Demonstration of Indicators and GIS Methods

Progress Report of the DGVII-DGXI-Eurostat-EEAWorking Group of the SEA of the TEN

April 1998

Spatial and EcologicalAssessment of the TEN:

Demonstration of Indicatorsand GIS Methods

Environmental Issues Series no. 11

Spatial and Ecological Assessment of the TEN

NOTE

The contents of this report do not necessarily reflect the official opinion of the EuropeanCommission or other European Communities institutions. Neither the European Environ-ment Agency nor any person or company acting on the behalf of the Agency is responsiblefor the use that may be made of the information contained in this report.

This is a progress report of the DGVII-DGXI-Eurostat-EEA working group, prepared by theEuropean Environment Agency. Main contributors to this report are:Chris Steenmans, EEA, Ivone Pereira Martins, EEA, Wim Devos, GIM Geographic Informa-tion Management, Ann Dom, DGXI, Johan Geert Koier, DGVII, Technical assistance map-ping: Klaas Scholte, University of Utrecht

A great del of information on the European Union is available on the Internet. It can beaccessed through the Europa server (http://europa.eu.int)

Cataloguing data can be found at the end of this publication

Luxembourg: Office for Official Publications of the European Communities, 1998

Cover and layout:Folkmann Design

ISBN xxx

© EEA, Copenhagen 1998

European Environment AgencyKongens Nytorv 6DK-1050 Copenhagen KDenmarkTel. (+45) 33 36 71 00Fax (+45) 33 36 71 99E-mail: [email protected] page: http://www.eea.eu.int

1. Introduction ............................................................................. 7

2. Executive Summary ................................................................. 8

3. Context ................................................................................... 10

3.1. Background .................................................................... 10

3.2. The Commission Work Programme

on a SEA of the TEN ...................................................... 10

4. Methodology ......................................................................... 12

4.1. Purpose and Scope of a SEA of the TEN ...................... 12

4.2. Environmental Objectives and Targets .......................... 12

4.3. Impact Indicator Selection ............................................. 12

5. Data and Information Sources .............................................. 16

5.1. Major Data Sources and Limitations .............................. 16

5.1.1. Transport Infrastructure Data............................... 16

5.1.2. Environmental Data ............................................. 17

6. TEN alternatives .................................................................... 18

6.1. Selection of Alternatives ................................................ 18

6.2. The TEN in Perspective ................................................. 21

7. Nature and Biodiversity ........................................................ 22

7.1. Main Objectives Concerning Nature and Biodiversity .. 22

7.2. Vicinity to Sites of Nature Conservation Interest .......... 22

7.2.1. Methodological Approach .................................. 23

7.2.2. The Results .......................................................... 24

7.3. Partitioning of Land ....................................................... 26


7.3.2. The Results .......................................................... 27

Table of Contents

Table of Contents

Spatial and Ecological Assessment of the TEN4

8. Water Resources .................................................................... 29

8.1. Main Objectives Concerning

the Protection of Water Resources ................................ 29

8.2. Crossing of Surface Waters ........................................... 29


8.2.2. The Results .......................................................... 31

9. Noise ...................................................................................... 32

9.1. Main Objectives Concerning Noise ............................... 32

9.2. People in the Vicinity of Infrastructure .......................... 32


9.2.2. The Results .......................................................... 33

9.3. Crossing of Tranquil Zones ............................................ 33

9.3.1. Methodological Approach ............................................ 33

9.3.2. The Results .......................................................... 34

10. Land Resources ...................................................................... 36

10.1. Main Objectives Concerning Land Resources ............... 36

10.2. Land Take ....................................................................... 36

10.2.1. Methodological Approach ................................ 36

10.3. Sensitive Zones Mapping .............................................. 39

10.3.1. Methodological Approach ................................ 39

10.3.2. The Results ........................................................ 43

11. Evaluation and Conclusions .................................................. 46

11.1. Indicator Overview ........................................................ 46

11.2. Conclusions .................................................................... 49

5Table of Figures

Table of Figures

Figure 4.1: Relevant spatial and ecological issues for a

strategic environmental assessment of the TEN ... 13

Figure 5.1: Available digitalised data concerning

European TEN infrastructure ................................. 16

Figure 5.2: Available (digitalised) data for environmental

assessment on European scale .............................. 17

Figure 6.1: Decision No 1692/96/EC Annex III

priority projects ..................................................... 18

Figure 6.2: The three transport modes included in this

study and a multi-modal overview including

the 14 specific projects .......................................... 19

Figure 6.3: Example of the multi-modal TEN network

compared to the national and regional network... 20

Figure 6.4 Estimated length of the trans-European

transport network compared to the overall

European infrastructure (kilometres) ..................... 20

Figure 7.1: Site description according to its legal statute....... 23

Figure 7.2: Designated sites potentially affected by the

development of the trans-European transport

network. Status in mid 1997 (overlaps between

sites not taken into account) ................................. 23

Figure 7.3: Indication of potential conflicts between

complete TEN and nature sites. Nature site

classification according to neighbouring links ...... 25

Figure 7.4: Indication of potential conflicts between

complete TEN and nature sites. TEN

infrastructure classification according to ...................

neighbouring nature sites ...................................... 26

Figure 7.5: Number of land units and their cumulative

surface - existing and TEN planned situation ....... 26

Figure 7.6: Partitioning of Land effects of the complete TEN .... 28

Figure 8.1: Water crossings of the existing trans-European

transport network .................................................. 29

Figure 8.2: Comparison of the number of water

crossings under different TEN variants ................. 30


Figure 8.3: Complete Multimodal TEN classification

according to potential conflicts with surface water .. 30

Figure 9.1: Population susceptible to noise exposure

due to TEN infrastructure ...................................... 32

Figure 9.2: Population exposed to noise impacts

per TEN variant ...................................................... 33

Figure 9.3: Population density map and multi-modal

TEN network .......................................................... 33

Figure 9.4: Intersection with tranquil zones per TEN variant ..... 34

Figure 9.5: Intersection with tranquil zones

according to TEN variants ..................................... 34

Figure 9.6: Noise disturbance zones around the

Complete Multimodal TEN network ..................... 35

Figure 10.1: Estimated surface requirements per

infrastructure type ................................................. 36

Figure 10.2: Land take by TEN variant ...................................... 37

Figure 10.3: Distribution of the major land cover

types in Europe ...................................................... 38

Figure 10.4: Land take by land cover type for TEN variants ..... 39

Figure 10.5: Sensitivity thresholds attributed to

baseline data sets .................................................. 40

Figure 10.6: Assignment rules for determining

sensitivity per group .............................................. 40

Figure 10.7: Sensitivity map based on Group I criteria

(physical-environmental criteria) ............................ 41

Figure 10.8: Sensitivity map based on Group II criteria

(environmental legal status criteria) ...................... 42

Figure 10.9: Sensitivity map based on Group III criteria

(demographic criteria) ........................................... 43

Figure 10.10: Identification of high sensitivity cells

crossed by infrastructure links ............................... 44

Figure 10.11: Sensitivity index per impact group and

TEN variant ............................................................ 44

Figure 10.12: Sensitivity index graphs per impact group

and TEN variant ..................................................... 45

Figure 11.1: Summary of indicator characteristics ..................... 47

Figure 11.2: Assessment of used indicators according to

OECD criteria ........................................................ 48

7

1. Introduction

Introduction

The present report is written in view of the Commission’s aim to develop appropriate meth-ods for the strategic evaluation of the environmental impacts of the trans-European transportnetwork (TEN). In order to study the possibilities of evaluating a particular type of environ-mental impact – namely spatial and ecological impacts – the Directorate-General for Trans-port (DGVII) and the Directorate-General for Environment, Nuclear Safety and Civil Protec-tion (DGXI) in March 1997 initiated a co-operative effort with the European EnvironmentAgency (EEA) and Eurostat.

The Commission’s request led to the creation of a working group consisting of DGVII,DGXI, Eurostat and EEA representatives, and the present study was steered and developedby this working group. The EEA European Topic Centre on Land Cover carried out most ofthe technical work, including the GIS development. The EEA European Topic Centre onNature Conservation provided data on nature conservation and gave methodological feed-back.

The objectives of the present methodological study were:

I. to develop and evaluate a number of indicators and GIS assessment methods and todemonstrate their feasibility - taking account of the present data available - for anassessment of the overall trans-European transport network, and consequently providea common discussion platform between and within Commission services, with externalexperts, policy makers and other stakeholders on this issue;

II. to identify issues for further research as a result of this broad and multi-disciplinaryconsultation.

In this context, the reader should bear in mind that all the results and figures that are presented in thismethodological report serve only to illustrate the potential output of the various assessment methods.They are not the final results of the assessment.


2. Executive Summary

In accordance with the provisions of theEuropean Union (EU) guidelines for thetrans-European transport network (TEN), aEuropean Commission working group hasinvestigated the feasibility of carrying out afull spatial and ecological assessment of theTEN. The aim of the study was to developand test a number of indicators, to considerthe availability of data and to identify issuesfor further research, if necessary. The studyand the resulting report are intended topromote further discussion between allstakeholders on the issue of strategic envi-ronmental assessment (SEA) in relation toTENs.

On the basis of EU environmental objectives,mainly specified in the 5th EnvironmentalAction Plan, the working group identified anumber of indicators that could be takeninto account in a spatial and ecologicalassessment of the TEN. Five different envi-ronmental themes were highlighted:biodiversity and nature conservation, waterresources, coastal zones, noise and landresources.

