The application of the SEAWAT variable density code for the Lake Wieringen bonte-biesheuvel

Full text

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 1
Abstract— Presently, very advanced regional modelling tools
are available to quantify the changes in groundwater flow and
groundwater quality. The application of these modelling tools is
increasingly becoming a part of environmental assessments of
large projects. In this paper we examined the application of the
SEAWAT variable density model for the Lake Wieringen Project.
Firstly, an assessment is made of the reliability of the predicted
salinity trends through comparison of model predictions with
hydrochemical data and surface water data. Secondly, the
application of the results of the modelling in the environmental
assessment of the lake Wieringen Project is described.
Index Terms—groundwater modelling, variable density
I. INTRODUCTION
he Lake Wieringen project comprises the establishment
of a fresh water lake with an area of roughly 700 hectares,
located on the fringe of the former island of Wieringen and the
Wieringermeerpolder1 in the north of the Netherlands. The
establishment of the Lake Wieringen aims to increase the
socio-economic resilience of the region and make the region
less dependent on agriculture as a main source of income.
Because water played an important role in the design of the
lake, the regional design of the plan was backed by an
integrated water and ecology impact assessment. This served
both to evaluate the effects of the lake on the surrounding land
and to optimise the design of the water management system of
the lake itself and the surrounding polder. This assessment
included 1) a hydrological study aiming to quantify safety
against flooding, 2) an aquatic ecological and surface water
quality assessment to predict the future ecological state of the
lake (Witteveen+Bos,2006a), and 3) an hydrogeological study
aiming to assess the effects of the lake on groundwater levels,
groundwater seepage and groundwater quality in the region
(Witteveen+Bos, 2006b).
The hydrogeology of the region is characterised by varying
occurrence of fresh and saline groundwater. This situation is
the result of the geological history and land reclamation and
cultivation. In order to understand the impact that the Lake
Wieringen will have on groundwater quantity and quality,
whilst considering the varying salinity of groundwater, a
variable density groundwater model was developed using the
SEAWAT code. With this model, the historical and future
Witteveen+Bos. Almere, the Netherlands.
E-mail: m.bonte@witteveenbos.nl
1 “Polder” is Dutch for reclaimed land. Many of the areas in the west of
Holland consist of reclaimed land and have the word polder in their name.
trends in groundwater salinity were simulated. This was
followed by scenario modelling of several design alternatives
for the lake.
This paper describes the results of the hydrogeological study
and contains: 1) the construction of a variable density
groundwater model using the SEAWAT code, 3) the
calibration and validation of the model, 4) the results of the
scenario modelling of the autonomous development and the
establishment of the Lake Wieringen. The results of the
scenario modelling are discussed in terms of reliability and the
usability of the results in an Environmental Impact
Assessment.
II. DESCRIPTION OF THE WIERINGERMEER REGION
A. The region’s history and its relation to groundwater
salinity
The Wieringermeer and Wieringen are located in the north
of North-Holland (figure 1). The region is characterised by a
complex natural history, which is reflected in the groundwater
salinity pattern. The highlights of the region’s history are
described below.
From around 2000 years B.C. through to the Middle-ages, a
fresh water swamp system was located in the region, where
The application of the SEAWAT variable density code for the Lake Wieringen
project, the Netherlands
M. Bonte
1
, A. Biesheuvel
1
T
Fig 1: Location of the Wieringermeer and Wieringen

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 2
peat was formed (van de Ven, 2003). During this period fresh
water recharged the underlying aquifer. Between the 9th and
12th century, land cultivation, dewatering and peat digging for
fuel, led to a decline in topographical elevation from several
meters above mean sea level to several meters below mean sea
level. Remnants of several dikes suggest that these early
inhabitants of the region were fighting a continuous battle
against the rising sea in the increasingly lower lying land (van
de Ven, 2003). In those days, several sea inlets existed along
the western shore line of Holland through which the sea
threatened the peat digging areas in the region.
When during the end of the 12th century and start of the 13th
century several storm-floods from the north and east broke
away remaining peat and inundated the area, a large inland
arm of the Northsea called the Zuiderzee was formed in this
region. Due to the occurrence of glacial moraine deposits,
Wieringen itself is situated on topographically higher ground.
