TRU, Alm.del - 2021-22 - Bilag 241: Orientering om Deltares' uafhængige tredjepartsgranskning af de hydrauliske analyser af Lynetteholms påvirkning af vandgennemstrømningen i Øresund, fra transportministeren

Transportudvalget 2021-22
TRU Alm.del Bilag 241
Offentligt

DRAFT

Transportudvalget 2021-22

TRU Alm.del - Bilag 241

Offentligt

Independent review of the

Hydrodynamic Studies on the impact of

Lynetteholm on exchange of water and

salt through Øresund

TRU, Alm.del - 2021-22 - Bilag 241: Orientering om Deltares' uafhængige tredjepartsgranskning af de hydrauliske analyser af Lynetteholms påvirkning af vandgennemstrømningen i Øresund, fra transportministeren

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on

exchange of water and salt through Øresund

2 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange of

water and salt through Øresund

Client

Contact

Reference

Keywords

Land reclamation, Lynetteholm, exchange flow, Øresund, environmental impact, Baltic Sea

By & Havn

Document control

Version

Date

Project nr.

Document ID

Pages

Classification

Status

0.1

18-03-2022

11207757-002

11207757-002-HYE-0001

Confidential until further notice

draft

This is a draft report, intended for discussion purposes only. No part of this report may be relied upon by

either principals or third parties.

Author(s)

Doc. version

0.1

Author

Arnout Bijlsma

Reviewer

Wouter Kranenburg

Firmijn Zijl

Approver

Dirk-Jan Walstra

Publish

3 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Summary

This review of the Hydrodynamic Studies on the impact of Lynetteholm on the exchange of

water and salt through the Øresund evaluates the report of the Hydrodynamic Studies by DHI

and additional material provided in response to questions to By & Havn and DHI, considering

the methodology, the set-up and verification of the modelling, the interpretation of the model

results and the conclusions with respect to the exchange of water and salt.

The key results of the Hydrodynamic Studies on the impact of Lynetteholm on the exchange

of water and salt through the Øresund are the estimate of the blocking effect for flow and salt

at Drogden Sill of -0.19% and -0.24% for Main Proposal 1 and 2 of the Lynetteholm land

reclamation, respectively, with a 95% confidence range of ±0.12%. These results, indicating a

reduced exchange, are based on a relative approach in which the Main Proposals 1 and 2 of

Lynetteholm are compared to the reference situation using a 3D numerical model of the

Øresund and the conditions of 2018. The 95% confidence range of ±0.12% for Lynetteholm is

half of the ±0.25% confidence range established in the Øresund Link modelling.

The main findings of the review are:

The blocking effect defined in the Hydrodynamic Studies is an appropriate measure

for changes in the exchange of water and salt through the Øresund. (Note that a

negative blocking effect means that the exchange is weakening).

The approach based on the Øresund model with fixed boundary conditions is valid as

long as Lynetteholm does not affect the hydrodynamic conditions at the locations of

the open boundaries.

The set-up and calibration of the models is not well documented, but the horizontal

and vertical grid resolution and the type of model (software, physical processes) are

appropriate for the purpose.

The verification of the models is limited, e.g. a verification of currents or transports

trough the Øresund is missing. The additional comparison during the review of the

computed exchange of water and salt through the Øresund to global numbers from

literature supports that the model and the selected conditions based on 2018 are

appropriate for the evaluation of the blocking effect of Lynetteholm.

Therefore, the blocking effect of -0.19% and -0.24% resulting from the Hydrodynamic

Studies for Main Proposal 1 and 2 at Drogden Sill is

considered realistic.

Based on

the review we find a 95% confidence range of ±0.25% more reasonable.

It is outside the scope of the review to evaluate whether the blocking effects found are

acceptable and/or negligible or not.

In case the above estimates of the blocking effect and confidence range are considered not

fully acceptable and/or negligible further substantiation might be needed. For this the

following suggestions can be made.

Since the approach to estimate the impact of Lynetteholm on the exchange of water

and salt assumed that Lynetteholm does not influence the boundary conditions of the

Øresund model, it would be useful to learn to which degree this estimate is

conservative or not, either from previous studies or literature, or from a sensitivity

study with a numerical model of the Baltic Sea and the Danish Straits.

It could be useful to improve the reporting of model set-up, calibration and

verification. This could help to provide more confidence in the quality of the models.

Particularly, the verification of the transports in the Øresund model should be

considered. Such a verification for 2018 or for another year could enhance the

confidence in the model performance, e.g. by showing that calibration on water levels

4 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

and water level gradients and certain aspects of salinity is sufficient to obtain correct

water and salt transports.

A 95% confidence range smaller than ±0.25% is perhaps achievable but would

require further substantiation.

5 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Contents

Summary

1.1

1.2

1.2.1

1.2.2

1.2.3

1.2.4

2.1

2.2

2.3

3.1

3.2

3.3

3.4

3.5

4.1

4.1.1

4.1.2

4.1.3

4.1.4

4.1.5

4.1.6

4.1.7

4.2

4.2.1

4.2.2

4.2.3

4.2.4

4.2.5

4.2.6

4.2.7

4.3

Introduction

Background

Set-up of the review

Aim and scope

Received input

Approach and reading guide

Limitations of the review

Discussion of problem definition

Area of interest and proposed layouts of Lynetteholm

Water and salt exchange through the Danish Straits

Discussion and research questions

Evaluation of the methodology

Numerical modelling approach

Software

Blocking effect

Interpretation of blocking effect

Conclusion

Evaluation of model set-up

DKBS2 model

Computational mesh

Bathymetry

Simulation period and initial condition

Atmospheric forcing and boundary conditions

Physical input parameters

Numerical input parameters

Calibration

Øresund model

Computational mesh

Bathymetry

Simulation period and initial condition

Atmospheric forcing and boundary conditions

Physical input parameters

Numerical input parameters

Calibration

Conclusion

Evaluation of model verification

6 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

5.1

5.1.1

5.1.2

5.1.3

5.1.4

5.1.5

5.2

5.2.1

5.2.2

5.2.3

5.2.4

5.2.5

6.1

6.2

6.3

7.1

7.2

Results of DKBS2 model verification

Water levels

Currents

Salinity and temperature

Water and salt exchange through the Danish Straits

Conclusion

Results of Øresund model verification

Water levels

Currents

Salinity and temperature

Water and salt exchange through Øresund

Conclusion

Evaluation of the impact of Lynetteholm on water and salt exchange

Blocking effect for water exchange

Blocking effect for salt transport

Results in perspective

Synthesis of findings

Discussion of impact of Lynetteholm

End conclusion

References

7 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

1.1

Introduction

Background

Lynetteholm is a development project of By & Havn involving an artificial island off the coast

of Copenhagen, see Figure 1.1. The project involves a land reclamation of 275 ha in the

Øresund area. For the project an environmental impact assessment (EIA) has been

performed, see [1]. The Hydrodynamic Studies for the EIA have been carried out by DHI.

Figure 1.1

Artist impression of the Lynetteholm land reclamation in the center with Copenhagen in the

foreground and the Øresund with the Swedish coast in the background (Source: By & Havn).

In a presentation given by By & Havn to Deltares the result of the Hydrodynamic Studies was

summarized in short as follows:

The environmental impact assessment (EIA) of Lynetteholm, documents that

Lynetteholm will influence the exchange of water and salt through Øresund.

The reduction in exchange flow (the blocking) is calculated to approximately 0.25%.

The impact is negligible compared to existing fluctuations in the Baltic Sea.

In time the blocking effect will be compensated by effects of climate change.

Since it has been concluded that the artificial island Lynetteholm will affect the exchange flow

between the North Sea and the Baltic Sea through the Øresund, meetings have been held

with Danish and Swedish Authorities on the environmental impact of Lynetteholm under the

Espoo Convention 1991, as required by international and Danish legislation. Under the

Espoo Convention or by other legislation no formal and objective criteria or requirements are

in place for acceptance of the results of the transboundary environmental impact.

We further understand that the results of the Hydrodynamic Studies presented in the Espoo

meetings gave rise to discussion. This resulted in the request for an independent review of

which the scope was agreed upon in the Espoo meetings. The present report describes this

independent review carried out by Deltares.

8 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

1.2

1.2.1

Set-up of the review

Aim and scope

The objective of the independent review is to evaluate the methodology, the execution of the

modelling, the interpretation of the results and the conclusions drawn in the Hydrodynamic

Studies with respect to the impact of Lynetteholm on the exchange of water and salt through

Øresund.

In the presentation by By & Havn the scope of the review was defined as follows:

1. Impact of Lynetteholm on exchange of water and salt through the Øresund (Blocking

effect).

2. The calculation grid (bathymetry, location of the north and south edges of the model,

resolution of the calculation grid, consistency in the setups in order to eliminate the

effect of numerical noise…).

3. The driving forces (boundary conditions for salt, temperature, water level and current,

waves, wind…) including whether the uncertainty in the model's boundary data is

assessed in the final conclusion on environmental impacts.

4. The selected time period is representative and sufficient to accommodate long-term

changes in current conditions.

5. The quality of calibration and validation of the model in the Sound (current, water

level, salt, temperature…).

This scope was agreed between the Swedish and the Danish Authorities.

1.2.2

Received input

On 8 and 11 October 2021 the following reports were received by Deltares from By & Havn:

DKBS2 Hydrodynamic Model Setup and Validation.

DHI Technical Note (translated

from Danish), see [2].

Construction of Lynetteholm, EIA – Technical Background Report No 1, Hydraulic

Surveys

(Selected Chapters).

DHI Report, (Translation of parts from report in Danish

of 2 November 2020), see [3].

Anlæg af Lynetteholm, VVM – Teknisk Baggrundsrapport nr. 1, Hydrauliske

undersøgelser.

Complete DHI report in Danish, see [4].

Of the DHI reports provided by By & Havn, the first report [2] (‘DKBS2 report’) describes the

model setup and validation of the updated large scale numerical model of the Belt Sea and

the Baltic Sea, the DKBS2 model. This model is applied in hindcast and forecast mode as

part of DHI’s Water Forecast service. The model has been applied for the period 2008 – 2017

(10 years) to demonstrate the ability of the model to simulate water level variations,

circulation and stratification in the system.

The second report (‘Øresund report’), see [3] and [4], is the most important report for the

review. This report describes the modelling work carried out to estimate the impact of two

versions of the Main Proposal for Lynetteholm on the local hydrodynamic conditions (water

levels, currents, salinity, temperature, waves) and the exchange of water and salt between

the Baltic Sea and the Kattegat through the Øresund. This has been done by comparing the

computational results of a detailed Øresund model for the two Lynetteholm variants and the

existing situation for the year 2018. The boundary conditions for 2018 were obtained with

help of the DKBS2 model. The impact on the exchange of water and salt through the

Øresund is presented in terms of a blocking parameter, determined from the computational

results.

or ‘Hydrodynamic Studies’

9 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

The report also describes the discharge and mixing of surplus water from the borrow area

and the environmental pollutants emitted during the construction and operational phases, as

well as the effects of waste during digging associated with the replacement of the gyttja-

containing sediments found along the entire outer perimeter. However, this review only

considers the aspects related to the exchange flow.