The working group considered only quanti-tative evaluation methods, using a Geo-graphical Information System (GIS) as ananalysis tool. The methodological approachwas mainly based on land-suitability orsensitivity analysis. This type of analysisfocuses on the environmental characteristicsof the land and relies on expert judgement(based on scientific knowledge). The quanti-tative approach involves indicator categoriesselection, impact estimation and ratings orpriorities assignment. Map overlay tech-niques are the most suitable for carrying outthe analysis; therefore, GIS tools - given thescale of the analysis and the geographicaldifferentiation of the impacts - have provento be the most useful. With these types ofmethodologies, impacts are implied but notexplicitly reported. Qualitative results may beintegrated subsequently but an evaluationwithout any degree of quantification was notconsidered feasible, given the objectives ofthe study.

Six different TEN network developmentvariants - considering only linear road, railand water infrastructures - were used todemonstrate how the indicators could be

used. For the current study, it was decided todefine overall network variants that do notrepresent true policy options, but that areuseful in raising awareness about the possi-ble environmental impacts of the trans-European transport network. The variantsare:

I. a “do-nothing” option: no further exten-sions to the existing network;

II. a “do-minimum” option: the existingnetwork is extended by the execution ofthe fourteen specific projects agreed atthe Essen European Council in 1994;

III. the existing network is further developedthrough the carrying out of all roadprojects;

IV. the existing network is further developedthrough the realisation of all railprojects;

V. the existing network is further developedthrough the implementation of allinland waterway projects;

VI. the existing network is completed by thedevelopment of all planned projects.

Two broad types of indicators were tested.The first type is based on identifying specificsites in the vicinity of planned infrastructurelinks. In principle this technique can be usedto evaluate the potential impact on, forexample, designated biodiversity or naturesites and water catchment areas.

The second category of indicators is basedon the identification of different land covertypes according to their degree of vulnerabil-ity to certain impacts. For example, certainhabitat types may be more vulnerable from abiodiversity point of view than others.Alternatively, one could define non-frag-mented areas that are more likely to have ahigh value for nature and biodiversity thanzones that are already fragmented by existing(TEN) infrastructure. This type of indicatorcan also be used to combine differentenvironmental aspects into an overall vulner-ability/sensitivity index per zone.

It must be noted that the results of anassessment at EU-level must be interpretedwith great care, especially when it concernsspatial and ecological impacts whose signifi-cance very much depends on local circum-stances. The design and alignment of the

Executive Summary 9

infrastructure, the vulnerability of thespecies in the area concerned, the trafficvolume and several other factors determinethe extent of the ecological impacts. Moreo-ver, mitigation measures at project level canreduce (sound barriers) or even eliminate(tunnelling) adverse impacts. Therefore, anEU-level network assessment should prima-rily be seen as a warning system for potentialrisks.

The feasibility study clearly underlined thepractical difficulties that arise in an assess-ment of the overall network. The quality andavailability of data at EU-level turned out tobe rather poor. For example, digitalisedmaps of the TENs are only available on avery large scale (1:1.000.000), informationconcerning the possible alignment of newlinks is often lacking, the actual boundariesof designated areas are unknown or notavailable in GIS-format, and overlaps be-tween different designation categoriescannot be identified.

The insufficiency of information limits theperformance of the assessment methods.Based on the currently available data, themeasurability and analytical validity of theproposed indicators cannot be fully assured.Therefore, either additional data must becollected at EU-level or a different approachmust be followed, namely a more detailedbut still strategic, assessment per infrastruc-ture scheme. An example of this method isthe way in which strategic assessments arecarried out for the GermanBundesverkehrswegeplan. Furthermore,additional data collection would be particu-larly important for improving informationon planned TEN infrastructure. Anotherpriority action is the completion of theNATURA 2000 network, as this wouldprovide a more consistent basis for assess-ment of impacts on habitats.

The models and parameters that were usedin this study have been applied in a ratherstraightforward manner, and have beenemployed using expert judgement, whichinevitably involves a certain degree of subjec-tivity. The key assumptions that have beenmade and their transferability to EU scaleshould be improved by a broad and multi-disciplinary consultation of experts. Inparticular the EU wide definition of sensitivezones, which inevitably relies to a certainextent on a subjective use of sensitivityvalues, should be further researched.

As a broad conclusion, the present methodo-logical study can be considered successful asit realised most of its original objectives. Itdemonstrated the technical feasibility, butalso the limitations, of a European-wide SEAof the TEN, and stressed the usefulness ofmulti-disciplinary data collection at anEuropean scale.

As a result, the present exercise helpedsuggest further research and methodologicalimprovements:

1. Refining the indicators and methodsthrough a wide and multi-disciplinaryconsultation process;

2. Filling the major data gaps on the TENand on the environment;

3. Carrying out a full spatial and ecologicalassessment of the TEN, including theintegration into the GIS of the results ofthe traffic-related impacts and a compre-hensive multi-criteria analysis of thepredicted impacts;

4. Initiating longer term research on theabove outlined issues, in order to gradu-ally improve methods, data and tools forthe ecological and spatial impact assess-ment of transport plans;

5. Performing “zoom analyses” or “samplesurveys”, enabling a testing of the inter-relation between different planning/geographical levels and correspondingassessment techniques.


3. Context

3.1. Background

In July 1996 the European Parliament andthe Council adopted Community guidelinesfor the development of the multi-modaltrans-European transport network (TEN)¹.By 2010, it will gradually be established byintegrating land, sea and air transportinfrastructure networks throughout theCommunity. The transport infrastructure willcomprise road, rail and inland waterwaysnetworks, seaports and inland waterwayports, and other interconnection points. Theguidelines provide the framework for thedevelopment of the TEN and specify objec-tives, principles and outline maps. One ofthe main objectives of the TEN is “...toensure the sustainable mobility of personsand goods within an area without internalfrontiers under the best possible social andsafety conditions...” (article 2 of the guide-lines). The network should not onlystrengthen the economic and social cohe-sion of the Union, but should also help toachieve the Union’s environmental objec-tives. Integration of environmental concernsinto the design and development of the TENis considered a key priority.

The Commission has long recognised thatthe development of a process of strategicenvironmental assessment (SEA) is a prereq-uisite in this regard. This view has beenconfirmed in the guidelines. Article 8 §2 ofthe TEN guidelines states that “the Commis-sion will develop appropriate methods ofanalysis for strategically evaluating theenvironmental impact of the whole network[...] and will develop appropriate methods ofcorridor analysis covering all relevant trans-port modes without prejudice to the defini-tion of the corridors themselves [...]”. Thearticle also stipulates that the result of thework will, as and when appropriate, be takeninto account by the Commission in its reporton the revision of the guidelines, which is tobe published by July 1999.

It should be noted that the concept ofstrategic environmental assessment is notlimited to transport policies or plans. Inseveral Member States and in a number ofcountries outside Europe, SEAs are alreadycommonly carried out for land use anddevelopment plans. Moreover, strategic

environmental assessments may becomemandatory for a wide range of plans andprogrammes if the Commission’s proposalfor a Directive on SEAs is adopted². In April1997, the Council of Ministers of the Euro-pean Conference of Ministers of Transportformally endorsed the principles of SEA at itssession in Berlin and called for SEA to beput into practice.

3.2. The Commission Work Programme on a SEA of the TEN

Following the adoption of the TEN guide-lines, the Directorate-General for Transport(DGVII) and the Directorate-General forEnvironment, Nuclear Safety and CivilProtection (DGXI) of the European Com-mission started to co-operate closely . A state-of-the-art review showed that a wide range ofSEA methods and techniques already existand that SEA practice in the transport sectoris growing. However, there is little experi-ence in applications on a scale comparableto the trans-European transport network.Thus, a practical-oriented work programmewas developed and adopted by both DGs. Itincluded the following major tasks:

• carrying out a pilot SEA of the overalltrans-European transport network, in co-operation with the EEA and in consulta-tion with the Member States;

• supporting a number of pilot assessmentsof TEN corridors conducted by theMember States in co-operation and withco-financing of the Commission;

• developing a methodological handbookfor SEA of infrastructure networks andcorridors, featuring the findings of paststate-of-the-art studies, case studies andthe above mentioned pilot SEAs. Practi-cal guidance is also given.

The present work forms part of the pilot SEAof the overall trans-European network, which- for pragmatic reasons - has been split inthree main parts:

I the assessment of spatial and ecologicalimpacts of the TEN. This part focuses onthe mere physical impacts of the infra-structure. The idea is to evaluate landuse, disturbance and fragmentation of

1) Decision N° 1692/96/EC on CommunityGuidelines for the

development of thetrans-European

transport network, OJ L228, Volume 39, 9

September 1996

2) Proposal for aCouncil Directive on the

assessment of theeffects of certain plans

and programmes onthe environment -COM(96)511 final

3) European Commis-sion - DGXI, Spatial and

ecological assessmentof the trans-European

transport network:scope, methods, data

and research needs,Workshop proceedings,

Brussels, 24 - 25 April1997.

11Context

nature areas, etc. by using geographicalinformation system tools;

II the assessment of traffic-generatedimpacts such as emissions of greenhousegases, acidifying gases and pollutants,energy consumption, safety and - ifpossible - noise. This task, which requiresthe use of predictive traffic and environ-mental models, will be carried out by aninternational research team in thecontext of the Commission’s 4th Framework Programme on Transport RTD;

III the aggregation of the results of theprevious two steps and the comparativeevaluation of the predicted impacts inthe light of the Community’s environ-mental objectives and targets.

The present methodological study wasconducted as a preparatory phase for thefirst part of the pilot assessment, i.e. thespatial and ecological assessment of the TEN.It builds on the findings of a technicalworkshop on this issue organised by theCommission in April 1997³. The workshopwas attended by representatives of theEuropean Commission (DGXI, DGVII, andEurostat), the EEA and its Topic Centres on

Land Cover and Nature Conservation, anumber of international NGOs, as well asconsultants and scientists.