The inundation of the Zuiderzee separated Wieringen from
land further south and made it an island. At the inundated part
of the region, where coarse sands occurred in the underground,
saline seawater intruded the aquifer rapidly caused by density
induced free convection. At areas where low permeable clay
and loam layers occurred in the underground, fresh water
resources were protected from salinisation (Post, 2004).
During the depression years (1920’s to, 1930’s) large scale
labour projects were undertaken in the Netherlands to
stimulate the economy. One of these projects was the
reclamation of Wieringermeerpolder. In 1924, a dike was built
between the island and the main land and the
Wieringermeerpolder was again land. In the same decade the
remainder of the Zuiderzee was disconnected from the North
Sea through the construction of a dam named the “Afsluitdijk”.
From then on the Zuiderzee was named the IJsselmeer.
Through several decades, the IJsselmeer gradually became a
fresh water lake, fed by a tributary of the Rhine.
B. Water management and hydrogeological setting
Wieringen was formed in the former last Ice Age (the
Saalien) when a glacier pushed a ridge with a topographical
elevation up to 10 meters above mean sea level and a glacial
moraine was deposited under Wieringen. The occurrence of
these relatively poorly permeable glacial deposits make that a
large part of the rainfall discharges locally in lower parts on
the former island. These lower parts are called ‘kogen’. Fresh
water lenses are developed under the higher areas. Brackish to
salt water discharges in the lower areas.
The Wieringermeerpolder is a typical deep polder with
polder drainage levels varying from 4 to 6 meters below
average sea level. To facilitate the deep dewatering in the
polder, an intensive drainage system was constructed when the
polder was established. This drainage system is combined with
a separate water inlet system which conveys fresh water from
the adjacent IJsselmeer, through the Amstelmeerkanaal, to
farmers in the region.
In the region a shallow and deep aquifer can be recognised
comprising unconsolidated coarse sand deposited by glacial
rivers in the early and late Quaternary. The shallow and deep
aquifer are separated by the before mentioned glacial moraine
deposits. At Wieringen these deposits are nearly at the ground
surface and the shallow aquifer has a thickness of several
meters. South of Wieringen, in the northern half of the
Wieringermeerpolder, the moraine deposits are not present and
the shallow and deep aquifer are in direct hydraulic
connection. Further to the south, the glacial deposits at around
30 meters below ground surface. In the south, the deep aquifer
is characterised by several minor aquitards. Figure 2 shows a
hydrogeological cross section of the region. It can be seen
from this cross section that the occurrence of fresh and saline
groundwater is closely related to the hydrogeological structure
of the region. Clay layers protect relicts of fresh groundwater
present in the underground.
III. THE PROJECT
Figure 3 shows an overview of the proposed Lake
Wieringen. The Lake Wieringen project will be undertaken in
four phases. In the first phase part of the lake will be built at
the Polder Waard-Nieuwland and recreational housing will be
realised. In the second phase, the entire Lake Wieringen will
be constructed by dredging and construction of a dike. In the
last two phases the main infrastructural works will be
undertaken (roads and water infrastructure).
Fig 2: Hydrogeological cross section
Fig 3: Layout of the Lake Wieringen

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 3
The construction of the four phases of the project will be
completed around 2025. The design surface water level of the
proposed Lake Wieringen in each of the scenario’s will be 0.4
meters below mean sea level which is around 4 meters above
the present polder surface water level. For this reason, the
establishment of the lake is expected to impact both on
groundwater flow and salinity patterns. The Lake receives
surface water inflow from the IJsselmeer though an inlet
construction located in the harbour. Water is discharged to the
Amstelmeer which is located west from the lake.
IV. CONSTRUCTION OF VARIABLE DENSITY MODEL AND
CALIBRATION
A. Model construction
To assess the effect of the establishment of the Lake
Wieringen on groundwater conditions, a variable density
groundwater flow and solute transport model was developed
using the SEAWAT 2000 code (Langevin, 2003). The
SEAWAT code is based around the existing codes
MODFLOW 2000 (Harbaugh et al., 2000), simulating
groundwater flow and MT3DMS (Zheng & Wang, 1999),
simulating solute transport. The two models have been coupled
and combined with a variable density flow module (the VDF
package) that calculates the water pressure field for each time
step based on groundwater heads and solute concentration
using a linear relationship between solute concentration and
density. The darcian flow equation in MODFLOW has been
adjusted to account for buoyancy effects.