The present review addresses the exchange of water and salt through the Øresund in the

operational phase. The DHI reports focus on the result of the model simulations. For the

review a clear picture of the underlying question and context regarding the impact on the

exchange of water and salt through the Øresund is also needed. Some additional information

was found in the Environmental Impact Report [1] and in a text book on the Baltic Sea [5].

Furthermore, during the review additional questions were asked to By & Havn and DHI on the

methodology followed to address the question, and on various aspects of the modelling and

the analysis of the results. In this respect the following additional information has been

received and included in the review:

An update of Figure 6-126 in [3] and [4] to cover the entire year 2018 instead of the

first 6 months (19 January 2022), see [7].

An Excel file with an additional blocking analysis for flow and salt for the most

decisive cross-section, the Drogden Sill, by combining the data of the East and West

Peberholm sections (21 January 2022), see [8].

A memo addressing the representativity of the evaluation period of 2018 and

providing further background on the blocking effect and further details on the set-up

and calibration of the Øresund model (24 January 2022), see [9].

All input received in Danish has been assessed using general available translation

functionality from MS Word or Google Translate. Although this method of translation might

have its limitations, it was considered sufficient for the purpose of the review.

1.2.3

Approach and reading guide

With the aim and scope mentioned above, the independent review of the Hydrodynamic

Studies on exchange of water and salt through the Øresund addresses the following subjects:

The evaluation of the methodology

The evaluation of the set-up of the models

The evaluation of the verification of the models

The evaluation of the impact of Lynetteholm on the exchange of water and salt

through the Øresund.

Each subject is closed with a conclusion section. The review ends with a synthesis of findings

and overall conclusions.

A clear problem definition is essential for conducting a proper review. Therefore, we will first

discuss in Chapter 2 the extent to which the Lynetteholm plans affect the geometry of the

area of interest, and the physical processes that might be influenced by these plans. This

leads to the problem definition, which forms the basis of the review.

In Chapter 3 the methodology including the general modelling approach based on the DKBS2

and the Øresund models and the blocking effect as a measure for the influence of

Lynettehom is reviewed.

The review of the set-up of the DKBS2 model and the Øresund model in Chapter 4 aims to

verify whether key aspects like the model grid resolution, model bathymetry, (location of)

open boundary conditions, etc. are suitable for the purpose of the study. Furthermore, the

10 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

substantiation of the physical and numerical input parameters will be verified on the basis of

the reports and additional enquiries.

In the reports of the Hydrodynamic Studies the performance of the calibrated models is

verified in simulations of which the results have been compared with measurements. The

quality of the verification of the models for the reproduction of currents, water levels, salinity

and water temperature will be inspected. Lastly, the fitness of the models for the intended

(relative) approach will be evaluated. These results will be evaluated as presented and

described in the reports, see Chapter 5.

In Chapter 6 the determination of the impact of Lynetteholm on the water and salt exchange

through the Øresund is reviewed.

The review is concluded with a discussion of the various intermediate findings and the

assessment of their relevance for the end result of the study. Based on this, a final conclusion

on the validity of the outcome of the Hydrodynamic Studies with reference to the exchange

flows through the Øresund will be given, together with some recommendations, see Chapter

1.2.4

Limitations of the review

The review of the Lynetteholm Hydrodynamic Studies is limited to the exchange of water and

salt through Øresund. In the review, the models and the results of the simulations have been

evaluated as presented and described in the reports, i.e. without access to the original input

and output files of the numerical models. Furthermore, no new simulations have been carried

out for this review by DHI, or by Deltares. Finally, since By & Havn indicated that no legal

criteria for acceptance of effects on the exchange flow exist, the review does not address the

question whether the resulting effects of Lynetteholm on the exchange of water and salt

through the Øresund are acceptable or not. This is beyond the scope of the review.

11 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Discussion of problem definition

Good insight in the setting of the problem is essential for the review. Therefore, we will first

describe our understanding of the way in which the Lynetteholm plans affect the geometry of

the area of interest, and which (physical) processes might be influenced by these plans. This

provides the basis for the problem definition. This is especially relevant since the problem

definition is not explicitly described in the reports provided.

2.1

Area of interest and proposed layouts of Lynetteholm

The area of interest and proposed layouts of the artificial island Lynetteholm are described in

the Øresund report [3]. Lynetteholm is located in the Øresund on the east side of

Copenhagen. The Øresund is the eastern of the three Danish Straits that connect the Baltic

Sea to the North Sea. A bathymetry map of the Øresund with geographical names is given in

Figure 2.1. The eastern part, including Ven island, belongs to the Swedish territorial waters.

The western part of the Øresund, including Santholm and Lynettehom, is part of the Danish

territorial waters. The shallowest cross-section, Drogden Sill, is located south of Santholm,

and is intersected by the channels of Drogden and Flinterenden. Lynetteholm will block the

Kongedybet channel, the smaller of two channels north of Drogden, see Figure 2.2. This is

Figure 2.1 Øresund model area with bathymetry and geographical names (Source: Figure 4-1 in [3])

12 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

illustrated in more detail in Figure 2.3 where the contours of Main Proposal 1 and 2 of

Lynetteholm are shown on top of the bathymetry. Main Proposal 1 is a reclamation without a

coastal landscape (pink curve) and Main Proposal 2 is a reclamation containing a coastal

landscape (red curve). The difference between the investigated layouts (Figure 2.3) and the

final layouts (Figure 2.4) is relatively small and does not affect the blocking of Kongedybet [3].

The depths in the figures are given with reference to DVR90, the Dansk Vertikal Reference

1990.

Figure 2.2 Bathymetry with channels, banks and islands in the southern part of the Øresund (Source: Figure

4-2 in [3]).

Figure 2.3 Indication of investigated layouts for Main Proposal 1, a reclamation without a coastal landscape

(pink curve) and Main Proposal 2, a reclamation containing a coastal landscape (red curve). (Source: Figure

3-3 in [3]).

13 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Figure 2.4 Indication of final layouts for Main Proposal 1 (pink curve) and Main Proposal 2 (red curve). The

red curves show the top of the coastal profile, the water line and where the foot of the profile reaches the

natural seabed. (Source: Figure 3-3 in [3]).

2.2

Water and salt exchange through the Danish Straits

Through the Øresund and the other Danish Straits (Little Belt and Great Belt) water is

exchanged between the North Sea and the Baltic Sea. These exchange flows are extensively

described in a general textbook on the physical oceanography of the Baltic Sea [5]. The

exchange flows are a determining factor for the environmental conditions in the Baltic Sea.

Especially the irregularly occurring major inflows of oxygen rich and saline water are

important with respect to eutrophication, the most serious environmental problem in the Baltic

Sea. In view of this, a concise description of the exchange flows through the Øresund and the

other Danish Straits based on [5] has been included in the review for further reference.

Table 2.1 describes the global water and salt balance of the Baltic Sea, based on Figure 4.1

and 4.9 in [5]. Generally, the outflow estimated at about 1660 km

/yr is 480 km

/yr larger than

the inflow estimated at 1180 km

/yr, due to the river discharges and the net result of

precipitation and evaporation in the Baltic Sea. Typically, the exchange flows are much larger

Table 2.1 Global water and salt balance Baltic Sea

Water Balance

(km

/yr)

Rivers

Precipitation -

Evaporation

Inflow Danish Straits

Outflow Danish Straits

440

40 (215- 175)

1180

1660

Volume (km

)

Baltic Sea

21205

Salt content (G ton)

159

Salinity (psu)

7.5

Salt Balance

(G ton/yr)

Salinity

(psu)

14 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

than the net outflow. The Øresund contributes to about 25-30% [5] (3/11 or 27% in [2]) to the

total exchange flow through the Danish Straits.

On the long-term, the inflow and outflow of salt are in balance at approximately 30 G ton of

salt per year. This corresponds to almost 1/5 of the salt content of the entire Baltic Sea each

year. The mean salinity at inflow is estimated at 25 psu, which is higher than the mean

salinity at outflow estimated at 18 psu, corresponding to the difference between inflow and

outflow volumes. The freshwater input to the Baltic Sea varies over the year and is largest

from April – July. This results in a seasonal variation of the exchange flows. The major

inflows therefore usually occur from September – January. Recent Major Baltic Inflow (MBI)

events have been reported for January 1993, January 2003, see [5] and an extreme event in

December 2014, see [10]. In 1993 and 2003 the inflow was about 310 and 200 km

for Little

Belt, Great Belt and Øresund combined, see Par.5.4.3. in [5]. It was estimated that 0.85 G ton

salt flowed in via the Øresund during the 2003 MBI, 42% of the total salt transport during this

event. For the 2014 event the estimates are 320 km

for the total inflow volume, of which 198

had a high salinity, and a total salt inflow of 4 G ton, see [11]. Of this 60 km

high salinity

water (30%) and 1.38 G ton salt (35%) flowed through the Øresund.

As described in [5], the outflow thought the Straits exists of low-density surface water from

the stratified Baltic Sea, while high salinity and dense North Sea water flows toward the Baltic

Sea in the bottom layer. Strong vertical mixing occurs in the Straits. Therefore, at the sills like

Drogden Sill in the Øresund, no steady two-layer flow exists. On the contrary, the flow on the

sills is predominantly barotropic and depending on the water level difference and the wind.

Due to vertical mixing the inflowing water to the Baltic Sea is less saline than North Sea

water, and the outflowing water is more saline than the Baltic surface water. The strength of

the inflow, the duration of the events and the mixing in the Straits determine the amount of

salt that enters the Baltic Sea. A large part of the total inflow of salt is caused by more regular

occurring moderate events. Major inflows, occurring on average once in 10 years, also play

an important role, however.

The time-scales of the hydrodynamic processes involved are quite long, given a typical

residence time of about 30 years for the Baltic Sea [5] and the occurrence of major inflow

events on a decadal scale. This means that in principle any changes in the Danish Straits

that influence the exchange flow will take more than 30 years to reach a new (dynamic)

equilibrium.