The work, steered and developed by aDGVII-DGXI-EEA-Eurostat working group,consisted of the following tasks:

• the development of an integrated data-base and Geographical InformationSystem (GIS) on the TEN, includingthematic data maps on infrastructure,land cover, demography, geography,environment and nature;

• the selection and review of indicators forassessing the spatial and ecologicalimpacts of the TEN, taking into accountthe current data availability and limita-tions;

• the development and testing of anumber of GIS assessment techniques;

• the compilation of the results in a GISdemonstration package to allow aninteractive demonstration of indicatorsand methods;

• a broad consultation on the findings anda consensus-building process regardingfull assessment methodology.


4.1. Purpose and Scope of a SEA of the TEN

The SEA is a systematic process of evaluatingthe environmental impacts of a strategicaction and its alternatives. The main aim of aSEA is to ensure that environmental consid-erations are addressed at the earliest appro-priate stage of decision-making and on alevel equal to that of economic and socialconsiderations.

In developing methods for the SEA of theTEN, it is important to keep in mind thegeneral aim of such an evaluation and theinformation that it is expected to generate.The main objective of developing a SEAsystem for the TEN is to raise awarenessabout the environmental consequences ofdeveloping the TEN, to improve planningand to assist investment decisions. The SEAof the TEN should therefore provide thefollowing information:

• an overall evaluation of the extent towhich the current TEN outline plan willhelp to achieve the Community (andother) objectives and targets for environ-mental protection. This evaluationshould distinguish the effects per trans-port mode, the potential modal shift andthe effect of induced traffic;

• a broadbrush impression of the geo-graphic distribution of positive andnegative impacts of the TEN. In particu-lar, the assessment should allow theidentification of potential conflict areas.The SEA will not address alignmentoptions per link or corridor, but can forexample be useful in determining thesize of the corridor that should subse-quently be assessed in more detail by theMember States at a more local planninglevel (e.g. corridor level).

The SEA should produce information that isof greatest relevance to the particular deci-sion-making level it addresses and on a levelof detail that is appropriate. Thus, theassessment of the TEN will differ in scopeand detail with the assessment of the TENcorridors: an assessment at TEN-level willinevitably be more broad and will focus onimpacts of “Community-importance” (e.g.greenhouse effect, acidification, loss ofbiodiversity). Corridor assessments will

4. Methodology

enable more detailed evaluations and willhave to take into account specific and moredetailed national and regional characteristicsand objectives.

4.2. Environmental Objectives and Targets

Strategic assessment is generally consideredas an objective-led process. The purpose of aSEA is to assess to what extent the strategicaction and its alternatives contribute to therealisation of certain environmental objec-tives and targets. A framework of environ-mental objectives and targets is essential forthe identification of proper indicators andfor the evaluation of impact predictions.

The broad (environmental and other)objectives of the TEN are set out in Article 2of the TEN guidelines. For a SEA at the TENlevel, the targets of the 5th EnvironmentalAction Programme (5EAP)4 and its recentreview5 provide a minimum framework. TheWhite Paper on the Common TransportPolicy (CTP) also identified a number ofpriority environmental objectives. As regardsnature conservation, the designated areascovered by the Directives 79/409/EEC(Birds Directive) and 92/43/EEC (Habitats)and other (inter)nationally designatedconservation areas should be incorporated.The present study focuses on Community-level targets. At a later stage, an investigationshould be conducted as to whether and hownational or regional strategic options andobjectives can be included within the frame-work of the SEA of the TEN.

4.3. Impact Indicators Selection

An indicator is a variable that enables anevaluation of the impact intensity for aparticular environmental stake or objective.Environmental indicators can be measuredin physical units (e.g. emissions), on anordinal scale (e.g. landscape quality) or cansimply be described in a qualitative manner(e.g. biotopes).

One of the main aims of the working groupwas to identify possible indicators of spatialand ecological impacts of the TEN. In thecontext of this study, ecological and spatial

4) EuropeanCommission, DG XI

(1992): TowardsSustainability: A

European CommunityProgramme of Policy

and Action in Relationto the Environment and

Sustainable Develop-ment (The Fifth ActionProgramme). Brussels.

5) European Environ-ment Agency (1995):

Environment in theEuropean Union 1995.

Report for the Review ofthe Fifth Environmen-

tal Action Programme.Copenhagen.

Methodology 13

impacts are broadly defined to include thefollowing impact categories:

• Ecological impacts of infrastructure, i.e.impacts that are directly related toecosystems, habitats and species and thedegree of biodiversity. Examples arehabitat loss, habitat fragmentation anddisturbance or loss of species. Indicatorsare mostly related to habitats.

• Functional impacts, e.g. on agricultureand forestry: large-scale infrastructuredevelopment can create barriers, divid-ing functional land units such as parcelsof farmland. They can consume part ofthe agricultural area and may makefurther exploitation impossible substitu-tion effect).

• Impacts for which indicators with aspatial dimension can be used. Anexample of this is noise, which can beevaluated among other ways by estimat-ing the number of people living in thevicinity of the infrastructure or by defin-ing noise sensitive zones (“tranquil”zones) on the basis of a number ofspatial/ecological criteria.

The scope and choice of impacts and indica-tors should match the appropriate policylevel (in this case the EU level). For the SEAof the TEN simple indicators have to beused, which are applicable to the whole ofthe Union. For more local-level assessments(e.g. corridor assessments), other, moredetailed indicators can be used, which also

Figure 4.1: Relevantselected spatial andecological issues for astrategic environmentalassessment of the TEN

Theme Environmental objectives Environmental targets Potential indicator

Biodiversity and safeguard of biodiversity conservation of designated vicinity of sites to thenature conservation areas and the wider environment infrastructure

(5th EAP, CTP, Birds Directive,Habitats Directive, Ramsar density of sites within buffersConvention, Biodiversity around the infrastructureConvention, etc.)

consumption of naturalhabitats

Water resources sustainable use of water reduction of pollution, protection number of crossings of waterwaysresources of watersheds, prevention of

environmental damage from number of water and groundwatermaintenance and improvement shipping activities protection zones touchedof groundwater quality

maintenance of the ecologicalquality of surface fresh water

eduction of discharges intomarine water (5th EAP, CTP)

Coastal zones sustainable development of improve the balance of land use number of coastal zones touchedcoastal zones and their resources and conservation and the use of(5th EAP) natural resources size of coastal zone stretches taken

by linksperforme integrated planningand management

improve co-ordination betweenrelevant EC policies and betweenEC, national and regional policies

Noise avoid exposure to dangerous reduction of noise exposure area under influence along the linksnoise levels for health and quality (especially night-time exposure)of life (5th EAP) number of tranquil zones touched

Land resources sustainable maintenance of avoid disrupting functional units land take (agricultural land, forestry)economic activities

improve land-use planning

Figure 4.1Relevant selected spatial and ecological issues for a strategic environment assessment of the TEN


account for more specific characteristics ofthe immediate environment. In short, theinitial selection of spatial and ecologicalindicators has been made based on thefollowing criteria:

• relevance in regard to Communityenvironmental objectives and targets;

• applicability on EU scale (+ Switzerland).

As a second step, the working group assessedthe feasibility of the impact indicators interms of the availability of suitable data andevaluated the uncertainties involved.

Figure 4.1 summarises the impacts andpotential indicators that were identified bythe working group. The figure also indicatesthe policies that relate to various impacts andindicators. It distinguishes five differentthemes or issues: biodiversity and natureconservation, water resources, coastal zones,noise and land resources. The environmen-tal objectives and targets concerning theseissues primarily follow what is laid out in the5th Environmental Action Plan, the WhitePaper on the Common Transport Policy, theBirds Directive, the Habitats Directive as wellas a number of other international agree-ments. In general, the aim is to protect andconserve the areas concerned and in thecase of noise, to reduce the number ofpeople exposed to certain noise levels.

According to its modal scope this studyfocuses on terrestrial, linear TEN infrastruc-ture, i.e. motorways, conventional and highspeed rail links and inland waterways.

Airports, maritime ports and terminals arenot included, mainly because the TENguidelines do not identify specific plannedprojects for nodal points (with the excep-tion of priority project 10, Malpensa Air-port). Naturally, an assessment requiresproposed plans or actions. For airports theguidelines do contain outline maps, but themaps only subdivide the main Europeanairports into four categories (internationalconnecting points, EU connecting point,regional and accessibility point, and airportas part of an airport system) without indicat-ing any planned projects. As to maritimeports and inland terminals, the guidelinesprovide a general reference, but provide nooutline maps nor an indication of plannedprojects. The Commission has meanwhileprepared a proposal for the integration ofports, inland ports and terminals in thetrans-European network, in a similar way tothat of airports.

In general, an impact assessment consists ofthree different steps:

impact prediction; making an objective esti-mate of the type and magnitude of theactions likely to have future impacts;

impact evaluation; the subjective appraisal ofthe significance of the predicted impacts,taking into account the sensitivity or impor-tance of the receiving environment, and bycomparing (e.g. using multi-criteria analysis)the impact levels with environmental objec-tives and targets;

analysis of the uncertainties that underlie theimpact prediction and evaluation. Onemanner of dealing with uncertainties in a SEAis by basing the analysis on a relative compari-son of alternatives, rather than on absoluteimpact predictions. It is of course importantthat the uncertainties at strategic level are notso large as to make subsequent assessments atmore local levels of decision-making signifi-cantly different (and perhaps contradictory).

The working group focused on the first andthird steps of impact assessment. The currentreport is an attempt to demonstrate howimpacts could be predicted using the indica-tors suggested by the working group. Thereport does not include an evaluation of theimpacts, since the predicted impacts should atthis stage be seen as methodological tests. Anevaluation of the impacts would be prema-ture. However, the working group has consid-ered the degree of uncertainty linked to thedifferent indicators, taking into account thequality of the currently available data.