The Lake Wieringen-model was based on an earlier model
developed for a pre-feasibility assessment using the
MOCDENS3D code (Grontmij, 2001; Oude Essink, 2002).
Both MOCDENS3D and the SEAWAT codes are based on
MODFLOW, facilitating a fairly easy model conversion. The
reason to convert the model to SEAWAT 2000 and not use the
initial MOCDENS3D model was mainly that the SEAWAT
2000 model is publicly available from the USGS website and
comes with an extensive model documentation. During model
construction, a number of additional advantages were found:
1) SEAWAT 2000 can be run using the 3rd order TVD scheme
for the advection term for the solute transport part, which was
found to be less prone to numerical dispersion and artificial
oscillations than the various available Method of
Characteristic (MOC) schemes, 2) SEAWAT 2000 uses actual
measured groundwater heads, whilst MOCDENS3D uses
corrected fresh water heads, the former allows easier model
calibration; and 3) SEAWAT 2000 was found to be
numerically more stable. Especially in regions where
inversions in the salinity-depth profile occurred,
MOCDENS3D did not converge towards a stable solution. In
these areas the interpreted salinity pattern required manual
corrections (Oude Essink, 2002). Table 1 shows some details
of the Lake Wieringen model.
B. Temporal discretisation
Both a steady state model simulating present groundwater
flow conditions and a transient model simulating groundwater
flow conditions between 1970 and 2150 were built. The steady
state conditions were simulated using the interpreted chloride
concentration. The concentration field is translated in a density
field which is used in the steady state groundwater flow
calculation.
The transient model was set up using stress periods with a
length of 5 years throughout simulation period. It was
attempted to simulate a historical longer period to consider the
development in salinity pattern over a longer time frame
relating to the discussed historical development of the region,
but calculating time proved to be a limiting factor. The relation
of the numerical results and the natural history of the region
are therefore only discussed qualitatively.
In the period 1970 to 2000, boundary conditions are
assumed to be constant in time. This period serves to avoid
initial numerical transient effects in the chloride field. From
2000 onwards, sea water levels and recharge are adjusted to
account for the effects of Climate Change using the Middle
Scenario for Climate Change as described by the
Intergovernmental Panel of Climate Change (Commissie
Waterbeheer 21e eeuw, 2000). The following significant
changes in the hydrologic system are adopted for the year
2150 (autonomous developments):
- Sea level rise of 0.9 m;
- Increase in groundwater recharge of 10.5%;
- Soil subsidence of 0.15 m (NW4, 1997).
The soil subsidence is included in the sea level rise term,
implying that the head of the General Head Boundary
condition is increased with 1.05 m over a time period of 150
years. It is noted that the intra-annual climatic changes are not
included in the model because the length of stress periods does
not allow this. This means that the increase in drought periods
TABLE I
DETAILS OF NUMERICAL MODEL
Real world Model representation
Model domain 23 x 32 km2
Horizontal discretisation 200 x 200 m2 at boundaries of the
model to 50 x 50 m2 near Lake
Wieringen
Vertical discretisation 22 layers
Transmissivity and leakance
values
Hydrogeological database REGIS
(TNO-NITG, 2005) and model
calibration
Polder drainage and watersystem Simulated using RIVER and DRAIN
packages. Conductance values are
based on soil mapping and model
calibration. Chloride values of
boundaries are based on average
values from measurements.
Surface waters IJsselmeer and Waddenzee are
simulated using General Head
Boundaries (GHB)
Groundwater recharge Long term climatic data, crop factors
and GIS land use data.
Initial heads Steady state calculation.
Initial chloride concentration Initial chloride concentrations were
interpreted from groundwater
salinity data from 1930 to present.

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 4
(often associated with salinity) can not be directly quantified
using this modelling approach.
C. Calibration
The steady-state model was used for calibration against
measured groundwater heads and discharge rates measured at
the polder pumping stations.
The calibration steps involved: 1) adjusting the
transmissivity values from the REGIS database to achieve an
acceptable difference between measured and calculated heads
measured in deep monitoring wells; and 2) adjusting the
conductance of the GHB, DRAIN and RIVER packages to
achieve an acceptable difference between measured and
calculated phreatic groundwater levels and polder discharge
rates. This led to an increase in transmissivity values with a
factor ranging from 1.5 to 2.5 and conductance values ranging
from 60 to 200 m2/day (corresponding to drainage resistance
values ranging from 200 to 700 days).