On a somewhat longer time-scale, in the order of 100 years, climate change is expected to

influence the environmental conditions in the Baltic Sea as well. Several aspects related to

climate change may affect the hydrodynamic processes either directly, or indirectly via

changes in the exchange flows through the Danish Straits, for example: relative sea level

rise, changes in fresh water run-off to the Baltic Sea, changes in temperature and changes in

the local wind climate, see [5]. Such changes may also affect the occurrence of major inflow

events. The influence of the changes on the thermohaline circulation of the Baltic Sea can be

explored by numerical modelling but it is understandable that the uncertainty in such studies

is fairly large. Furthermore, the assessment of environmental effects of climate change would

require a more integral approach of hydrodynamics, water quality and ecology as other

processes than the hydrodynamics are probably also affected.

2.3

Discussion and research questions

First of all, in the Hydrodynamic Studies the problem of the assessment of the impact of

Lynetteholm on the Baltic Sea environment has been simplified by using the impact on the

exchange of water and salt through the Danish Straits as an approximation for the

environmental effect on the Baltic Sea. In this review this is regarded as a starting point and

not evaluated further.

15 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

In the Øresund the Drogden Sill is the main bottleneck for the exchange flow, as described in

the previous Section. According to the proposals, Lynetteholm will block the Kongedybet, one

of the two channels west of Santholm that lead to Drogden Sill from the north. It is

conceivable that this blockage may for instance locally influence the saline inflow to the Baltic

Sea. On the other hand, the exchange flows east of Santholm, and further away in the Little

Belt and the Great Belt, will not be hindered and may even partly compensate any blockage

that may occur in the channels west of Santholm. From that perspective it is considered

appropriate to study at least the exchange flows throughout the entire Øresund.

If the impact of Lynetteholm on the exchange of water and salt through the Øresund would be

negligible, there is no reason to expect influence on the exchange through the other Danish

Straits and on the long-term response of the Baltic Sea circulation. For this reason these

aspects may be disregarded and the Hydrodynamic Studies may focus on the exchange flow

through the Øresund only, and justify these limitations afterwards when proven that the

influence of Lynetteholm is negligible indeed.

Given the variability of the exchange flows it will be necessary to select a natural period or

perhaps an artificial sequence of conditions which is sufficiently representative for the

exchange phenomena occurring for the situation that Lynetteholm is operational, like average

transports, seasonal variations, regular moderate inflow events, and exceptional inflow

events. It is also necessary to evaluate the effects on both water and salt transport as these

may differ in principle, and to address the accuracy of the estimated effects.

The investigation in the Hydrodynamic Study should therefore answer the following research

questions:

What is a representative natural period or an artificial sequence of conditions to

investigate the effect of Lynetteholm on the exchange of water and salt through

Øresund?

How large and variable is the water and salt exchange through the Øresund under

the selected conditions for the existing layout (without Lynetteholm)?

How large and variable is the water and salt exchange through the Øresund under

the selected conditions for Main Proposal 1 and 2 of Lynetteholm?

What is the effect of Lynetteholm, based on the differences in the exchange of water

and salt in comparison to the reference situation, and how accurate is this?

A judgement whether the found differences are acceptably small or not, is not part of the

review. Nevertheless, attempts that have been made in the reports provided by By & Havn to

put the resulting differences into perspective will be commented.

16 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Evaluation of the methodology

The next step in the review is to evaluate whether the methodology including the general

modelling approach is able to meet the objective of the Hydraulic Studies and answer the

research questions with respect to the effect of Lynetteholm on the exchange flows through

the Øresund. The methodology has not been explicitly described in the reports and therefore

the description below is derived from the reports [2], [3] and [4] and memo [9]. Any review

remarks are clearly indicated as such. The chapter ends with a short evaluation of the

methodology applied.

3.1

Numerical modelling approach

The impact of Lynetteholm on the exchange of water and salt through the Øresund has been

investigated by numerical modelling of the hydrodynamics in this area. The focus is on the

modeling of the hydrodynamics in the Øresund, see [3]. We note that this makes it possible to

model the exchange flows in much more detail than would be possible when the other straits

and the entire Baltic Sea were included since that involves a much larger area and very long

simulations periods. Strictly speaking this approach is only valid when the influence of

Lynetteholm is negligible at the locations of the open boundaries, see the argument in

Section 2.3. Some compensation of any adverse effects on the exchange flow in the

channels west of Santholm via the channel east of Santholm is possible, however.

Furthermore, the effects have been determined via a relative approach: for selected

conditions the results of simulations for two proposals of the Lynetteholm layout have been

compared to the results of a simulation for the existing situation (without Lynetteholm) under

the same conditions [3]. We note that the advantage of this relative approach is that

uncertainties in the simulations (e.g. model set-up, conditions) have less effect on the

accuracy of the impact of Lynetteholm which is being estimated.

All simulations have been performed under the same conditions, which means that the same

initial condition, atmospheric forcing, boundary conditions and river discharges haven been

applied. The initial and boundary conditions have been derived from a larger scale model, the

operational three-dimensional model of the Baltic Sea, Belt Sea, Kattegat and Skagerrak, the

DKBS2 model [3]. Water level measurements along the Øresund have been used to improve

the water level boundary conditions, see [9].

For the representative natural period or an artificial sequence of conditions to investigate the

operational phase of Lynetteholm (see Section 2.3) the year 2018 was selected, a full recent

year to cover all seasonal variations due to varying river discharges and the annual cycle in

water temperatures, see Par.4.2.4 in [3]. A more detailed explanation for the selection of

2018 as a representative year to investigate the blocking effect was given in [9], considering

the annual net flow, the occurrence of many weak and stronger inflow and outflow events,

including two small or medium size MBI’s. Furthermore, it was noted in [9] that by considering

a full year the resulting blocking effect is less sensitive for the actual period selected,

compared to the 72 day’s simulations carried out at the time in the studies for the Øresund-

Link. We note that ideally, an extreme MBI, e.g. of December 2014, would also have been

part of the simulations.

Although the operational phase of Lynetteholm is somewhere in the future, we accept the

approach to select the conditions of a recent year for all simulations. This year should contain

the relevant physics to a sufficient degree (see further our evaluation in Section 5.2.4).

17 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

The system used to investigate the effects of Lynetteholm on the exchange of water and salt

through the Øresund is defined in [3] by:

The geometry (bed level) in an area covering the entire Øresund from the Gilleleje –

Kullen cross-section in the north to the Stevns – Skanör cross-section in the south

and including Copenhagen Harbour as shown in Figure 2.1.

The 3D hydrodynamic processes governing the development of water levels,

currents, salinity and temperature.

The driving forces existing of water levels (tides, surges), currents, salinity and water

temperature at the open boundaries, and winds, atmospheric pressure gradients,

atmospheric heating and cooling.

According to DHI the cooling water recirculation of the Amager Power Station (Par. 4.3 and

6.2 in [4]), which is affected by Lynetteholm, was not included in the investigation of the

blocking effect. Since the changed cooling water recirculation may in principle have some

effect on the density currents in the channels west of Santholm, it would have been

appropriate to motivate the neglect of this process.

3.2

Software

The Hydrodynamic Studies are carried out using the MIKE 3 software, see [2] and [9]. For the

DKBS2 model also the version is documented (version 2017), see [2]. For the Øresund

model this information is not reported.

The DKBS2 model is applied with MIKE 3 in hydrostatic mode (Par 2.1 in [2]). We assume

that the same holds for the Øresund model.

3.3

Blocking effect

The key parameter in the Hydrodynamic Studies to evaluate the impact of Lynetteholm on the

exchange of water and salt through the Øresund is the blocking effect (dq) defined as

∑(|

−

∑|

with being the water transport (m

/s) through a vertical cross section (Drogden Sill), index

referring to the situation including Lynetteholm, and index

referring to the baseline or

reference situation, see Par 6.1.6. in [3]. The summation is carried out over the entire

evaluation period (2018). A similar formula holds for the blocking of the salt transport (kg/s).

According to [9] this approach has been successfully applied in the studies for the Øresund-

Link.

For interpretation purposes we rewrite the blocking effect in terms of average inflow of water

= 2

with summation

and salt and net outflow of water. Using

∑|

period

the blocking effect for flow can be expressed as

(

−

+ 0.5

)

in which the overbar indicates long-term averaging,

stands for the inflow of water,

for

Lynetteholm and

for Baseline.

is the net outflow through Øresund, assumed for

simplicity independent of the situation with regard to Lynetteholm.

In a similar way the blocking effect for salt can be expressed as

−

18 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

in which

stands for the inflow of salt. The average net salt transport is assumed to be

zero

under current conditions.

These expressions confirm that the blocking effect as defined in the Hydrodynamic Studies is

an appropriate measure for changes in the exchange of water and salt through Øresund.

Note that a negative blocking effect means that the exchange is weakening, and a positive

blocking effect that the exchange increases.

In the Hydrodynamic Studies, the integration or summation has been carried out for a full

calendar year (2018), averaging the effects of the seasonal cycle. Spin up effects at the

beginning of the simulation should be excluded from the evaluation. Furthermore, upon

request the blocking effect is evaluated at Drogden Sill [8], where the flow is considered

generally barotropic, meaning that locally the density currents can be neglected. Accuracy

estimates of the blocking effect are derived from the Øresund Link studies [3], [9].

3.4

Interpretation of blocking effect

In an effort to put the computed effect of the Lynetteholm land reclamation on the water and

salt exchange thought the Øresund into perspective, DHI made a comparison with estimates

of the effect of the expected sea level rise and with the blocking criteria applied for the

Øresund Link.

The sea level rise was implemented in the Øresund model by simply increasing all water level

initial and boundary conditions by a few centimeters. All other aspects of the model set-up

remained unchanged. Basically, this means that the model zone existing of sigma-layers

including initial and boundary conditions was stretched a little bit. Furthermore, the duration of

the simulation was limited to the first half year of 2018, according to the plots in [3]. Next, the

period in which the blocking effect by Lynetteholm is approximately neutralized by the effect

of sea level rise is estimated on the basis of a sea level rise of 1.55 mm /year.

As mentioned in Section 2.2, sea level rise is one aspect of climate change, and other

changes, e.g. in North Sea conditions and wind climate may also affect the exchange of

water and salt in the Danish Straits. Without further motivation or background information we

have our doubts on the validity of this approach. In Section 6.3 we provide further comments

on this approach.

In the discussion of the results in reports [3] and [4] the final conclusion is based on

comparison with the accuracy of the ‘Zero Solution’ that was required for the Øresund Link

(apparently described in [6]). We think that this can be of use, if the requirements of that time

are still valid, and e.g. confirmed by later long-term monitoring (from 2000 – present). This will

also be discussed further in Section 6.3.

3.5

Conclusion

The methodology to estimate the impact of Lynetteholm on the exchange of water and salt

through the Danish Straits by numerical modelling of the hydrodynamics in only the Øresund

area is acceptable when it proves that the influence of Lynetteholm on the exchange through

Øresund is negligible. Some compensation of adverse effects on the exchange flow in the

channels west of Santholm is possible via the channel east of Santholm.