In this demonstration phase, a range ofassessment techniques was tested, including:

• A proximity analysis of the planned TENinfrastructures to the (legally and scien-tifically) designated sites. This provides avery rudimentary indication of thepotential risk that planned TEN linksmay constitute in terms of disturbance ordamage to habitats.Calculations of singleindicators for the whole network, such astotal land take (per habitat type).

• A vulnerability mapping analysis, i.e. aconfrontation of the TEN alternativeswith sensitive zones. The zonesaredefined on the basis of a combination ofvarious indicators and are evaluatedusing indices of significance.

All assessment methods described in thepresent report were designed for consistency

Methodology 15

with geographical information system (GIS)techniques. A GIS is a computer-basedsystem used for entering, storing, manipulat-ing, visualising and analysing geographicaland tabular data together. It is clear thatcomputer hardware and software play acentral role in a GIS; but data, models,expertise and documentation - orgware - areequally, if not more, important.

GIS tools are particularly useful for environ-mental assessments, because they can com-bine different databases and make overlaysof different maps. The ability to present dataand indicators in the form of maps greatlyfacilitates the understanding and interpreta-tion of the evaluation. It should be notedthat the maps included in the presentdocument are only very small-scale represen-tations of the actual output of the system anddo not do justice to the degree of detail thatis in fact present in the databases.

In order to test the different indicatorsmentioned above, several GIS techniqueswere applied:

• conversion algorithms were used to bringthe data into the appropriate formats.Raster formats proved more suitable foroverlay processes, vector formats wereused for neighbourhood functions;

• spatial analysis operations were carriedout including masking, proximity analy-sis, overlay, and data set re-combination;

• generation functions enabled the crea-tion of specific support data such assampling grids and perimeters.

Each individual indicator should be consid-ered a result of a series of data transforma-tion and manipulation steps using several ofthe above activities.


5. Data and Information Sources

A first important task in the demonstrationwork consisted of bringing together allnecessary data in a single information systemand checking their quality and appropriate-ness for use in the assessment. Guidance onthe range of assessment methods to be testedwas obtained through a review of variousexisting studies and SEA practices, which canbe found on a regional, national and, to alesser extent, international level.

5.1. Major Data Sources and Limitations

5.1.1. Transport Infrastructure DataThe use of transport infrastructure in theSEA requires information on the location,the mode, the type and the state of theEuropean Networks. Eurostat-GISCO pro-vided a number of spatially-referenced datasets covering the necessary information.These databases were reorganised to en-hance their suitability for analysis.

It is important to note that the infrastructuredata have important limitations. The outlineplans that are included in the TEN guide-lines only give very rough indications aboutthe (future) location of the new infrastruc-ture links. For some links, such informationwas completely lacking and their location hasbeen entered in the Geographical Informa-tion System on the basis of a “best guess”. Inaddition, there is very limited information atEuropean level currently on the status of theplanned infrastructure links. In the case ofroads, conventional rail and inland water-ways, the guidelines do not distinguishupgrades of existing infrastructure fromcompletely new links. Of course, this distinc-

Theme Content Source Year

trans-European rail Network as published in the Official Journal Eurostat 1996

trans-European road Network as published in the Official Journal Eurostat 1996

trans-European water Network as published in the Official Journal Eurostat 1996

European Roads 1/1.000.000 road network Eurostat updated 1996

European Railways 1/1.000.000 rail network Eurostat updated 1997

Specific TEN projects 13 projects as published by the CEC EEA 1997

Figure 5.1 Available digitalised data concerning European TEN infrastructure

tion is an elementary requirement for aproper assessment of the ecological impactof a planned road or rail connection.

Also, comprehensive and digitalised informa-tion concerning, for example, the number ofmotorway lanes or rail tracks is lacking.

The European Commission has initiated anumber of activities to enhance the existingdatabase on the TEN projects. These activi-ties involve co-operation with the MemberStates and different international organisa-tions (e.g. WERD and CCFE). Through thisdata collection exercise, the Commission’sGIS will gradually contain more informationabout the quality of the TEN links. However,it is not expected that more detailed infor-mation concerning the location of new linkswill be obtained.

It should also be noted that the scale of theavailable maps for the whole trans-Europeantransport network is much larger than thescale of the maps used in any SEA applica-tions used thus far. In fact, a recent study ona SEA concluded that “any GIS mapping atlarger than 1:200.000 scale would be inad-equate for SEA purposes”6.

The working group that prepared thepresent document fully acknowledges thelimitations of the available infrastructuredata and therefore stresses that the outputshould not be seen as actual predictions ofthe possible environmental impacts of theplanned TEN links. The aim of the currentstudy was primarily to investigate how indica-tors and GIS could best be used and to testwhether meaningful results can be pro-

6) Steer Davies Gleave,“State of the art on

strategic environmentalassessment for transportinfrastructure”, prepared

for the Directorate-General for Transport of

the EuropeanCommission, Final report

July 1996, p. 113.

Data and Information Sources 17

duced, taking into account the quality of theavailable data.

5.1.2. Environmental DataThe term “environmental data” refers to alarge variety of spatially referenced data sets.It is the combination of the different datathat enables the creation of a model of theEuropean terrain. From the basic material,derived data sets are produced for assess-ment purposes.

The figure below describes the core raw dataused for the present study. All databases havea nominal scale of 1/1.000.000, unlessspecified otherwise.

As Figure 5.2 illustrates, some of the databasesare rather heterogeneous as far as age isconcerned (land cover and populationdensity in particular). The stability columnprovides an indication of the need to updatethe data sets. “Stable” indicates that thedatabase concerned does not require fre-quent updating (e.g. topography, coastalzones and water pattern) whereas “dynamic”databases need regular updating over time(administrative units and nature inventories).

The available environmental spatially refer-enced data on the EU level also suffercertain limitations, either in scale, coverageor content. Certain impacts (e.g. impacts onsoil quality and groundwater) had to beexcluded from the assessment because of a

Figure 5.2Available (digitalised) data for environmental assessment on European scale

theme content source year stability

Coastal zone derived from administrative regions Eurostat 1996 stable

Topography 1 km resolution digital elevation model Eurostat 1994 stable

Administrative units mask defining the European Territory Eurostat 1996 dynamic

Population population density by smallest Eurostat 1981/93 averageadministrative units SFSO(CH)

Land cover 1/100.000 CORINE land cover classes ETC/LC 1986/97 averageof the EU

Ecological regions 1/3.000.000 digitised map of the ETC/NC 1997 stableEuropean ecological regions

Designated sites sites designated under international ETC/NC 1993 averageconventions version

Nature inventories sites complying to scientific criteria of ETC/NC, 1989/95 dynamicimportance to nature conservation Birdlife Int.

Water pattern surface water pattern CEC DGXI, 1985 stableETC/IW

lack of data. The EEA and its Topic Centresare making continuous efforts to solve theproblems encountered in the environmentaldata sets.

The main problems are the following:

• coastal zones have only been defined in alinear way; there is no EU definition thatenables the identification of the real ortotal area that should be considered;

• the database on population density (tobe used for noise impact assessment) isheterogeneous in the sense that datacome from different years; Eurostat-GISCO is working on an improveddatabase;

• the land cover data is a compilation ofvarious (national) databases. The EUdatabase is not yet fully harmonised, butthe European Topic Centre on LandCover is working on this issue;

• the databases on designated sites andnature inventories only contain informa-tion on the size and the centre co-ordinates of the sites; information on theactual site boundaries is not (yet) avail-able. DGXI has started initiatives toobtain information on site boundaries;

• the database on water patterns is heterogeneous and coverage is incomplete.The European Topic Centre on InlandWaters is currently involved in updatingthe database.


6. TEN Alternatives

6.1. Selection of Alternatives

The comparison of alternatives forms thecore of each SEA. In the case of the TEN,the choice of alternatives must focus onmodal choices and on varying degrees ofnetwork extensions.

In line with the concept of “tiering” and forpractical reasons, it seemed appropriate thatan assessment of an European network doesnot consider alternatives for each individuallink, but rather illustrates the (potential)environmental impact of different overallnetwork settings (or scenarios). The selec-tion of the alternatives to consider is an issuein itself that requires further developmentand also a wide consultation process. Itdemonstrates that the European networkassessment should primarily be seen as a toolto raise awareness about the possible envi-ronmental consequences of the TEN, whichmay compare different general orientationsfor the TEN.

For this pilot assessment, six alternativenetwork developments have been selected.They must be considered only as examples ofpossible options. The six network alterna-tives are:

1. the existing TEN network (existingprimary and secondary road, rail andwater infrastructure is not included)

2. the existing TEN network plus thefourteen priority projects that wereagreed at the 1994 Essen EuropeanCouncil (except project N° 10, theMalpensa Airport)

3. the existing TEN network plus allplanned rail projects

4. the existing TEN network plus allplanned inland waterway projects

5. the existing TEN network plus allplanned road projects

6. the complete multi-modal TEN

The first alternative represents the “do-nothing” option (or present situation) thatcan be considered as a reference situationagainst which other alternatives can beevaluated. Variant 2 could be considered a“do-minimum” alternative. The other vari-ants represent differing degrees of imple-mentation of the TEN plans. Variants 3 to 6are based on the outline maps 2 to 4 inAnnex 1 of the TEN guidelines; plannedprojects are those projects indicated as“planned” in the outline maps. The projectsof specific interest are those projects thatwere adopted by the European Council at

Project Mode Member States Location

1 High speed rail DE, AU, IT Berlin-Nürenberg, Brenner axis

2 HSR UK, FR, BE, NL, DE Paris-Brussels-Köln-Amsterdam-London

3 HSR FR, ES Madrid-Montpellier, Madrid-Dax

4 HSR FR, LU, DE Paris-Karlsruhe, Mannheim, Luxembourg

5 Rail NL, DE Betuwe Line

6 Combined FR, IT France-Italy

7 Road GR Greek Motorways

8 Road PT, ES Motorway Lisboa-Valladolid

9 Rail IR L link Cork-Dublin-Belfast-Larne-Stranraer

11 Rail/road DK, SE Øresund link (Denmark-Sweden)

12 Rail/road SE, FI Nordic Triangle

13 Road IRL, UK, NL, BE Ireland/UK/Benelux road link

14 Rail UK West Coast main line

Figure 6.1 Decision No 1692/96/EC Annex III priority projects

TEN Alternatives 19

Essen on 9 and 10 December 1994 and arecontained in Annex III of the TEN guide-lines. Project n° 10, the Malpensa Airport, isnot considered in this report, as this study’sscope is limited to linear infrastructureplans.