Figure 4 presents the scatter plots of measured against
calculated hydraulic heads and discharge rates. The mean
absolute error between measured and calculated heads after
calibration varied from around 0.2 meter in the shallow model
layers to 0.5 meter in the deeper model layers. The increase in
calibration error with depth is probably due to the increasing
effect of chloride concentration on the pressure head. The total
hydraulic gradient in the model region is around 8 meters
which means that the relative error is less than 10 % which was
the calibration objective.
The relative error between measured and calculated
discharge rates for five pumping stations after calibration was
less than 10 % and around 15% for two pumping stations. This
calibration result was considered adequate because the error in
discharge measurements at pumping stations ranges from 10 to
30%.
V. RESULTS
A. Autonomous development
An important question is what will happen to the
groundwater quality when the Lake Wieringen is not
constructed. As described, the proposed Lake Wieringen will
be completed in 2025. So to make a reliable comparison of the
effects of the project, the hydrogeological situation in this year
needs to be described and compared to the effects of the Lake
Wieringen. Further, the historical trends in groundwater
salinity prove not to be constant in time, which means that the
development of groundwater salinity with and without the
Lake Wieringen needs to be compared over a longer time
frame. Figure 4 shows the predicted salinity of shallow
groundwater over 150 years from 2000 onwards. These results
indicate that the northern half of the Wieringermeerpolder is
characterised by increasing salinity, whilst in the southern half
1.E+03
1.E+04
1.E+05
1.E+06
1.E+03 1.E+04 1.E+05 1.E+06
Meas ure d polder disch arge (m3/day)
Calculated polder discharge (m 3/d ay)
Wierin germeer polder
Wierin gen
1:1 line
-8.00
-7.00
-6.00
-5.00
-4.00
-3.00
-2.00
-1.00
0.00
1.00
-8.00 -6.00 -4.00 -2.00 0.00
Mea sure d groundw ater levels (m - se ale vel)
Calculated groundw ater levels (m - sealevel)
modellayers 4 - 9
modellayers 10 - 15
modellayers 16 - 18
modellayers 19 - 22
1:1 line
Fig 4: Calibration result
Fig 5: Autonmous salinity development

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 5
groundwater freshening occurs. The increasing salinity in the
north is primarily caused by the gradual upward movement of
deeper, more saline groundwater, and secondly by rising sea
levels.
Interestingly, it was found that the increase in groundwater
seepage and salt-load levelled out over time, while sea level
rise continues. This is due to increasing salinity of shallow
groundwater which can almost completely compensate the
effect of rising sea-levels. This is caused by the higher density
of saline groundwater, which causes an increase of downward
hydraulic pressure whilst maintaining a constant groundwater
head. On the island of Wieringen the fresh water lens gets
gradually smaller. This effect is primarily due to the rising sea
level and in contrast to the Wieringermeerpolder where sea
level rise has a smaller effect.
Apart from changing groundwater quality, modelling
predicts that Wieringen will be confronted with increasing
groundwater levels. This results in an increased risk for water
logging and inundation. Model simulations show these effects
will become important from 2050 onwards. Rising
groundwater levels are negligible in the Wieringermeerpolder.
B. Hydrogeological effects of the Lake Wieringen
The effect that the establishment of the Lake Wieringen will
have on shallow groundwater is illustrated in figure 6. Figure 6
shows that groundwater salinity increases rapidly following
construction of the Lake. This effect is added to the effects of
autonomous development (indicated in figure 5), and of
comparable order of magnitude in the area directly
surrounding the Lake Wieringen. The salinity of the shallow
groundwater is determined by rainfall and by groundwater
seepage and its respective salinities. The predicted increase in
salinity is of comparable order of magnitude to the increased
groundwater seepage. This means that the increasing
salinisation is the result of the increase in groundwater seepage
in the polder areas, and not the result of mobilisation of more
saline groundwater. That is also the reason that increasing
groundwater salinity is predicted to occur almost immediately
after the lake is constructed over a relatively large area.