The impact of Lynetteholm on the exchange of water and salt through the Øresund is based

on the determination of the blocking effect of Lynetteholm from simulations of the situation

This holds in principle for all Danish Straits combined, see Section 2.2. Individual straits may differ somewhat.

19 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

with and without Lynetteholm for representative conditions. For these conditions the full year

of 2018 has been selected. The simulations have been carried out using the MIKE 3

software. The review confirmed that the blocking effect as defined in the Hydrodynamic

Studies is an appropriate measure to quantify changes in the exchange of water and salt

through Øresund. Furthermore, for the accuracy estimate of the blocking effect reference is

made to the results of the Øresund Link studies.

In the Hydrodynamic Studies, two methods have been presented to provide some

perspective on the computed blocking effects.

Firstly, in the Hydrodynamic Studies the blocking effects computed for Lynetteholm have

been compared to requirements that were in place for the Øresund Link. We consider that

this can be useful when it can be shown that such requirements are still valid.

Secondly, simulations have been carried out for sea level rise scenarios. For the reference

situation and the two Main Proposals the water level was simply increased by 2 cm.

Assuming a rate of sea level rise of 1.55 mm /year, the period in which the blocking effect by

Lynetteholm is approximately neutralized by the effect of sea level rise, is estimated. We

have strong reservations on the validity of this approach since sea level rise is only one

aspect of climate change. In principle the relevance of sea level rise should be evaluated in

comparison to other changes like those in the tide, temperature and salinity in the North Sea,

and wind climate, precipitation, river run-off, and even water quality and ecological processes

in the Baltic Sea.

20 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Evaluation of model set-up

The review of the model implementation (or model set-up) aims to verify whether key

elements like the grid resolution, applied bathymetry, boundary conditions, etc. meet with the

formulation of the general approach. Furthermore, the substantiation of the physical and

numerical input parameters will be verified on the basis of the reports and inquiries.

4.1

DKBS2 model

The set-up of the three-dimensional DKBS2 model is described in general terms in Chapter 2

of [2].

4.1.1

Computational mesh

The DKBS2 model covers the entire Baltic Sea and the Belt Sea, Kattegat and Skagerrak.

Figure 4.1 shows a part of the computational mesh near Øresund. The horizontal resolution

varies from 500 – 1000 m in Belt Sea coastal areas to 4 – 6 km in Baltic offshore areas [2].

Apparently, the mesh of the 3D model is defined in spherical coordinates (latitude, longitude),

but details on the reference system are not reported.

Figure 4.1 Computational mesh of the DKBS2 model near the Øresund (Source: Figure 2.3 in [2]).

The vertical mesh exists of 10 sigma-layers of equal thickness above -10 m of depth and 233

z-layers below that level with a layer thickness of 1 m until -220 m and increasing to 20 m at -

610 m in Skagerrak [2]. At Drogden Sill in the Øresund, with depths up to -10 m, this results

in a layer thickness of 1 m or less.

4.1.2

Bathymetry

The bathymetry is based on an available 500 m x 500 m bathymetric data set created during

the Fehmarn Belt environmental studies (data of 2011 or earlier) [2]. It is questionable if this

21 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

resolution combined with the grid is sufficient to include the details of the Drogden Channel

and the Flinterden Channel given the limited fairway width of about 300 m and 360 m,

respectively.

4.1.3

Simulation period and initial condition

Simulations have been presented for the period 2008-2017 to illustrate the performance of

the model (see further Section 5.1). The origin of the initial condition at 1 January 2008 has

not been discussed in [2], although this is relevant for a system with residence times of about

30 years, since a considerable part of the solution is determined by the initial condition.

Furthermore, the simulation of 2018, which has been used for nesting of the Øresund model

is not discussed.

Atmospheric forcing and boundary conditions

The atmospheric forcing is prescribed at hourly intervals with a resolution of 0.1° x 0.1°

latitude and longitude (approximately 11 by 5.6 km at 60° N) and consists of wind,

atmospheric pressure, precipitation, air temperature, cloud cover and relative humidity

originating from a meteorological model (Par. 2.3 in [2]). Ice concentration fields are

prescribed at hourly intervals with a resolution of 0.2° x 0.2°.

At the open boundary in the Skagerrak water levels, current velocities, salinity and

temperature are used to specify the boundary conditions [2]. This data is provided by the

UKNS2 model of the North Sea. Apparently, the tidal heights of the North Sea model needed

to be improved, for which data of a global ocean tide model was used.

While the co-oscillating tide is prescribed at the open boundary in the Skagerrak, it is not

mentioned whether the tide generating forces are applied to the model. Though water level

variations in the Baltic Sea depend mostly on the atmospheric forcing, tides are also

generated within Baltic Sea. Though the tides are generally weak, the tidal range may still

reach 0.2 m in the Gulf of Finland.

Furthermore, the fresh water inflow of all major rivers is prescribed in the form of time-series

with a daily or monthly interval. Several corrections for incomplete data were necessary to

obtain a realistic estimate of the water balance [2]. Such estimates may have a large

uncertainty, which influences the exchange flows. This was not addressed or investigated by

means a sensitivity analysis.

The forcing and boundary conditions have not been evaluated further. The type of forcing and

boundary conditions seem in general appropriate for this type of large-scale model, however.

The verification in Section 5.1 will show the actual performance.

4.1.5

Physical input parameters

The description of the physical input parameters in Par. 2.6 of [2] is far from complete, and

the parameters that are mentioned are not always clear (e.g. “default parameters”, missing

wind velocities for the corresponding drag coefficients).

The review is therefore limited to comparing available parameters to the choices made for the

Øresund model, see Section 4.2.5.

4.1.6

Numerical input parameters

In the report [2] no information is given on choices for numerical input parameters that

determine the accuracy of the solution. We can only assume that the appropriate choices

have been made that would result in a sufficient accurate and stable solution.

4.1.4

22 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

4.1.7

Calibration

The model covers a large area and has a reasonable resolution and suitable forcing to

simulate the thermohaline circulation in the Baltic Sea. However, from the DKBS2 report is

not clear what has been calibrated and how, and which parameters have not been calibrated

but where specified on the basis objective information. Results of sensitivity simulations, e.g.

with respect to river run-off, are not reported in [2].

Therefore, a conclusion on the suitability of the DKBS2 model for the application in the

present project, will depend on the verification, which is discussed in Section 5.1.

4.2

Øresund model

The set-up of the three-dimensional Øresund model is partly described in Par. 4.2.3 of [3].

Additional information on the set-up and calibration of the Øresund model is given in [9].

4.2.1

Computational mesh

The Øresund model covers the entire Øresund between the Gilleleje-Kullen section in the

north and the Stevns-Skanör section in the south, where the open boundaries are located,

see Figure 2.1. In Figure 4.2 the computational mesh is shown, existing of a unstructured

mesh in the form of triangles and quadrangles [3]. Apparently, a local map projection has

been used, with northing and easting as coordinates (unit: m). Details on the reference

system are not documented in the report. Along the northern open boundary the mesh size is

about 500 m and along the southern open boundary it is about 1 km, possibly in agreement

with the mesh of the DKBS2 model in this area. In the central part of the Øresund model the

grid is refined, and the highest resolution occurs near Lynetteholm. Also, the channels

through Drogden Sill, (Drogden and Flinterden) seem to be resolved by the refined mesh.

The vertical mesh consists of combined z-sigma-layers. At the surface 10 sigma-layers of

equal thickness are present above -15 m of depth and below that level z-layers with a

gradually increasing layer thickness of 1.5 m to 3 m [3]. At Drogden Sill in the Øresund, with

depths up to -10 m, the 10 sigma-layers result in a layer thickness of 1 m or less. In the

channels near Lynetteholm, the Kongedybet and the Hollænderdybet with depths generally

less than 15 m the layer thickness will be 1.5 m or less.

Above -10 m of depth, e.g. at Drogden Sill, the vertical resolution of the Øresund model and

the DKBS2 model are identical (10 sigma-layers both). In areas with depths below -10 m the

vertical resolution in the Øresund model is generally lower, e.g. at -15 m of depth the layer

thickness is 1.5 m, compared to 1 m in the DKBS2 model. It is somewhat surprising that the

more detailed model in the horizontal is less detailed in the vertical. According to DHI the

vertical resolution in the Øresund model has been slightly reduced compared to the DKBS2

model in view of computational time.

4.2.2

Bathymetry

The bathymetry of the Øresund model is also shown in Figure 4.2. It has been derived from

bathymetric survey data from By & Havn for the waters of Copenhagen. Elsewhere, data from

nautical charts have been used [3]. Note that the higher resolution of the Øresund model, e.g.

on Drogden Sill, and the adjustments of the deep channel bathymetry north of Ven (see

Section 4.2.7), will affect the dynamics relative to the DKBS2 model.

According to DHI the geometry in the operational phase of Lynetteholm has not been

adjusted for morphological changes compared to the current situation since the area of

potential (local) erosion on the Middelgrunden exists of sand lenses combined with hard

material for which it is not possible to predict the erosion. Some local capital dredging has

also been neglected.

23 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Figure 4.2 Computational mesh and bathymetry of the Øresund model (Source: Figure 4-5 in [3]).

4.2.3

Simulation period and initial condition

The simulations for the reference situation and for Main Proposal 1 and Main Proposal 2 of

Lynetteholm have been carried out for the year 2018 and started from initial conditions

interpolated from results of the DKBS2 model. A 7-day spin-up period (from 25 December

2017 on) was applied to let the interpolated DKBS2 conditions adjust to the Øresund model,

and to the differences in layout at Lynetteholm [9]. This period should be sufficiently long for

barotropic processes, but whether it is also sufficient for baroclinic processes and the effect

of the land reclamation is hard to judge. No statement was given that this period proved to be

sufficiently long, e.g. supported by sensitivity simulations.

Atmospheric forcing and boundary conditions

The atmospheric forcing is identical to the forcing of the DKBS2 model [9]. The resolution of

11 by 5.6 km (see Section 4.1.4) means that the model is roughly covered by 10 points in

both north-south and east-west direction. Effects of the land-water transition on the

4.2.4

24 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

atmospheric forcing, which may become relevant in smaller scale modelling, have not been

mentioned.

We assume that the fresh water inflow to the Øresund model is also consistent with the

DKBS2 model, although this is not stated in [3] or [9].

The boundary conditions have been interpolated from hourly data of the DKBS2 model,

accounting for the difference in vertical resolution [9]. For water levels this is very coarse; 10

minute data would be more appropriate. No output has been provided to show that

irregularities along the boundaries due to time and space interpolation are absent.