Hence, the alternatives should not be inter-preted as true policy options. They reflect(hypothetical) extreme situations, but makepossible the identification of the range ofpossible outcomes concerning the environ-mental impact of the TEN and allow an

Figure 6.2The three transport modes included in this study and a multi-modal overview including the 14 specific projects

evaluation of the modal liability for environ-mental impacts.

As mentioned before, plans for non-linearinfrastructure (maritime ports and airports)have not been taken into account.

The working group acknowledges that bydefining the variants in the above mentionedway, the “added value” for decision-making islimited, though important in raising aware-ness about the environmental implicationsand modal liability of the TEN development.


It would be preferable to define alternativesthat represent true strategic options andpossibly different investment levels. Onecould for example think of a developmentvariant that focuses on key corridors, adevelopment variant that primarily aims atstrengthening cohesion, a variant thatrequires limited investment and a variantthat requires high investment levels. How-ever, definition of such variants requiresselection or prioritisation of individual linksand therefore requires careful preparation

and extensive consultation and discussionwith all parties involved, in particular theMember States. The working group consid-ered the definition of such TEN strategyoptions outside the scope of its primarilymethodological work, but recommends thisas an important issue for further research.

Figure 6.2 consists of four maps showing theexisting and planned infrastructure that isconsidered in the current study.

Figure 6.3 Example of the multi-modal TEN network compared to the national and regional network

mode existing TEN (1996) planned TEN (2010)* existing overall Europeaninfrastructure**

upgrade new

road 49,598 n.a. +26,875 49,024***

conventional rail 48,477 n.a. +1,372 155,836

high speed rail 4,901 +14,408 +10,088 2,406

inland waterways 12,239 n.a +1,412 30,191

* All planned roads, conventional rail links and waterways are assumed to be new** Source: European Commission, “EU transport in figures - statistical pocketbook”,

2nd issue 1997 (figures refer to 1995)*** Motorways only

Figure 6.4 Estimated length of the trans-European transport networkcompared to the overall European infrastrukture (kilometres)

TEN Alternatives 21

6.2. The TEN in Perspective

In order to facilitate the understanding ofthe scope of the trans-European transportnetwork, Figure 6.3 sketches the density ofthe TEN compared to the national andregional road and rail networks.

Figure 6.4 shows a different way of putting thetrans-European network into perspective. Itprovides graphical estimates of the length of

the planned links. The estimates wereperformed at a cartographic scale of 1/1.000.000 and provided the following cumu-lative lengths of the network alternatives.The calculations are based on the assump-tions mentioned earlier, i.e. all plannedroads, conventional rail and inland water-ways are considered to be new infrastructure.A comparison of these measured values ismade with the reference values made byDGVII.


7. Nature and Biodiversity

7.1. Main Objectives Concerning Nature and Biodiversity

The importance of safeguarding nature andbiodiversity is undisputed. It is also one ofthe key priorities of the EU’s 5th Environ-mental Action Plan.

The construction and use of roads, railwaysand canals produces several negative effectson nature and biodiversity. The infrastruc-ture in itself may cause a loss of nature sitesor lead to partitioning and isolating ecosys-tems and species populations. Secondly, thetraffic that is using the infrastructure candisturb nature through noise, vibrations andaccidents.

Different conventions and agreements havebeen established to protect nature sites andbiodiversity, ranging from internationallegislation to regional programmes. Atinternational level, the United NationsConvention on Biological Diversity (CBD)plays an important role. The signatories tothe Convention (including all EU MemberStates) have committed themselves to theconservation of biological diversity and thesustainable use of its components.7 In 1998,the European Commission adopted a Com-munity Biodiversity Strategy in view of thefulfilment of the European Community’sobligations under the CBD.

At European level, the 5th EnvironmentalAction Plan obviously constitutes a centralelement, together with the Directive on theconservation of natural habitats and of wildfauna and flora (commonly referred to asthe Habitats Directive) and the Directive onthe conservation of wild birds (Birds Direc-tive)8. The EU policy aims to create anecological network of protected areas acrossthe Union, the so-called Natura 2000 net-work. The network will consist of areasdesignated under the Habitats and BirdsDirective.

When it comes to assessing a potentialconflict the TEN might create in respect tonature conservation interests and biodiver-sity maintenance, the question ariseswhether only impacts on formally designatedsites should be examined or whether natureand biodiversity should be interpreted in a

broader sense. The latter would includeother sites, like habitats and species, as wellas buffer-zones and corridors. This leads totwo different types of analyses.

The first analysis focuses on the vicinity ofthe existing and planned TEN infrastruc-tures to the (legally and scientifically)designated sites of nature conservationinterest. If a link is located in or near adesignated site, it is assumed to constitute apotential risk to the site concerned.

The second evaluation method specificallyconsiders the partitioning effect of the TENinfrastructure. The partitioning, or fragmen-tation effect constitutes an important ele-ment of the adverse environmental impactsthat infrastructure may have. The evaluationuses information concerning designatedareas, but also uses other information.

7.2. Vicinity to Sites of Nature Conservation Interest

A first way of estimating the impact of theTEN infrastructure on nature andbiodiversity is through identifying the vicinityof planned and existing TEN links to desig-nated nature sites. If a link runs through oris located near a designated site, it may beassumed that there is a potential risk ofadverse impacts to the site concerned.

This type of analysis has been suggested andtested before by the Royal Society for theProtection of Birds (RSPB) in its report “Theimpact of trans-European networks onnature conservation: a pilot project” (1995).

It should be noted that the strategic networkassessment can only identify potential risks;the actual impacts naturally depend on theexact location of the link (and site), the typeof construction and the mitigation measuresthat are taken. However at the strategic level,it is neither possible nor appropriate to takeaccount of detailed information per link (inaccordance with the principle of “tiering”,this should be left to corridor or projectassessment). The aim of the network assess-ment is simply to present an overall pictureof the possible environmental impact of theTEN. Of course, by presenting potential risks

7) The CBD defines“biological diversity” as

the variability amongliving organisms from allsources including, interalia, terrestrial, marine

and other aquaticecosystems and the

ecological complexes ofwhich they are part; thisincludes diversity within

species, betweenspecies and of eco-

systems.

8) Council Directives 92/43/EC and 79/409/EC.

Nature and Biodiversity 23

rather than actual impacts, the assessmentpaints a “worst-case scenario”: not all poten-tial risks will become real impacts.

7.2.1. Methodological ApproachUsing a methodological approach, the firstquestion to be answered is what sites aretaken into account. Several thousand sitesare defined throughout Europe. The major-ity, however, can only be mapped on a localscale. Digitalised information on the Euro-pean scale is currently only available for sitesdesignated on a formal international orEuropean basis.

Several limitations of the data must be noted.Firstly, the existing site lists contain onlyextremely limited information. For example,the actual site boundaries are in general not

Sites Description

Designated sites protected under community legislation

Bird Directive (SPA) Directive 79/409/EEC provides a general system of protection for all speciesof wild birds. Its purpose is to regulate the hunting, capturing, killing andsale of such species. Its main objective is the conservation and restoration ofnatural habitats. Data is not complete for all countries (only a sample of siteshas been tested).

Designations under international conventions

European Diploma Network of the CoE to promote better management of important naturalsites, scientific research and awareness raising.

Barcelona Convention (16 Feb. 1976). The Convention for the Protection of the Mediterranean Seaconstitutes the main legal instrument for the protection of theMediterranean environment. It is a framework convention that has beenprogressively developed through the adoption of further instruments andprotocols.

Biosphere Reserve The UNESCO “Man and the Biosphere” programme aims to encourageinterdisciplinary research, training and demonstration activities with a viewto improve the understanding and management of the resources of thevarious ecosystems. A BR aims to reconcile the conservation of biodiversitywith its sustainable use by involving local communities.

Biogenetic Reserve A network of the CoE to include sites with special emphasis on endemicspecies and biodiversity.

World Heritage The UNESCO World Cultural and Natural Heritage Convention preservesoutstanding ecosystems and landscapes.

Ramsar Convention (2 Feb. 1971) The Convention on Wetlands of International importance, withemphasis on waterfowl habitat, aims to combine long-term national policieswith co-ordinated international action.

Scientific inventories

CORINE Biotopes A scientific inventory of sites important for nature conservation, regardlessof the designation given. Sites were selected according to the presence ofspecies, habitats or landscapes. The data varies considerably in age.

Important Bird Area (IBA) Birdlife International used its global network of national organisations tocreate a database to monitor the conservation of bird species and habitats,identify and secure conservation of the most important sites, and build agreater understanding of the relationships required for sustainabledevelopment.