The model simulations for the years beyond 2050, show that
freshening of groundwater occurs. This is due to the
infiltration of less saline water from the Lake Wieringen,
which will cause a slowly progressing freshening of
groundwater in the Wieringermeerpolder. This is a process
that will take several centuries before a new equilibrium is
reached.
VI. DISCUSSION
A. Verification of the predicted autonomous development
In the project area, there are no monitoring wells present
that have a reasonable salinity time series data. The absence
of groundwater salinity data means that direct calibration of
the predicted autonomous development (through comparing
measured against simulated chloride concentrations) is not
possible. Because (autonomous) salinisation is of great
concern, both the agricultural section, nature development, and
more recently in the process of implementing the EU
groundwater directive, it is essential to how reliable the
predictions are from variable density modelling.
To obtain an indication of the reliability of the model’s
transient predictions, the calculated salinity trends between
1970 and 2000 were compared to the hydrochemical
‘signature’ of groundwater samples and the trends in surface
water salinity.
1) Salinity trends inferred from major ion chemistry
Salinising or freshening trends can be identified by
analysing groundwater quality. A simple way to determine
these trends is to calculate the Base Exchange Index or BEX
(Stuyfzand, 1999). BEX is defined by calculating the sum of
Na, K and Mg in mili-equivalents corrected by the
contribution of sea-salts by the following formula:
meq/l] [ Cl 1.0716- Mg) K (Na BEX
+
+
=
(1)
A significantly positive BEX indicates freshening, while a
significantly negative BEX indicates salinisation. A neutral
BEX indicates adequate flushing with water of constant
composition. The BEX is based on the differences in cation
exchange between a fresh water intrusion and a salt water
intrusion.
Groundwater data from a number monitoring wells was
obtained from the Dutch groundwater database DINO and for
each groundwater sample taken between 1980 and 2000, the
Base Exchange Index (BEX) has been calculated.
Subsequently, the trends identified with the Base Exchange
Index are compared to the trends calculated with the variable
Fig 6: Effects of Lake Wieringen on groundwater salinity

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 6
density groundwater model between 1980 and 2000 and shown
in figure 7 and summarised in table 2.
A reasonable resemblance is found between hydrochemical
data and the groundwater model results. Figure 7 shows that
the predicted regional groundwater quality trend, which
salinisation in the north and freshening in the south of the
Wieringermeerpolder corresponds with the trends inferred
from the major ion chemistry. Although rather qualitative by
nature, this comparison gives a reasonable indication that the
trends calculated by the model are reliable.
2) Salinity trends inferred from surface water data
In contrast to groundwater quality data, there is a reasonable
monitoring network present for surface water quality. For each
surface water monitoring station, the trend in salinity is
calculated using linear regression between 1990 and 2000.
Figure 8 shows a comparison between the calculated trends in
chloride concentration of the surface water and the phreatic
groundwater. Table 4 shows the comparison quantified.
The comparison between trends in surface water and
groundwater salinity show that whilst the groundwater model
indicates both freshening and salinising of groundwater to
occur, the surface water shows only freshening trends.
Obviously, the trends in groundwater and surface water can
not directly be compared to each other. However the
discrepancy appears rather high when it is realised that the salt
load in surface water is entirely derived from groundwater
discharge. The difference between trends in surface water and
groundwater may be due to an increase in flushing of the
surface water system with fresh water.
Local farmers take fresh water in from the adjacent
IJsselmeer to irrigate their crops and flush the surface water.
The water intake may have increased in the area, both to
irrigate higher value and more sensitive crops which require
more and fresher water, and potentially also the flush the
surface water system in response to increasing groundwater
salinity. There are no data available on water inlet volumes
from the IJsselmeer so this can not be verified. However, the
freshening in surface water salinity does indicate that local
farmers probably don’t really have a problem with salinity yet.
This was also confirmed through conversations with local
farming organisations.