4.2.5

Physical input parameters

A description of the physical input parameters of the Øresund model is given in [9]. It is

slightly more extensive than for the DKBS2 model. Notable differences are:

The bed roughness length is 0.1 m in areas of less than 6 m deep in view of

generally occurring vegetation, and 0.03 m in the remaining areas, based on

calibration [9]. Note that in the DKBS2 model a much lower value of 0.005 m was

used.

According to DHI the Øresund model also includes extra friction to represent the

Øresund bridge piers.

The diffusivity factor for temperature and salinity in the horizontal: ‘scaled eddy’ is

equal to 1, and in the vertical: ‘scaled eddy’ equal to 0.03 [9]. This is different from

the DKBS2 model: horizontal 0-1.0 (temperature/salinity) and vertical 1.0

(temperature) / 0-1.0 (salinity) [2].

Without explanation in the reports or further study of the software and the models we are not

able to reflect further on the input parameters.

Numerical input parameters

In the report [3] or in [9] no information is given on choices for numerical input parameters

that determine the accuracy of the solution.

Calibration

In the Øresund model the following parameters were adjusted during the calibration of the

model, see [9]:

The water level at the boundaries to reproduce the water levels in and water level

gradients along the Øresund.

The bottom roughness length and the scaling factors for dispersion on the basis of

the vertical salinity profiles observed south of Ven Island and in Køge Bay.

Furthermore, the deep channel bathymetry north of Ven was adjusted (smoothed) to

reduce impact of artificial bumps blocking for the near-bed inflow of saline water.

It is not clear whether the calibration and the verification (see Section 5.2) have been carried

out with independent data sets (e.g. over different periods).

The Øresund model was nested in the DKBS2 model and during the calibration of the

Øresund model the water level boundary condition has been calibrated. In itself this can be

considered as an improvement. By result, the transports through the Øresund are different

compared to the DKBS2 model, while results of the DKBS2 model were accepted for the

long-term effects in the Baltic Sea (see Section 5.1). Therefore, a discussion of the effect of

the water level calibration on the transports though the Øresund would have been

appropriate.

4.2.6

4.2.7

25 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

4.3

Conclusion

The reporting on the model set-up is such that a full assessment cannot be given in this

review. Elements like model domain, grid resolution, applied bathymetry, atmospheric forcing,

and the boundary conditions, are appropriate for the purpose. However, it is not clear

whether the spin-up period is long enough, and if time interval of the boundary conditions is

sufficient.

The physical input parameters cannot be fully verified on the basis of the input provided.

Notable differences between the models, which describe the same exchange flows through

the Øresund, are

The Øresund model has a higher grid resolution in the horizontal, as may be

expected, but a lower grid resolution in the vertical compared to the large scale

DKBS2 model.

The higher horizontal resolution of the Øresund model, e.g. on Drogden Sill, and the

adjustments of the deep channel bathymetry north of Ven, will affect the dynamics

relative to the DKBS2 model.

The physical parameters may differ considerably, like the resistance due to bed

friction and the parametrization of bridge piers of the Øresund bridge, and the

diffusivity factors for temperature and salinity.

With respect to the numerical input parameters we need to trust that the appropriate choices

have been made, resulting in sufficient accurate and stable solutions. Therefore, the

verification of the models in Chapter 5 is to provide the main basis for confidence in the

models.

Note that due to the calibration of the water level and water level gradients in the Øresund

model, the results of the DKBS2 model basically serve to provide initial and boundary

conditions for currents, salinity and temperature. The suitability of these conditions derived

from the DKBS2 model for 2018 depend solely on the verification for the period 2008 – 2017,

which is discussed in the next chapter.

26 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Evaluation of model verification

The verification of the models has been assessed based on the results as presented and

described in the reports and additional information provided, but without access to the original

definition the numerical models and the output files of computations. The results of the

DKBS2 model verification are discussed in Section 5.1 and the results of the Øresund model

verification for the reference situation (without Lynetteholm) in Section 5.2.

5.1

Results of DKBS2 model verification

In Chapter 3 of the DKBS2 report [2] the model results have been verified with measured

water levels, currents, salinities and temperatures. The flows through the Danish Straits have

been compared to values from literature. The entire computation covers 10 years, from 2008

– 2017. The verification is generally presented for shorter periods, however.

5.1.1

Water levels

The computed water levels are compared to measured water levels for selected periods in 13

tide gauge stations distributed over model area in the DKBS2 report (Par. 3.1.1 in [2]):

A period of one month (15 February to 15 March 2016) for 5 stations with tidal

influence (Skagerrak, Kattegat and Danish Straits), and

A period of 5 months (July -November 2016) for 8 stations in the inner Baltic where

the water level variations are predominantly determined by the wind, although small

tidal variations are also present in the Gulf of Finland (Helsinki, Kronstadt/Saint

Petersburg) in both model and measurements.

Visual inspection of the presented plots (periods of the two sets not overlapping) show that in

these periods the main variability is well represented, especially in the Baltic Sea. A

numerical evaluation and/or a more detailed discussion of specific phenomena like tides,

wind events (surges), water level differences over the Straits or MBI’s, has not been

presented, however.

Currents

The computed and measured currents are compared in 5 stations, of which 2 are located

north of the Danish Straits in the Skagerrak and Kattegat, 2 are located in the Arkona Basin

roughly several 100 kms southeast of the Straits and 1 in the Northern Gotland Basin in the

middle of the Baltic Sea, see in the DKBS2 report (Par. 3.2.2 in [2]):

Väderöarna (Skagerrak) at 4 m and 28 m depth July – September 2014,

Läsö Ost Boj (Kattegat) at 2 m depth February – October 2009,

Fino station (Arkona Basin) at 5 m and 20 m depth August 2013 – January 2014,

BSH Arkona Becken (Arkona Basin) station at depth 5-6 m and 40 m March – July

2012,

Huvudskär Ost Boj (Northern Gotland Basin) at depth 2m January – September

2016.

Time-series of current speed and direction and current roses have been compared for

surface currents (between 2 and 6 m depth) and when available for the currents at a larger

depth (20 – 40 m). The time-series vary in length from 3 – 9 month and are not overlapping.

The interpretation of the quality of fit of current time-series and current roses is generally

difficult as the measured currents near the surface may have been influenced by local

variations in wind or by (subgrid) bottom topography near the bed. A discussion of specific

phenomena occurring in these periods, like tides, storms and particular extreme MBI’s (e.g.

December 2014) is desirable but not available, however. The suggested fair agreement in the

report is therefore not entirely convincing.

5.1.2

27 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

The integrated parameter of flow through the Danish Straits is discussed in Section 5.1.4.

5.1.3

Salinity and temperature

More important for the present investigation is the quality of the salinity and temperature

development and stratification on both sides of the Øresund, since these parameters are

used for nesting of the Øresund model. Salinity and temperature measurements at the

surface, the bottom and sometimes and at an intermediate level have been graphically

compared to time-series of modelled salinity and temperature for the period of 2008 – 2017 in

21 stations [2]. Unfortunately, there is no further discussion of these results in [2], for instance

assessing the response of the model on the more extreme events (MBIs). The

measurements, were roughly obtained with a monthly interval over periods varying between 5

and 9.3 years.

Focussing on the stations in or near the Øresund (AnholtE in the Kattegat, KBH431 in central

Øresund, and BY2 in the Arkona Basin) we notice a strong variability in modelled salinity near

the surface in the first two stations, in contrast to the third station where this is the case near

the bottom. Such variability can only be very roughly compared with monthly measurements.

The range of variation seems correct, however. The salinity near the bottom in the first two

stations (high salinity ~ 33 psu) and near the surface in the third station (~8 psu) are much

more stable and show a fair to good agreement with measurements.

Measured and modelled temperatures generally show a good agreement on seasonal and

interannual time scales, with exception of the temperature minima near the bottom in

particular years in AnholtE and KBH431. It is not clear what might have caused these

deviations.

Salinity and temperature profiles to check the height of the modelled and measured

thermoclines and pycnoclines were not presented in the DKBS2 report. It is recommended to

verify those too. Salinity profiles in the Øresund during 2018 are discussed in Section 5.2.3.

Given the large scale of the model, the presented results show a good match with the

measurements on the long-term (9 years), as well as on seasonal and interannual scales.

5.1.4

Water and salt exchange through the Danish Straits

The flows through the Danish Straits, Little Belt, Great Belt and Øresund have been

graphically presented for 2011 in the DKBS2 report (Par. 3.2.1 in [2]). The mean outflow

computed by the model for the period 2008 – 2017 is 533 km

/year for all straits combined

[2]. This is 11% higher that the value of 480 km

/year mentioned in literature (see Table 2.1

and [5]). We consider this is an acceptable level of agreement for such a comparison, since

the accuracy of mean outflow depends on the water balance of the Baltic Sea which depends

on estimates of the inflow of rivers, precipitation and evaporation, where larger uncertainties

are not uncommon. In the model the flows are roughly distributed between Little Belt, Great

Belt and Øresund according to the ratio of 1:7:3 as shown on the basis of the instantaneous

flow results of 2011 in [2]. This means that generally 27% of the water flows through

Øresund. This is in agreement with the ratio of 25 -30% mentioned [5]. Using the 27% ratio

and the 533 km

/year for all straits combined, the mean flow through the Øresund over 2008

– 2017 is estimated at 144 km

/year (the corresponding model result for 2008-2017 was not

reported

The exchange of salt has not been addressed in [2]. But from [10], referenced in [9], we

understand that the distribution of salt transport over the Straits can be different from the

In [9] DHI reports a 31% higher net flow of 189 km

/year for the period 2002-2019. The origin of this value and the

reason for the difference is not known at present.

28 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

ratios provided for the flows. E.g. for small Major Baltic Inflow (MBI) events the part of the salt

transport through the Øresund can be much larger than 25-30% of the total transport.

However, in the comparison of the time-series of measured and modelled salinity the long-

term fit over 10 years is stable. Although this is still less than the residence time of about 30

years, it suggests that the total salt exchange through the Danish Straits is modelled

reasonably well.

The accuracy of water and salt exchange though the Øresund in the DKBS2 model has not

been addressed in [2], but since the water levels are further calibrated and salinity is further

verified in the Øresund model this is less relevant for the present study.

5.1.5

Conclusion

Considering the large space and time scales of the simulations with the DKBS2 model, the

general quality of the results over the period 2008 – 2017 is reasonably good. By definition,

the model lacks finesse in and near Øresund, and the accuracy of the parameters used for

nesting of the Øresund model has not been addressed. Comparison to values from literature

suggest an accuracy in the order of 10% for the total net flow through all Danish Straits. For

the salt exchange such information was not available. However, the stable long-term results

for the salt dynamics in the Baltic Sea suggests that the total salt exchange through the

Danish Straits is modelled reasonably well.

The DKBS2 model has been used to provide boundary conditions and initial conditions for

the Øresund model in the year 2018. However, the application of the DKBS2 model to 2018

has not been presented or discussed.