Figure 7.1Site description according to its legal statute

known. Instead, the databases provide centreco-ordinates and an indication of the surfaceof the site. Secondly, certain databases donot cover all countries and/or contain dataof varying age (this is particularly true forthe CORINE Biotopes ). Thirdly, areas maybe double counted, as the same site may bedesignated under various conventions. Atypical example is the Camargue area in thesouth of France. The problem of doublecounting may be solved in the future oncemore detailed site descriptions becomeavailable under Natura 2000.

The nationally designated sites have notbeen considered in this study either, due tothe current lack of consistency and heteroge-neity of the data (definitions of the sitesdepend on national legislation that vary


enormously). The EEA European TopicCentre for Nature Conservation is compilinga list of all national sites (without howeverharmonising the definition of the sites) inthe common database on designated areas.

As already indicated, the information con-cerning the location of new TEN links, aswell as the boundaries of the designatedsites, is rather imprecise. The databases onthe TEN provide only rough indications ofthe locations of new links (sometimes “bestguesses”) and the nature sites can often onlybe represented symbolically by a circle with aradius that is proportional to the site surface.

In order to take account of these uncertain-ties, a buffer-zone around the infrastructurecentre lines can be used. Of course, thewidth of the buffer-zones is a significantdetermining factor for the outcome of thestudy. On the one hand, the buffer-zoneshould be wide enough to reflect all poten-tial impacts in a reasonable way. On theother hand, the buffer-zone should benarrow enough to avoid unnecessary overes-timation of the possible risks. Ideally thebuffer-zone should be a fair reflection of the“average uncertainty” concerning the loca-tion of the nature site and the plannedinfrastructure link. Simultaneously, it shouldtake account of the fact that the impact ofthe infrastructure - through for example

noise and vibrations - stretches further thanthe actual borders of the infrastructure itself.An additional complication arises from thefact that many sites have an oblong shape,thus are not compact. The working groupacknowledges that by making randomsample surveys, an indication of the “averageuncertainty” of the geographical informationmight be obtained. Such sample surveys havenot been carried out, but could be a subjectfor further research.

An indication for choosing the width of thebuffer-zones can be obtained from otherstudies. The RSPB used for example corri-dors of 2 and 10 kilometre wide9. TheFrench Ministry of Transport generallyapplies 10 kilometre buffer-zones in theassessments that are carried out on corridorlevel10. Consequently it was decided for thecurrent analysis, a 10 kilometre buffer-zoneshould apply.

7.2.2. The ResultsThere are different ways in which the resultscan be presented. A first option is to identifyper TEN development variant, the numberof sites that is potentially at risk. Figure 7.2shows the proportion of potentially threat-ened sites for the six network alternatives,according to its statute. It suggests that thedevelopment of the planned road networkwill increase the number of potentially

9) RSPB, idem.

10) See Ministère del’Equipement, des

Transport et duTourisme - Direction

des Routes, “Flux est-ouest au sud du bassinparisien, dossier pourun débat”, Décembre

1994. It must be notedthat the scope and

nature of this study isslightly different from

the network assessmentat European level.

variant (% of sites within 10 km)

site type and inventories 1 2 3 4 5 6 number of sites

S.P.A. 53 53 55 53 60 61 743 (EU12)

Barcelona C. site 43 44 44 44 47 49 94

Eurodiploma site 44 53 47 44 56 59 34

Biosphere Reserve 66 68 66 68 72 75 47

Biogenetic Reserve 52 53 54 52 57 59 253

World Heritage 20 20 20 20 20 20 5

Ramsar C. site 71 72 72 71 79 80 246

CORINE Biotopes inventory 49 50 51 49 56 58 7740 (EU12)

Important bird area 51 52 53 51 58 59 1644

Variants: 1) existing TEN network2) existing TEN network + 13 specific projects3) existing TEN network + all planned rail projects4) existing TEN network + all planned inland waterway projects5) existing TEN network + all planned road projects) complete multi-modal TEN

Figure 7.2 Designated sites potentially affected by the development of the trans-European transport network.Status in mid 1997 (overlaps between sites not taken into account)


threatened sites by 14 percent. The develop-ment of the rail projects has a considerablesmaller impact. This is barely surprisingconsidering that the extension of the roadnetwork is assumed for this exercise toconsist of around 27,000 kilometres of newinfrastructure, whereas the rail network isonly extended by around 11,000 kilometres.

The presented results should only be consid-ered as hypothetical, considering the uncer-tainties both for the infrastructure data

(distinction between new links and upgrad-ing) and for the environmental data (doublecounting of sites).

Figure 7.3 gives a geographical representationof the sites that are potentially at risk. Themap distinguishes four different risk catego-ries. The sites indicated with a green dot arenot within the 10 kilometre buffer-zone ofany of the existing or new TEN infrastructurelinks. The sites represented by a blue dot are“threatened” by one or two links, the orange

Figure 7.3Indication of potential conflicts between complete TEN and nature sites.Nature sites classification according to neighbouring links


ones by 3 or 4 links and the red sites by 5 ormore links.

Instead of showing the results per site, theoutcome can also be presented per link.Figure 7.4 shows the average number ofdesignated sites at risk for each link. Themap is based on a density calculation perlink because a longer link naturally has ahigher chance of neighbouring sites than ashorter link.

Combining approaches allows the identifica-tion of problematic links and sites with a

high probability to exposure. It could beconsidered that special caution is requiredwhen dealing with projects on these loca-tions (red flag approach).

7.3. Partitioning of Land

A particular effect that linear infrastructurehas on nature - and landscape - is that itpartitions land or, more specifically, frag-ments habitats.11 Habitat fragmentation canbe defined as the splitting of natural ecosys-tems into smaller and more isolated units

Figure 7.4 Indication of potential conflicts between complete TEN and nature sites.TEN infrastructure classification according to neighbouring nature sites.

11) More detailedresearch on the

problem of habitatfragmentation is

planned in theframework of COST

Action 341 (EuropeanCo-operation in the

field of Scientific andTechnical Research).


thus endangering the survival of animal andplant species and communities. The declinein biodiversity witnessed after a land unit hasbeen split into smaller fragments depends onthe condition of the original unit, the sizeand distribution of the newly formed partsand the barrier effect of the infrastructuredelineating the unit. Moreover, partitioningcan constitute a form of visual intrusion tovaluable landscapes.

7.3.1. Methodological ApproachFor the above stated reasons, a good indica-tor for measuring the partitioning effectshould take both size and ecological qualityinto account. The size of a unit can bedetermined graphically; yet to assess itsecological quality is more difficult. Twoapproaches to the quality problem have beenexplored, i.e. quality evaluation based on:

I. the proportion of the land unit area thatis designated by international conven-tions

II. the proportion of the land unit area thatis covered by forest or semi-naturalhabitats.

If a land unit scores above the Europeanaverage on both observations, it is consid-ered of high quality; if it fails on both ac-counts, it is considered of lower quality.

For the present exercise, only the TEN linkswere included. The other existing primaryand secondary road and rail infrastructurewas not taken into account. It follows thatthis is a simplistic assumption and that thederived results can only be considered as“dummy” results that merely serve to illus-trate the methodology. The working group

acknowledges that further testing is requiredto determine how and to what extent influ-ences other than the TEN, can be dealt withproperly.

Another simplifying assumption has beenmade-to ignore original land units (deline-ated by the existing TEN network) that weresmaller than 1000 km². This was done purelyfor scale reasons. It does not imply thatsmaller land units of biological interest areconsidered to be unimportant. However,because of the rough scale of environmentaldata on EU level and the uncertainty of thelocation of TEN links, it is not possible, atthis level, to evaluate effects on smaller landunits. This is a task for national, regional orlocal level assessment.

A similar type of analysis has been tested bythe French consultancy BCEOM for theassessment of national infrastructure plans(“Schémas Directeurs”).12 The French studyconsidered all main railway lines, all roadsand motorways with an average daily trafficof more than 4000 vehicles per day, all canalsand all natural fragmentations (such asvalleys). Even though large-scale maps wereused for other parts of the assessment,BCEOM used relatively detailed maps for theevaluation of habitat fragmentation. Themethodology proved very feasible in a testapplication at a regional level.

7.3.2. The ResultsAgain there are different ways in which theresults can be presented. Figure 7.5 shows agraph with the number of land units com-pared to their cumulative surface. The greenline represents the existing situation; the redline indicates the situation after implementa-

12) BCEOM, “Etudestratégique d’impactsur l’environnement -essai méthodologique”,Ministère de l’Environ-nement - Direction dela Nature et du Paysage- Sous-direction del’Aménagement et duPaysage, Décembre1994.

Figure 7.5Number of land units and their cumulative surface - existing and TEN planned situation


tion of the trans-European transport net-work. The red line runs clearly below thegreen line. This illustrates that the samecumulative area (for example 2,000,000 ha)is formed by more individual land units afterthe realisation of the TEN than before. Thedisadvantage of this presentation is that itdoes not take account of the quality of thefragmented land.

Figure 7.6 below illustrates the partitioningeffect taking into account the characteristicsof the land units. Map 7.6-A shows the

existing situation and map 7.6-B shows thesituation after the development of the TEN.Map 7.6-C illustrates the differences betweenmap 7.6-A and 7.6-B. The green areas onmap 7.6-C have not changed due to theexecution of the TEN; whereas the brownareas have deteriorated from “high quality”land (with an above average proportion ofdesignated land and an above averageproportion of land covered by forest or semi-natural habitats) to “poor quality” land (abelow average proportion of designated landand forest/natural habitats).

Figure 7.6 Partitioning of Land effect on the complete TEN

A. Quality of land unitsaccording to size and

ecological criteria.Existing situation.

B. Quality of land unitsaccording to size and

ecological criteria.Complete TEN planned

situation.