B. Using the modelling results
The result of modelling the proposed Lake Wieringen
predict that groundwater levels in the area surrounding the lake
will rise and groundwater will become more saline in the near
future. Whether this is a negative or positive effect will depend
on the type of land use in the area. Most important land uses
include agriculture and nature. In order to use the results of the
hydrogeological modelling in an Environmental Impact
Statement (EIS), the effects on these functions should be
Fig 7: Comparison of autonomous trend with major ion chemistry
TABLE III
VERIFICATION OF CALCULATED SALINITY TRENDS USING THE BASE
EXCHANGE INDEX
TREND CALCULATED WITH SEAWAT
Salinising No Trend Freshening
Salinising 16 3 11
No Trend 12 1 3
Trend
inferred from
BEX Freshening 8 2 22
Fig 8: Comparison of autonomous trend with surface water trends
TABLE IV
VERIFICATION OF CALCULATED SALINITY TRENDS USING SURFACE
WATER TRENDS
TREND CALCULATED WITH SEAWAT
Salinising No Trend Freshening
Salinising 0 0 1
No Trend 0 0 1
SURFACE
WATER
TREND Freshening 13 4 16

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 7
described in terms of crop yield, ecological health and
inundation risk.
In order to interpretate the hydrogeological effects relating
to groundwater quantity aspects (water levels and seepage), the
WaterNOOD-code (STOWA 2003) was used. WaterNOOD is
a GIS based tool that quantifies agriculture yield and
ecological health using calculated groundwater levels,
groundwater seepage, soil type and crop type or ecological
species. The agricultural yield and ecological health are
represented as a percentage of the crop yield or ecological
health in an ideal hydrological situation for a certain soil type
and land use. The WaterNOOD instrument provides a very
powerful tool in quantifying hydrological effects in terms of
effects on different functions in an area.
Unfortunately, the WaterNOOD instrument can not be used
to translate changes in groundwater quality into agricultural
effects and effects on nature. One way of assessing the effect
of changing groundwater quality is using salt stress functions
developed for typical Dutch crops by Alterra (2003). These
salt stress functions describe the relationship between rising
salinity and decreasing crop yield. The problem with these
relationships is however that salt damage to crops of
vegetation occurs when salinity increases in the unsaturated
zone around the plant root. The salinity in the unsaturated zone
is determined not only by the shallow groundwater salinity but
also by precipitation, irrigation and the type of vegetation or
crop. One way of dealing with this for irrigated crops is by
assuming a constant leaching fraction (that is the fraction of
irrigation water that passes the root zone). That way the soil
moisture salt stress can be expresses as a salt stress due to
chloride concentration in irrigation water. This is illustrated in
Figure 9.
Because the Lake Wieringen will provide the agricultural
water supply, the results from the variable density modelling
were used in SOBEK and solute-balance calculations to
investigate the impact of the lake on the surface water quality.
Surface water quality has been studied in terms of expected N-
P- and Cl-concentrations in 2030. It appeared that polder
Waard-Nieuwland, at the low lying part of Wieringen, plays an
important role in the surface water quality of Lake Wieringen.
In the current situation water is discharged from polder
Waard-Nieuwland to the Amstelmeerkanaal (figure 10). In the
dry season, water from the Amstelmeerkanaal is used to supply
the Wieringermeerpolder.
Calculations indicated that if the additional salt-load from
the lower areas of Wieringen caused by the autonomous
development is discharging directly to the Lake Wieringen, the
level of salinity of the surface water will increase significantly,
and it will be unusable for agricultural water supply. This
information was directly used to optimise the design of the
surface water system through deflection of polder discharge
aiming to reduce surface water salinity (figure 10).
The calculations show that the increasing salinity levels can
mitigated through optimising the surface water system. This is
efficient in the area because currently there is already an
extensive fresh water infrastructure present in the region.
VII. CONCLUSIONS
Presently, very advanced regional modelling tools are
available to simulate groundwater flow and groundwater
quality. The calibration of these models is sometimes difficult
when considering the very few groundwater quality data that is
available. This means that the reliability of model predictions
are hard to verify.