The limited accuracy information and the lack of data on the general performance in 2018

does not have to be a major problem, since certain model aspects have been subject to

further calibration (water level boundary condition) or verification (salinity) in the Øresund

model, see Section 5.2.

5.2

Results of Øresund model verification

In Annex A of the Øresund report the results of the calibrated model have been verified with

measured water levels and salinities for the year 2018 (see [4]). The results for currents and

temperatures were not verified. The results of (integrated) flow and salt transport for the

reference situation (without Lynetteholm) are presented in Par 6.1.6 of the Øresund report.

5.2.1

Water levels

In Annex A.1 of [4] the time-series of measured and modelled water levels in the 11 water

level stations used for calibration are presented for the entire year 2018. 6 stations are

located on the Danish shore and 5 on the Swedish shore of the Øresund. Additionally,

scatterplots and frequency of occurrence plots of measured and modelled water levels are

presented. These plots also include a numerical evaluation, but this is not discussed in [4].

From these results we conclude that due to the calibration of the water level boundaries of

the Øresund model the fit of the time-series is good, with an average root-mean-square error

(RMSE) of 0.06 m.

In view of the water level gradient driven part of the flow through Øresund, a comparison of

the difference in water levels between stations near the northern boundary and the southern

boundary for the model and the measurements would be of interest. Probably this error is

only slightly larger.

29 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

5.2.2

Currents

Comparison of modelled and measured currents in transects or fixed points on Drogden Sill

would provide a strong verification of the Øresund model. This is especially relevant since the

transport through the Øresund determined by the DKBS2 model was changed by the

calibration of the water levels. However, no verification of currents is available in the Øresund

report, nor is it explained that current measurements are not available in the Øresund in

2018, or alternatively in other periods. The verification of the flow therefore depends on the

assessment of the accumulated flow in general terms, see Section 5.2.4.

Salinity and temperature

In Annex A.2 of [4] modelled and measured salinity profiles have been compared for 22

moments distributed over the year in the stations P2 and P4 located about 30 km north and

25 km southwest of Drogden Sill in respectively 35 and 14 m deep water. These results are

not discussed in [4] and therefore a short discussion is included here. Inspection of the

profiles shows that the salinity near the bed in station P2 is on average roughly 4 psu too low

compared to the measured profiles. The measurements also show a (much) stronger

stratification than the model. In station P2 the surface salinity increases during the MBI event

in September 2018 in both model and measurements. This resulted in station P4 also in a

significant increase (between 10 - 15 psu) in salinity over the entire water depth in September

in both model and measurements (salinity values exceeding 17 psu). In the shallower station

P4 the difference between the salinity at the surface and the bed is generally small or even

negligible.

During the inflow event of December 2018, no increase of salinity is noted in the stations P2

and P4. This is perhaps (partly) due to the large interval between the measured profiles (26

November – 20 December 2018).

5.2.3

Figure 5.1 Vertical section of salinity during various stages of exchange flow along Drogden Channel and

Hollænderdybet computed for the reference situation: (top) after a more or less stable period on 8 August

2018 (middle) during a period of saltwater intrusion on 15 September 2018 and (bottom) after a longer period

of low salinity outflow on 20 November 2018 (Source: Figure 6-152 to Figure 6-154 in [4]).

30 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

The salinity distributions in various stages of exchange flow have been illustrated in [4] by the

computed salinity distributions in a vertical section along Drogden Channel and

Hollænderdybet, see Figure 5.1. From top to bottom the plots show the longitudinal section

after a more or less stable period on 8 August 2018, during a period of saltwater intrusion on

15 September 2018 during the MBI event, and after a longer period of low salinity outflow.

The stages can be verified from the plot of the accumulated salt transport in Figure 5.2. In all

stages a vertical density gradient is present at Drogden Sill. Note that in areas with a water

depth of less than 15 m in the vertical section ‘sigma-effects’ can be found at places where

the isohalines follow the variation of the sea bed.

It is recommended to add time-series of modelled surface and near bed salinity and their

measured values, to verify the high variability of surface salinities noted in DKBS2 model. In

addition, it is recommended to add time-series and profiles of modelled and measured water

temperature to the verification.

5.2.4

Water and salt exchange through Øresund

The water and salt transport computed by the model for 2018 cannot be compared to

measurements. We can compare the net transport and the exchange flow of water and salt to

the characteristic values given in Table 2.1, however. For the MBI events present in the

simulation, we have used estimates from literature for comparison.

In Par 6.1.6.2 of [3] the net and the accumulated flow through the Øresund is discussed

amongst others for the reference or baseline situation. From Table 6-5 in [3] it follows that the

net flow through the Øresund in the model is 175 km

/year over 2018. This is 35% higher

than the average flow through the Øresund of about 130 km

/year derived from literature by

taking 27% of 480 km

/year for all Danish Straits, see also Section 5.1.4.

An estimate of the mean water exchange per year through the Øresund can be obtained by

taking half of the mean absolute flow of 29 652 m

/s (see Table 6.1), or 467.6 km

/year and

correct this with half of the net flow, or 87.5 km

/year. This leads to a mean water inflow of

about 319 km

/year and a mean water outflow of 555 km

/year through the Øresund in 2018.

This is respectively 19% and 24 % higher than the estimated mean values of 319 and 448

/year that one would get for the Øresund by taking 27% of the global values given in

Table 2.1 for all Danish Straits.

The largest inflow events are estimated from the plot of the accumulated flow for the Øresund

cross-section at Santholm, see Figure 6-128 in [3], at ~50 km

in September and ~40 km

December 2018. These were identified as small/medium size MBI events in [9]. Mohrholz

estimates the water transport through the Øresund at 35 km

during the September event and

42 km

during the December event at a different ratio of 22% (compared to 27% above) of

the total water exchange during the events, see [10] and the

link

given in [9] . Despite the

differences found, this confirms that the transport of water during the MBI events is in the

right order of magnitude taking into account the different nature of the model results and the

estimate by Mohrholz.

In Par. 6.1.6.5 and Par. 6.1.6.6 of [3] the net and the accumulated salt transport though the

Øresund is discussed for the reference or baseline situation, amongst others. The

instantaneous salt transport has a range of approximate -2 000 000 kg/s to +1 000 000 kg/s

(positive is north going), see Figure 6-143 in [3], while the long-term net transport would

almost be zero, see Table 2.1. The range of the accumulated salt transport through the

Øresund is between +900 and -400 M ton salt, see Figure 5.2. At the end of the year 2018

the annual net salt transport for the Øresund is not necessarily zero due to inter annual

variations. In this case it is small and slightly negative with -100 M ton salt. This is roughly in

31 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Figure 5.2 Accumulated transport of salt in the cross-section through the Øresund with the existing conditions

(black curve), reclamation without landscape (green curve) and reclamation with landscape (blue curve).

Source: Figure 6-148 in [4][3].

agreement with the balance in the salt transport shown in Table 2.1 as it is only 0.3% of the

total salt exchange of 30 G ton/year for all Danish Straits.

A fair estimate of the mean salt exchange per year through the Øresund can be obtained by

taking half of the mean absolute salt transport of 350 000 kg/s given in Table 6.2. This leads

to a salt exchange of about 11 G ton/year through the Øresund in 2018 or 37% of the 30 G

ton/year through all Danish Straits mentioned in literature, see Table 2.1.

Regarding the two small/medium size MBI events identified in [9], the inflow of salt can be

estimated from the accumulated salt transport through the Øresund in Figure 5.2. In the event

of September 2018, the accumulated salt transport changes from about +560 to -370 M ton,

so during the event about 930 M ton or 0.93 G ton of salt flowed into the Baltic Sea. In a

similar way an inflow of about 0.74 G ton salt is estimated for the event of the beginning of

December 2018. In the reference and link given in [9], Mohrholz estimates the salt transport

for the Øresund at 0.65 G ton during the September event and 0.75 G ton during the

December event at ratios of 57% and 52% of the total salt exchange during the events.

Despite the differences, this also confirms that the transport of salt during the MBI events is in

the right order of magnitude.

De total salt transports estimated by Mohrholz for the September and December events is

1.16 and 1.45 G ton. These are somewhat lower than the 1.6 G ton of salt transport that

corresponds to a MBI occurring on average once per year, based on the exponential curve in

Figure 5.3.

Figure 5.3 Total frequency of inflow classes in the time series DS0 for the three subsequent 40-year periods

1896-1935, 1936-1975 and 1976-2015 (source:

https://www.io-warnemuende.de/major-baltic-inflow-statistics-

7274.html)

32 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

In summary:

The net flow through the Øresund in the model is 175 km

/year over 2018, 35%

higher than estimated on the basis of global numbers in Table 2.1.

The net salt exchange is very small, as expected.

The water exchange in the model is estimated from the results at about 319 km

/year

inflow and about 555 km

/year outflow through the Øresund in 2018, respectively

19% and 24 % higher than estimated on the basis of average numbers in Table 2.1.

The salt exchange in the model is estimated from the results at about 11 G ton/year

through the Øresund in 2018 (inflow and outflow), or 37% of the average salt

exchange through all Danish Straits mentioned in Table 2.1.

During two small/medium size MBI events the inflow thought the Øresund in the

model is about 35 km

in September and about 42 km

in December. The

corresponding inflow of salt into the Baltic Sea is estimated at 0.93 G ton and 0.74 G

ton. Both the inflow of water and salt during these MBI events seem in the right order

of magnitude for MBI events occurring more than once per year on average.

Although the modelled flow and salt transport though the Øresund could not be verified with

measurements of 2018, we conclude from a comparison with published characteristic values

that the modelled flow and salt transport in 2018 is possibly somewhat stronger than average

but realistic in magnitude. The year also contains two small/medium size MBI events and

therefore the simulation of 2018 might be sufficiently representative for the investigation of

the effect of Lynetteholm on the water and salt exchange through the Øresund, provided that

the blocking effect is not sensitive for the occurrence of these events (see Section 6.2).

5.2.5

Conclusion

The performance of the Øresund model was verified for time-series of water-levels and

profiles of salinity measured in 2018. Unfortunately, no verification of currents and water

temperature has been provided. Since the adjustment of the water levels prescribed at the

open boundaries (originally derived from the DKBS2 model by nesting) may have changed

the transport through the Øresund , it is necessary to verify the transport of water and salt at

least in some way. Therefore, we compared the transport of water and salt and the two

small/medium size MBI events identified in [9] to values published in literature. From this we

conclude that the flow and salt exchange computed by the model for 2018 is possibly

somewhat stronger than average but realistic in magnitude. Therefore, we consider the

simulation of 2018 sufficiently representative for the present purpose, to estimate the effect of

Lynetteholm on the water and salt exchange through the Øresund by comparing model

results of a simulation including Lynetteholm with the reference situation without Lynetteholm

provided that the blocking effect is not sensitive for the occurrence of MBI events (see

Section 6.2). Concrete data on the model accuracy has not been provided, except for the

water levels. The accuracy of the water and salt exchange will be discussed the next section

in terms of the accuracy band of the blocking effect.

