SHE model using the Neuenkir-chen
catchment as case study. Water Resources Management 11,
219–239.
Zhang, W., Montgomery, D.R., 1994. Digital elevation model grid
size, landscape representation, and hydrologic simulations.
Water Resources Research 30 (4), 1019–1028.

Mike she usa

Recommended

Recommended

More Related Content

What's hot

What's hot (19)

Viewers also liked

Viewers also liked (7)

Similar to Mike she usa

Similar to Mike she usa (20)

Recently uploaded

Recently uploaded (20)

Mike she usa