The reliability of the simulated salinity changes assessed
through a comparison with hydrochemical data. A reasonable
resemblance is found between the salinity trends inferred from
the hydrochemical data and the SEAWAT simulations. The
model predictions are however not confirmed by the trends in
surface water salinity. This may be caused by an increase in
0
10
20
30
40
50
60
70
80
90
100
0 1000 2000 3000 4000 5000
Chloride concentration of irrigation water (mg/L)
salt stress (%)
orchards tulips fruit trees gre enhouses potatoes
corn grass grain sugerbeet
Fig 9: Salt stress functions (Alterra, 2003)
Hippolytus-
hoeverkoog
Waard
Nieuwland
Ooster-
lander-
koog
Gester/
Oeverkoog
Amstelmeerkanaal
Amstel
-
meer
Zuider-
haven
(IJssel
meer)
Wadden-
zee
Wieringermeerpolder, afdeling
1+2
1
2 3 4
6
Wester-
lander
koog
Hoelmer-
koog
Surface water management in the present situation
Wadden-zee
Hippolytus-
hoeverkoog
Ooster-
lander-
koog
Gester/
Oeverkoog
Amstel-
meer
Zuider-
haven
(IJssel
meer)
Wieringermeerpolder,
afdeling 1+2
Lake Wieringen
6
1
23
Waard
Nieuwland
8
Hoelmer-
koog
Wester-
lander
koog
Surface water management in the future situation
Fig 10: Surface water system in present and future situation. Arrows
indicate direction of surface water flow. The symbols next the numbers
indicate polder water pumping stations.

The application of the SEAWAT variable density code for the Lake Wieringen project, the Netherlands 8
flushing of the surface water system, to counteract the
increasing salinity of deeper groundwater.
In order to use the results of the numerical modelling to
assess the effects for agriculture and nature, WaterNOOD and
a surface water solute balance were employed. Direct
translation of the results from numerical groundwater
modelling in terms of crop yields and effects for nature is not
possible. This is due to the fact that the effects on these land
use functions are mainly dependent on salinity changes near
the ground surface. This means that the use of variable density
modelling on its one, is currently too much an
oversimplification to realistically assess the effects of
groundwater salinisation.
REFERENCES
[1] Alterra (2003) Actualisatie van de zouttolerantie van land- en
tuinbouwgewassen ten behoeve van de berekening van de zoutschade in
Nederland met het RIZA-instrumentarium.
[2] Dienst der Zuiderzeewerken (1936) Geohydrologische gesteldheid van
de Wieringermeer.
[3] Grontmij (2001) Geohydrologisch onderzoek Wieringerrandmeer.
[4] Harbaugh, A.W. and M.G.McDonald (2000) MODFLOW 2000, the
U.S.G.S. modualr ground-water model. Open file report 00-92.
[5] Langevin, C.D., W.B. Schoemaker, W. Guo (2003) MODFLOW-2000,
the U.S. Geological Survey Modular Ground-Water Model—
Documentation of the SEAWAT- 2000 Version with the Variable-
Density Flow Process (VDF) and the Integrated MT3DMS Transport
Process (IMT).
[6] NW4, Werkgroep Klimaatverandering en bodemdaling (1997).
Klimaatverandering en Bodemdaling. Onderzoeksrapport in het kader
van de voorbereidingen voor de vierde Nota Waterhuishouding.
[7] Oude Essink, G.H.P. (2002) Salinization of the Wieringermeerpolder,
the Netherlands; Proc. 17th Salt Water Intrusion Meeting, Delft, The
Netherlands
[8] Post, V.E.A. (2004) Groundwater salinization processes in the coastal
area of the Netherlands due to transgressions during the Holocene,
Ph.D. thesis, Vrije Universiteit Amsterdam. ISBN 90-9017404-4.
[9] TNO-NITG (2005) Van Gidslaag naar Hydrogeologische eenheid;
Toelichting op de totstandkoming van de dataset REGIS II.
[10] STOWA (2003) WaterNOOD instrumentarium handleiding V1.1.
Kenmerk 302/OF3/4C1/000782/LE d.d. 10-12-2003.
[11] Stuyfzand, P.J. (1999). Patterns in groundwater chemistry resulting from
groundwater flow. J. Hydrol. 7
[12] Van de Ven, editor (2003) , Liveable Lowlands, History of water
management en land cultivation in the Netherlands. ISBN 90-5345-190-
0. Matrijs publisher, Utrecht (in Dutch).
[13] Witteveen+Bos (2005) MKBA Wieringerrandmeer
[14] Witteveen+Bos (2006a) Integraal Effectrapport Wieringerrandmeer
[15] Witteveen+Bos (2006b) Geohydrologisch onderzoek
Wieringerrandmeer
[16] Zheng, C., Wang, P.P. (1999) MT3DMS: A Modular Three-
Dimensional Multispecies Transport Model for Simulation of
Advection, Dispersion, and Chemical Reactions of Contaminants in
Groundwater Systems; Documentation and User’s Guide. Contract
Report SERDP-99.