33 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Evaluation of the impact of Lynetteholm on water

and salt exchange

In the Hydrodynamic Study, the impact of Lynetteholm on the water and salt exchange

through the Øresund has been determined by applying the Øresund model to Main Proposal

1 and 2 of Lynetteholm and comparing the results with those obtained for the reference

situation. In the Øresund report the comparison presented for the cross-sections Øresund (an

east – west section across Santholm) and Peberholm East and Peberholm West are the most

interesting ones. Several parameters on the water and salt exchange are presented, notably:

the instantaneous flow, the mean absolute flow (Par. 6.1.6.1),

the accumulated flow and the annual net flow (Par. 6.1.6.2),

the blocking effect for flow (Par. 6.1.6.3).

the momentary salt transport (Par. 6.1.6.5),

the accumulated salt transport and the annual net salt transport (Par. 6.1.6.6),

the blocking effect for salt transport (Par. 6.1.6.7),

see [3] and [4].

The key parameter in this comparison is the blocking effect for water flow and salt transport,

as described in Section 3.3. The conclusions presented in the Øresund report are based on

the cross-section Øresund. In principle it is better to evaluate the blocking effect at the

smallest and shallowest cross-section, Drogden Sill, however. Therefore, we will make also

use of the additional blocking analyses in [8], where the blocking effect has been evaluated

by combining the sections of Peberholm East and Peberholm West, located close to Drogden

Sill.

6.1

Blocking effect for water exchange

In Table 6.1 the mean absolute flow over 2018 of the reference situation (baseline) and two

Lynetteholm variants (Main Proposal 1 and Main Proposal 2) is shown for the most relevant

cross-sections, together with the percentage of change, based on [8]. Compared to [8] we

have adjusted the sign of the percentage of change, to let it correspond with the definition of

the blocking effect for the entire year. The values for the cross-section Total Drogden Sill (red

text and numbers) were simply obtained by combining the mean absolute flow of Peberholm

East and Peberholm West and deriving the blocking effect/change percentage from that. This

introduces minor errors in the estimate, however. Therefore, in [8] a new estimate for Total

Drogden Sill (black numbers) was provided by analyzing the combined discharges

Peberholm East and Peberholm West (first determining the total discharge, and then the

mean absolute values). The resulting blocking effect for 2018 for the flow at Drogden Sill

(East & West Peberholm combined) is -0.186% and -0.244% for Main Proposal 1 and 2,

respectively. This is similar to the blocking effect originally computed for the Øresund section.

The blocking for Main Proposal 2 (-0.244%) is somewhat stronger than for Main Proposal 1 (-

0.186%) due to the coastal landscape in Main Proposal 2.

The accuracy of the calculation of the blocking effect is discussed in [9]. It is argued in [9] that

it is reasonable to accept the 95% confidence range of ±0.25% of the blocking effect

established for the Øresund Link modelling as an estimate of the accuracy for the present

modelling of the blocking effect in the Øresund. Although this is not explicitly stated, we

assume the confidence range holds both for flow and for salt. We can imagine that the

confidence range is within the given range of ±0.25% due to improvements in the modelling

approach such as higher resolution models and longer simulations compared to that applied

34 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

for Øresund Link. However, we do not agree to reduce the confidence range with 50%

because the initial blocking of Lynetteholm is about 50% of the initial blocking effect of the

Øresund Link, as suggested in [9].

We conclude therefore that for the flow through the Øresund the blocking effect is -0.186%

and -0.244% for Main Proposal 1 and 2, respectively, with 95% confidence range within

±0.25%.

Furthermore, it is relevant to check if the blocking effect is sensitive for the occurrence of the

small/medium MBIs described in Section 5.2.4. Inspecting the time-series of the blocking

effect for flow derived for the Øresund cross-section (Figure 6-133 in [3]) we do not see

notable effects in the periods of the MBIs (September, beginning of December).

Table 6.1 Mean absolute flow over 2018 (Annual mean gross water flow - calculated without sign) from [8].

Change percentage is equal to the computed blocking effect over 2018.

Baseline

Main Proposal 1 Main Proposal 2 Main Proposal 1 Main Proposal 2

m3/s

Change %

West Peberholm

10897.87492

10800.71993

10767.86938

-0.892%

-1.193%

East Peberholm

18789.44222

18829.02629

18844.63364

0.211%

0.294%

Total Drogden Sill

29687.31715

29629.74622

29612.50302

-0.194%

-0.252%

Total Drogden Sill

29651.52822

29596.23035

29579.10229

-0.186%

-0.244%

Oresund (across Saltholm)

29568

29513

29496

-0.186%

-0.244%

Cross-section

6.2

Blocking effect for salt transport

In Table 6.2 the mean absolute salt transport over 2018 of the reference situation (baseline)

and two Lynetteholm variants (Main Proposal 1 and Main Proposal 2) is presented for the

most relevant cross-sections, together with the percentage of change, in a similar way as was

done for the flow in Table 6.1. In this case the new estimate for the blocking effect for 2018

for salt transport at Drogden Sill (East & West Peberholm combined) in [8] provides with -

0.191% and -0.241% for Main Proposal 1 and 2, numbers that show a slightly stronger effect

than computed for the Øresund section across Santholm. Note that the new values are very

close to the values obtained for the blocking effect for the flow.

We conclude that for the salt transport through the Øresund the blocking effect is -0.191%

and -0.241% for Main Proposal 1 and 2, respectively, with a similar 95% confidence range

within ±0.25%, as for the flow.

Furthermore, we checked the sensitivity of the blocking effect on the occurrence of the

small/medium MBIs described in Section 5.2.4. Inspecting the time-series of the blocking

effect for salt derived for the Øresund cross-section (Figure 6-157 in [3]) the effects in the

periods of the MBIs (September, beginning of December) are very small for Main Proposal 1

and slightly larger for main Proposal 2, but not more than -0.01%. Therefore 2018 is

considered sufficient representative for the analysis of the impact of Lynetteholm on the

exchange of water and salt.

35 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Table 6.2 Mean absolute salt transport over 2018 (Annual mean salt transport - calculated without sign) from

[8]. Change percentage is equal to the computed blocking effect over 2018.

Baseline

Main Proposal 1 Main Proposal 2 Main Proposal 1 Main Proposal 2

kg/s

Change %

West Peberholm

136252.4264

134829.9941

134438.3022

-1.044%

-1.331%

East Peberholm

214002.6852

214718.4694

214933.8328

0.334%

0.435%

Total Drogden Sill

350255.1116

349548.4635

349372.135

-0.202%

-0.252%

Total Drogden Sill

349824.0955

349156.339

348981.0446

-0.191%

-0.241%

Oresund (across Saltholm)

362226.0053

361584.2347

361414.9652

-0.177%

-0.224%

Cross-section

6.3

Results in perspective

In summary, in the Hydrodynamic Studies, the blocking effect for water and salt at Drogden

Sill is estimated at -0.19% and -0.24% for Main Proposal 1 and 2 respectively, with a 95%

confidence range within ±0.25%. This indicates that a small reduction of the exchange of

water and salt through the Øresund is likely. Whether these ranges are acceptable or not, is

outside the scope of this review.

In an attempt to put this result into perspective, DHI compared it to the blocking criteria for the

Øresund Link (Par. 6.1.6.3) and to a rough estimate of the effect of the expected sea level

rise (Par. 6.1.6.4 and Par. 6.1.6.8 in [3] and [4]).

In [3] the blockage effect estimated for Lynetteholm was compared to the blocking criteria

that were applied for the Øresund Link. Citation: “In

connection with the Øresund Link, a zero

solution was required, where by compensation excavations an attempt was made to produce

conditions resulting in a zero-blocking factor for both water and salt. In these calculations, the

blocking requirement was set at less than 0.1% with an uncertainty spread estimated at +/-

0.25% within an uncertainty limited to a 95% confidence interval, Ref. /1/

. The uncertainty

accepted by the zero solution is thus higher than the estimated blockage than the

Lynetteholm reclamation for salt transport through Øresund.”

For a correct interpretation we

would like to add that this needs to be expressed in terms of an X% chance that the blocking

effect of Lynetteholm Main Proposal 2 is better (less negative) than the Zero Solution (-0.1

0.25%), and a Y% chance that the blocking effect is worse (more negative)

. Where X and Y

need to be determined from the two probability distributions.

In [3] the blockage effect estimated for Lynetteholm was also compared to the estimated

effects of Sea Level Rise (SLR) on the blockage in an attempt to relate the blocking effect for

water and salt transport due to Lynetteholm to other (autonomous) physical processes. The

approach was to increase the water levels in the model by 2 cm and determine the effect on

blocking after half a year (January – June 2018). This slightly increased the thickness of the

top (sigma-) layers in the model. Other parameters from the initial conditions and boundary

conditions were not changed. Other effects of climate change than SLR (river runoff,

precipitation, evaporation, wind) were ignored without discussion. The effect of 2 cm SLR on

the blocking effect of flow in the Øresund cross-section of both Main Proposals was +0.030

%, see Figure 6-148 in [3]. Assuming a rate of SLR of 1.55 mm/year this suggests that a

period of about 8 years is required to balance the effect of Main Proposal 1 and 10 years for

Main Proposal 2.

See reference [6]. This reference has not been made available.

In [9] a requirement for the blockage effect of 0% ±0.25% is mentioned for the Zero Solution of the Øresund Link

instead of -0.1% ±0.25%. When this is correct the probability statement should be adjusted accordingly.

36 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

For salt transport a similar approach is followed in [3]. The effect of 2 cm SLR on the blocking

effect of salt in the Øresund cross-section of both Main Proposals was +0.024 %, see Figure

6-162 in [3]. Assuming a rate of SLR of 1.55 mm/year this suggests that a period of about 10

years is required to balance the effect of Main Proposal 1 and 12 years

for Main Proposal 2.

Although it was stressed by DHI that the effect of the SLR scenario (with all limitations) was

intended for comparison and not as a mitigating measure, the result can also be interpreted

in the sense that Lynetteholm absorbs about 10 years of (autonomous) improvement of

exchange of water and salt due to climate change, assuming the estimate is correct. Note

that this statement holds only for the transport through the Øresund (about 27% of the total

flow, but the percentage can be much larger for salt transport) and not for the transports

through the Little and Great Belt (about 73% for flow).

The Øresund report mentions 25 years, which we believe is not correct.