C. Induced land qualitychanges by land

partitioning effect ofthe complete TEN

(Drop in indicator valueset at 20%).

Water Resources 29

8. Water Resources

8.1. Main Objectives Concerning the Protection of Water Resources

Infrastructure building and use can affectwater quality through pollution from exhaustemissions, de-icing salts, weed control alongroads and railways, fuel spills, sewage fromtrain discharges onto tracks, and pollutantsassociated with transport of hazardousmaterials. Transport infrastructure buildingcan furthermore affect the hydrologicalfunctioning of ground and surface water, e.g.can disturb the groundwater table and affectthe course and run-off of surface water.

Surface and groundwater quality are keyelements in a healthy environment as itforms the very basis of life itself. Nationaland international networks monitor waterquality , and various kinds of protectionmeasures are implemented throughout theUnion accordingly.

The European Union Fifth EnvironmentalAction Programme identified the mostimportant problems and set priorities. Withrespect to groundwater and surface freshwa-ter, it stressed the need to integrate resourceconservation and sustainable use criteria intoother policies including agriculture, industryand land-use planning. Concerning thequalitative aspects, and taking into accountsurface freshwater, it called for the safe-guarding of existing high quality and theimprovement of the ecological quality ofwater.

The proposal for an EU Action Programmefor Integrated Groundwater Protection andManagement (COM(96) 315 final), adoptedby the Commission in August 1996, togetherwith the proposal for a Water FrameworkDirective (COM(97) 49 final), issued inFebruary 1997, aim to protect groundwater,inland surface waters, estuaries, coastalwaters and groundwater, which form the

framework for the water policy. The WaterFramework Directive would require MemberStates to prepare a programme of measuresto attain “good” surface water andgroundwater status by the end of 2010.

There are some regional and internationalagreements that concur to the same broadobjectives and targets. Regional agreementsinclude the Rhine Action Plan, the ElbeAction Programmes, the Helsinki Conven-tion - Baltic Sea Joint Comprehensive Envi-ronmental Action Programme - 1993 to2012, the Oslo and Paris Commission[OSPAR] - North Sea Ministerial Confer-ences, the Hague Conference of 1990, theMediterranean Action Plan, and the ArcticMonitoring and Assessment Programme.International agreements include the Con-vention on the Protection and Use ofTransboundary Watercourses and Interna-tional Lakes.

8.2. Crossing of Surface Waters

8.2.1. Methodological ApproachIn principle, a similar type of indicator as forbiodiversity and designated areas could beused to identify the surface water areas orwater catchment areas at risk by determiningthe vicinity to new infrastructure links.Similar to the former chapter, the actualimpact on water quality and hydrology areextremely difficult to predict as it wouldrequire accurate knowledge and data, forexample, on the complex relation betweenthe emissions of pollutants and their disper-sion in the water. Again, the strategic assess-ment at European level can only aim toidentify potential impact risks on the basis ofa very simplified risk indicator.

It did not appear feasible to consider watercatchment areas simply because EU-widespatially referenced information on the

Figure 8.1Water crossings of the existing trans-European transport network

observed crossings road rail

number of crossings of the existing TEN network 3373 6050

density of crossings of the existing TEN network 6.8/100km 8.9/100km


Figure 8.2 Comparison of the number of water crossings under different TEN variants

Alternative road rail

existing TEN network 100 100

existing TEN network + 14 projects of specific interest 108 106

existing TEN network + all planned rail projects 100 116

existing TEN network + all planned road projects 166 100

complete multi-modal TEN (rail and road links only) 166 116

Figure 8.3 Complete Multimodal TEN classification according to potential conflicts with surface water

Water Resources 31

quality of surface water and groundwaterresources is lacking. Therefore, only a verysimplistic, morphological indicator wastested. It is based on the crossing of surfacewater. In other words, the density of cross-ings, measuring the potential proximitybetween links and surface waters, was consid-ered as an indicator for environmental riskto surface waters.

For self-evident reasons, the TEN inlandwaterways were not included in this exercise.

8.2.2. The ResultsFigure 8.1 summarises the number of watercrossings and the density of crossings (per100 kilometre of infrastructure) on theexisting situation. It suggests that on averagethe rail network crosses more water patternsthan the road network.

Figure 8.2 presents the increase in thenumber of crossings compared to the exist-ing network. The figures are given as indices(existing network = 100). The table suggestsan extremely large increase of water cross-ings due to the construction of new roads.However, similar to all results on the previ-ous chapter, the outcomes should not beseen as estimates of the actual impact, butrather as “dummies” to illustrate the outputof a methodology.

Results can also be presented per link.“Critical” links can be identified by measur-ing the density of surface water crossingsthey induce. Figure 8.3 includes a map drawnfor the complete TEN variant and providesan illustration of this type of presentation.


9. Noise

9.1. Main Objectives Concerning Noise

Noise pollution is one of the most tangibleand direct effects related to the use oftransport infrastructure. The importance ofthis issue on urban areas was pointed out inthe 5EAP - where it was shown that morethan 16% of the Community populationsuffers because of noise levels over Leq 65dB(A) during the night, mainly resultingfrom road and air traffic. Noise pollutionhas also recently been highlighted in theCommission’s green paper on a future noisepolicy13.

Noise is an impact that can be effectivelymitigated at project level (e.g. through theuse of tunnels, cuttings, noise screens). Still,it is also important to consider the impact atearlier planning stages, in particular at thelevel of route determination.

Noise emissions not only depend on trafficcharacteristics (e.g. traffic intensity, speeds)and vehicle technology but also on theinfrastructure design and materials. Further-more, the degree of nuisance that noiseemissions cause is very much determined bylocal elements, as noise propagation de-pends largely on the typology of the localenvironment (i.e. topography, existingobstacles and screens such as buildings orvegetation) and meteorological conditions.This means that although noise levels can beeffectively measured and modelled at a localscale, it is much less straightforward tomodel noise impact quantitatively at Euro-pean scale.

Nevertheless, two approaches have beentested. The first approach considered thepopulation within buffer-zones around theexisting and new infrastructure. The secondapproach consisted of the identification ofso-called “tranquil zones” that could be atrisk due to the TEN infrastructure develop-ment. In fact, the two approaches illustratetypical issues when assessing noise. On theone hand, the aim is to avoid noise exposurefor people in their dwellings, offices andwork places, and on the other hand, the aimis to maintain “tranquil areas” both forreasons of nature conservation and recrea-tion.

9.2. People in the Vicinity of Infrastructure

9.2.1. Methodological ApproachThe increase in the number of peoplesusceptible to noise disturbance could beused as an indicator. Ideally, the indicator fornoise impact should take into account thetraffic intensity for the infrastructure con-cerned and identify critical noise contours(e.g. 65 dBA). However, this would requiretraffic forecasts and modelling, which isoutside the scope of this report.

In order to measure the increase in thenumber of people susceptible to noisedisturbance, a database describing thedetailed position of dwellings, offices andother working places (e.g. special ones suchas schools and hospitals) for the entireEuropean Union would be necessary. Such adatabase is, however, not available. The most

13 ) EuropeanCommission (1996),Future Noise Policy,

European CommissionGreen Paper,

COM(96)540 final

Figure 9.1 Population susceptible to noise exposure due to TEN infrastructure

Influence zonesAlternative 250 m 1 km 5 km

Existing TEN network 100 100 100

Existing TEN network + 14 projects of specific interest 104 103 101

Existing TEN network + all planned rail projects 107 105 102

Existing TEN network + all planned water projects 101 100 100

Existing TEN network + all planned road projects 109 107 105

Complete multi-modal TEN 116 113 107

Noise 33

detailed information is on population and isavailable at local administrative level (NUTslevel 5, e.g. communes). By assuming thatthis population is evenly distributed over theterritory of the considered administrativelevel, one obtains a fairly accurate popula-tion density map at European scale.

An additional problem in assessing noiseimpact is that its harmful effects are ratherlocal. Risks will generally only occur within a250 metre buffer-zone around the infrastruc-ture. Inhabitants living at greater distanceswould only perceive disturbance from theconstant background noise from, for exam-ple, a motorway. On the other hand, theuncertainty around the exact location offuture TEN links is quite large. Partly to takeaccount of this uncertainty, the use of a 10kilometre wide buffer-zone was suggested inChapter 7 (analysis of nature sites). Takingaccount of the different kind of impacts ofeach subject, namely its spatial behaviourusing a similar buffer-zone for estimatingpossible noise impacts, would lead to adisproportionate overestimation of theeffect. Therefore, a 1 kilometre and a 5kilometre buffer exercise was performed.

9.2.2. The ResultsFigure 9.1 presents the proportional increaseof the population in the vicinity of the TENinfrastructure. The table sketches outcomesfor the different TEN variants using a 250metre, a 1 kilometre and a 5 kilometrebuffer-zone. The results are presented in theform of index figures (existing network =100).

As in the case of all other indicators, one canargue that for a proper evaluation of theresults, the effects of the different TENvariants should not be compared to theexisting trans-European network, but to theeffect of the general existing infrastructure.

Figure 9.2Population exposed to noise impacts per TEN variant

The results can also be presented in dia-grams (Figure 9.2) or on maps. Figure 9.3shows an overlap of the population densityand the multi-modal network.

9.3. Crossing of Tranquil Zones

9.3.1. Methodological ApproachThe French consultancy BCEOM suggestedin its 1997 report for the Ministry of Environ-ment to identify “tranquil zones.” This is to acertain extent comparable to the definitionof non-fragmented areas in Chapter 7.Tranquil zones have been defined as (semi-)natural areas - forests, semi-natural land andwetlands - that are at least 7.5 kilometresremote from existing TEN infrastructure14.