37 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

Synthesis of findings

The review will be closed with a discussion of the most important restrictions and

uncertainties in the reviewed studies. Next, the final conclusions of the review of the

Hydrodynamic Studies with reference to the exchange flows through the Øresund will be

given, together with some recommendations.

7.1

Discussion of impact of Lynetteholm

In the discussion of the impact of Lynetteholm on the exchange of water and salt we address

three important restrictions and uncertainties in the Hydrodynamic Studies.

The first restriction is that the approach based on the Øresund model, with identical boundary

conditions for the reference situation and the Main Proposals for Lynetteholm, is in principle

only valid when the blocking effect due to Lynetteholm is negligible. The degree to which a

non-negligible blocking effect estimated on the basis of the Øresund model with fixed

boundary conditions is conservative or not would be useful information for the interpretation

of the results. The reports of the Hydrodynamic Studies provided no insight in this; however,

it may have been addressed in previous studies or in literature. Otherwise it can be

investigated in a sensitivity study with the DKBS2 model.

The second subject is the improvement of the model set-up, calibration and the verification of

the Øresund model and the DKBS2 model.

Partly the improvement could be directed at the reporting of model set-up, calibration and

verification. Note that an assessment of the original input and output files of the applied

numerical models cannot replace such documentation as the report should contain the

substantiation of the choices that were made. A better reporting should help to provide more

confidence in the quality of the models.

Another matter is the verification of the transports in the Øresund model. Assuming that no

current or transport measurements were available for 2018, perhaps the application of the

same methodology in another year in which such measurements are available could provide

confidence in the model performance, e.g. by showing that calibration on water levels and

water level gradients and certain aspects of salinity is sufficient to obtain correct water and

salt transports.

Thirdly, the results of the Hydrodynamic Studies regarding the impact of Lynetteholm on the

exchange of water and salt might also benefit of an improved (and possibly reduced)

estimate of the 95% confidence range. This can perhaps be achieved in a qualitative

approach, by regarding the original Monte Carlo approach for the Øresund Link modelling

and estimate the effect of e.g. a more detailed model and longer simulation periods. Or

alternatively, by repeating the Monte Carlo approach.

When required, several options exist to provide more confidence in the results. However, we

cannot estimate whether the efforts involved will be proportional to the improvement of the

results. That also depends on the magnitude of the blocking effect that will be considered

acceptable and/or negligible by the Swedish and Danish authorities.

7.2

End conclusion

In this review of the Hydrodynamic Studies on the impact of Lynetteholm on the exchange of

water and salt through the Øresund, the report of the Hydrodynamic Studies carried out by

DHI and additional material provided in response to questions to By & Havn and DHI have

38 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

been evaluated considering the methodology, the set-up and verification of the modelling, the

interpretation of the model results and the conclusions with respect to the exchange of water

and salt through Øresund.

The

key results of the Hydrodynamic Studies

on the impact of Lynetteholm on the exchange

of water and salt through the Øresund are the estimate of the blocking effect for flow and salt

at Drogden Sill of -0.19% and -0.24% for Main Proposal 1 and 2, respectively, with a 95%

confidence range of ±0.12% for Lynetteholm. These results are based on a relative approach

in which the Main Proposals 1 and 2 of Lynetteholm are compared to the reference situation

using a 3D numerical model of the Øresund and the conditions of 2018. The 95% confidence

range of ±0.12% for Lynetteholm is half of the ±0.25% confidence range established in the

Øresund Link modelling.

Note that this result differs slightly from the reduction reported in the EIA of 0.23-0.25% for

the total flow through the Øresund and 0.21-0.23% for the salt transport for Main Proposal 1

and 2, respectively, because these values were based on a different cross-section.

The

main findings of the review

are the following:

The blocking effect defined in the Hydrodynamic Studies is an appropriate measure

for changes in the exchange of water and salt through the Øresund. Note that the

blocking effect becomes negative when the exchange is weakening, and positive

when the exchange increases.

The approach based on the Øresund model, with identical boundary conditions for

the reference situation and the Main Proposals for Lynetteholm, is valid as long as

Lynetteholm does not affect the hydrodynamic conditions at the locations of the open

boundaries.

The set-up and calibration of the models is not well documented, but the resolution,

the modelled processes and the type of forcing are appropriate for the purpose.

The verification of the models is not complete, particularly a verification of currents or

transports trough the Øresund is missing in the Øresund model. The additional

comparison during the review of the computed exchange of water and salt trough the

Øresund to global numbers from literature supports that the model and the selected

conditions based on 2018 are appropriate for the evaluation of the blocking effect of

Lynetteholm.

Based on the foregoing, the blocking effect of -0.19% and -0.24% for Main Proposal

1 and 2 at Drogden Sill found in the Hydrodynamic Studies is considered a realistic

result. Based on the review we find a 95% confidence range of ±0.25% more

reasonable.

It is outside the scope of the review to evaluate whether the blocking effects found are

acceptable and/or negligible or not.

In case the above estimates of the blocking effect and confidence range are considered not

fully acceptable and/or negligible, further substantiation might be needed. For this the

following suggestions can be made.

It would be useful to learn to which degree the approach based on the Øresund

model with fixed boundary conditions is conservative or not, either from previous

studies or literature, or from a sensitivity study with the DKBS2 model.

It could be useful to improve the reporting of model set-up, calibration and

verification. This could help to provide more confidence in the quality of the models.

Particularly, the verification of the transports in the Øresund model should be

considered. Such a verification for 2018 or for another year could enhance the

confidence in the model performance, e.g. by showing that calibration on water levels

and water level gradients and certain aspects of salinity is sufficient to obtain correct

water and salt transports.

39 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

A 95% confidence range smaller than ±0.25% is perhaps achievable but would

require further substantiation.

In addition to the estimates of the blocking effect for water and salt, the

Hydrodynamic

Studies

presented two ways to put the blocking effect for water and salt due to Lynetteholm

into

perspective:

a comparison with the ‘Zero Solution’ of the Øresund Link, and a

comparison with the effects of sea level rise.

In the first way, the

Hydrodynamic Studies

compared the estimated blocking effect for

Lynetteholm to the blocking requirement applied in the Zero Solution of the Øresund Link. At

the time the requirement was that the blocking was less than 0.1% with a 95% confidence

interval ±0.25%. This is interpreted as the blocking effect such as defined in the present study

is higher than -0.1% ±0.25%. The Hydrodynamic Studies concluded that the uncertainty

accepted by the Zero Solution is thus higher than the estimated blocking effect for

Lynetteholm.

In the

review

the following remarks are made:

We think that such a comparison can be of use, if it can be confirmed that the

requirements of that time are still valid, e.g. by later long-term monitoring (from 2000

– present) of the effects of the Øresund Link.

Given the blocking effect of Main Proposal 1 of -0.19% with a 95% confidence

interval of ±0.25% or the blocking effect of Main Proposal 2 of -0.24% ±0.25% a

probability statement would be more appropriate for comparison with the requirement

of the Zero Solution of better than -0.1 ±0.25% than to state that the uncertainty

accepted by the Zero Solution is higher than the estimated blocking effect for

Lynetteholm.

NB. a blockage requirement of 0% ±0.25% has also been mentioned for the Zero

Solution [9].

In the second way, the

Hydrodynamic Studies

presented a comparison with the estimated

effects of sea level rise on the exchange of water and salt through the Øresund. Simulations

were performed for the existing situation and the two Main Proposals with the Øresund model

in which all water levels were increased by 2 cm. Assuming a rate of sea level rise of 1.55

mm/year, the blocking effect due to Lynetteholm was expected to be equalized after a 10-

year period. For salt the blocking effect was expected to be equalized after a 25-year period.

It was stressed that the sea level rise scenario was intended for comparison of the order of

magnitude and not as a mitigating measure.

In the

review

the following remarks are made:

We have doubts about the approach since sea level rise is only one aspect of climate

change. Other aspects like North Sea conditions, wind climate, precipitation, river

run-off, and even water quality and ecological processes may also change, and lead

to different results.

Using the same data, we estimate that for flow ~8 and 10 years are required to

balance the effect of Main Proposal 1 and 2, resp. and for salt ~10 and 12 years are

required to balance the effect of Main Proposal 1 and 2 resp.

Assuming that these estimates are correct, it could be interpreted as Lynetteholm

absorbing about 10 years of (autonomous) improvement of the exchange of water

and salt through Øresund due to climate change.

However, given the doubts about the approach the present conclusions based on

sea level rise scenarios are not convincing and should not be used without further

substantiation.

40 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

References

[1]

[2]

Rambol (2020).

Lynetteholm Miljøkonsekvensrapport.

Project number 1100038380,

Version 7, 24-11-2020 (in Danish).

DHI (2021).

DKBS2 Hydrodynamic Model Setup and Validation.

DHI Technical Note,

Project number 11091910. (Translation of Draft Final version of 15 August 2018 in

Danish).

DHI (2021).

Construction of Lynetteholm, EIA – Technical Background Report No 1,

Hydraulic Surveys (Selected Chapters).

DHI Report, Project number 11823523-09,

Final version, 6 October 2021 (Translation of parts from report from 2 November

2020 in Danish).

DHI (2020).

Anlæg af Lynetteholm, VVM – Teknisk Baggrundsrapport nr. 1,

Hydrauliske undersøgelser.

DHI rapport 11823523-09, 2 November 2020. Full report

in Danish.

Leppäranta, M. & K. Myrberg (2009).

Physical Oceanography of the Baltic Sea.

Springer-Praxis Geophysical Sciences. ISBN 9783540797029.

DHI rapport i samarbejde med LICengineering a/s for Øresundsbro Konsortiet:

Zero

Solution in Relation to Effects on Cod Recruitment in the Baltic Sea,

December 2002

DHI (2022). Update of Figure 6-126 in [3] and [4] to covering the entire year 2018.

DHI email, 19 January 2022.

DHI (2022). Excel file with an additional blocking analysis for flow and salt for the

cross-section at the Drogden sill by combining data of the East and West Peberholm

sections. DHI 21 January 2022.

DHI (2022).

Clarifications and replies to Deltares review comments.

DHI memo, ref.

11823523-09, 24 January 2022.

Mohrholz, V. (2018).

Major Baltic Inflow Statistics – Revised.

Front. Mar. Sci. 5:384.

doi: 10.3389/fmars.2018.00384.

Mohrholz, V., Naumann, M., Nausch, G., Krüger, S., and Gräwe, U. (2015).

Fresh

oxygen for the Baltic Sea—An exceptional saline inflow after a decade of stagnation.

J. Mar. Syst. 148, 152–166. doi: 10.1016/j.jmarsys.2015.03.005.

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

41 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft

42 of 42

Independent review of the Hydrodynamic Studies on the impact of Lynetteholm on exchange

of water and salt through Øresund

11207757-002-HYE-0001, 18 March 2022, draft