Population density mapand multi-modal TEN network Figure 9.3

14) Ministère del’environnement (1997),Evaluation environne-mentale des schémasd’infrastructures detransport - Essaiméthodologique TheFrench report arguesthat, on average, noisegenerated by transportinfrastructure can bedetected as far as 7.5kilometres from thesource track.

incr

ease

120

115

110

105

100

95

250 m

Var

. 1

Var

. 2

Var

. 3

Var

. 4

Var

. 5

Var

. 6

incr

ease

120

115

110

105

100

95

1 km

Var

. 1

Var

. 2

Var

. 3

Var

. 4

Var

. 5

Var

. 6

incr

ease

120

115

110

105

100

95

5 km

Var

. 1

Var

. 2

Var

. 3

Var

. 4

Var

. 5

Var

. 6


This excludes areas where noise producinghuman activities are intrinsically present(urbanised areas, areas with predominantindustrial and agricultural activities).

Although noise generated by transportinfrastructure can be detected as far as 7.5km from the source track, at this distance itslevel will not be high enough to have actualhealth effects, even if it can still render acertain disturbance. “Silence” is a quality thatbecomes more and more difficult to find inour society. On the one hand, the possibilityfor people to temporarily escape urbanpressures and to relax in remote tranquilareas contributes to the general quality oflife. Tranquillity may also be an essentialprerequisite for nature conservation.

Figure 9.4 Intersection with tranquil zones per TEN variant

Intersection withAlternative tranquil zones (km2) %


Existing TEN network + 14 projects of specific interest 427 208 104


Existing TEN network + all planned water projects 413 277 101


Complete multi-modal TEN 512 556 125

Figure 9.5 Intersection with tranquil zonesaccording to TEN variants

A possible indicator for disturbance oftranquil zones is the intersection betweenthose zones and new TEN infrastructure(including a 7.5 km influence zone). Thiscan be estimated within a geographicalinformation system by overlaying the net-works with the tranquil areas. The surface ofthe tranquil area that falls within the 7.5kilometre buffer-zone around the infrastruc-ture is calculated relatively easily. In fact, bycalculating overlapping surfaces, this ap-proach goes a step further than a simplevicinity analysis that was used in Chapter 7.This can be justified for two reasons: firstly,the actual boundaries of the tranquil zonesare known, whereas the designated areaswere represented only by a circle around thecentre co-ordinates. Secondly, the numberand size of the tranquil zones (under thecurrent definition) is so much larger that itseems less appropriate to speak of individualtranquil zone, but rather consider it as aspecific land cover type.

9.3.2 The ResultsThe results can be presented in a table suchas the one in Figure 9.4, and by Figure 9.5,which summarise the overall increase in theintersection between tranquil zones andTEN infrastructure per variant, compared tothe intersection with the existing network.The effects of developing the rail projectsappear to be less than those of the develop-ment of the planned road network, whichtriggers both a partitioning and a deteriora-tion process. Both options seem to have acumulative effect on the amount of tranquilzones threatened.

Again, this indicator does not account forthe loss of smaller tranquil areas, as the dataavailable for an EU level assessment do notinclude such smaller-scale effects. Theseshould be addressed at regional or corridorlevel.

% in

crea

se

130

125

120

115

110

105

100

95Var. 1 Var. 2 Var. 4Var. 3 Var. 6Var. 5

Noise 35

Noise disturbance zones aroundthe Complete Multimodal TEN network Figure 9.6

A critique, which also applies to this indica-tor, is the fact that other infrastructures, notincluded in the TEN, were not considered.Possible improvement in the use of thisindicator could be achieved by also includ-ing non-TEN infrastructure in the definitioncriteria for tranquil areas. In this respect, aninvestigation should be made as to whichinfrastructure level should be taken intoaccount (i.e. should all roads included in theGISCO secondary road network be included,or only roads of a certain importance).

The results can also be shown on a map asthe one displayed on Figure 9.6, usefulmerely as an illustration of the indicatoranalysed.


10. Land Resources

10.1. Main Objectives Concerning Land Resources

In Europe, land resources are relativelyscarce, and reaching a sustainable balancebetween competing land uses is a key issuefor all development policies. Land useplanning and land resources are not dealtwith specifically as strategic themes in the5th Environmental Action Programme.However, their importance is fully refer-enced as a requisite for some strategic issuessuch as biodiversity and nature conservation,management of water resources and coastalzones, as well as for the sustainable develop-ment of several selected target sectors, suchas transport, agriculture and tourism.

Thus, identifying land take according to thevarious major habitat types of land cover canconstitute an indicator for a wide variety ofimpacts. For example, the calculation of landtake in semi-natural areas can be consideredan indicator for a potential decrease ofbiodiversity. Similarly, land take in urbanareas may indicate the risk of impacts onhumans, while the taking of agricultural orforestry land may indicate how the TEN mayaffect certain economic activities.

10.2. Land Take

The broadest way of assessing the spatial andecological impact of the TEN would be to

simply look at the land take of existing andnew infrastructure. It was not consideredappropriate to merely examine designatednature sites, populated areas, tranquil zones,water catchment areas, and so on separately.Instead, it was thought important to considerany land take as potentially indicatingadverse environmental impacts. For exam-ple, land take in semi-natural areas may leadto a decrease of biodiversity. Land take inurban areas could represent the risk ofimpacts on humans (safety and noise) andthe use of agricultural or forestry land couldbe considered harmful for nature in general,for landscape (visual intrusion) or even forcertain economic activities.

Naturally, looking at land take tout court isan ultimate simplification and a “worst-case”estimate, but given the flaws in the methodsand data (in particular) available for moresophisticated analyses, this crude indicatorhas its merits. One could go one step furtherand, as a minimum, take account of the landcover types, but the report will show that thismethod has the same limitations.

10.2.1. Methodological ApproachRoads, rail and waterways consume landdirectly or indirectly. Direct land take refersto the land needed for the actual infrastruc-ture and is a function of the number oftraffic lanes or tracks, the number of safetylanes, embankments, dug-outs, etc. Indirectland take is caused by land reallocation due

Source: EUROPEANCOMMISSION (1993),

The European HSRNetwork:

Environmental ImpactAssessment

Figure 10.1 Estimated surface requirements per infrastructure type

Traffic mode Total cross section (m) Surface (ha/km)

Rail Conventional 25 2.5

HSR upgrade 32 3.2

HSR new 35 3.5

Road 2x1 lanes 60 6

2x2 lanes 90 9

2x3 lanes 100 10

2x4 lanes 120 12

Land Resources 37

to the infrastructure construction, such asparking places, rail stations, etc.

In the present study, only direct land take hasbeen considered. The proposed indicator isdirect land take (derived from the lengths ofthe infrastructure development) differenti-ated by major land cover types that underliethe existing and planned links. Of course, thefact that the location of the routes is still notprecisely defined introduces a factor ofuncertainty into the estimates.

The surface required by transport infrastruc-ture is directly related to its constructioncharacteristics. Because these are increas-ingly set to international standards, thevariations between different countries anddata sources are marginal. Figure 10.1 illus-trates the surface requirements of differentinfrastructures. The figures are based on astudy carried out for DGVII of the EuropeanCommission in 1993.

To get an indication of the land use effi-ciency of each mode, the direct land takemust furthermore be put into the perspec-tive of the related traffic capacity that theinfrastructure offers. For example, takinginto account the capacities of a motorway(1500 person car units per hour per trafficlane) and of a HSR (15 trains per hour perdirection), the capacity of a HSR line equalsthat of a motorway with 2x4 lanes, butrequires much less land resources.15

Data on the number of traffic lanes for TENroads and the number of rail tracks per TEN-link are as yet not available. For the purposeof the present demonstration work, theamount of land take by linear infrastructureis estimated to take on average 3.2 ha/km forrail lines, 12 ha/km for inland waterways and10 ha/km for motorways. This corresponds

15) European Commis-sion (1993), The Euro-pean HST Network:Environmental ImpactAssessment

Figure 10.2Land take by TEN variant

Land takeAlternative Area (km2) Score

Existing TEN network 8,598 100

Existing TEN network + 14 specific projects 9,408 109

Existing TEN network + all planned rail projects 8,964 104

Existing TEN network + all planned inland waterway projects 8,767 102

Existing TEN network + all planned road projects 11,285 131

Complete TEN 12,632 137

to widths of 32 up to 120 meters. By combin-ing these factors with the estimated length ofthe existing and new network, total land takecan be obtained. Results are summed up inFigure 10.2, which presents total land take forthe existing network and the 5 networkvariants. The comparison of the figures isfacilitated by using index scores (the existingnetwork is 100). The smaller width of railinfrastructure explains the “favourable”score of the planned rail network.

It goes without saying that land take in itselfis not necessarily directly environmentallyharmful. The potential impact depends onthe type of land that is affected. A refine-ment of this indicator can be achieved takinginto account land cover types. Land covertypes were obtained from the EuropeanCORINE land cover inventory supplementedby national land cover inventories.

The land type or habitat information that wasused for this study is available at a resolutionof 250 metres, which seems compatible withthe infrastructure widths mentioned.

The CORINE database that was used for thisanalysis distinguishes seven major land coveror “habitat” types (in order of increasingenvironmental sensitivity):

• vegetated, strongly artificial land (whichincludes artificial non-agriculturalvegetated areas, arable land, permanentcrops and annual crops associated withpermanent crops);

• vegetated, less artificial land (which in-cludes pasture, complex cultivationpatterns, agriculture with natural vegeta-tion and agro-forestry areas);

• semi-natural land (which includes shruband/or herbaceous vegetation and openspace with little or no vegetation);


Figure 10.3 Distribution of the major land cover types in Europe

Land Resources 39

• artificial land (which includes urbanareas, industrial areas and mines);

• forests;• water bodies (which includes inland

waters, coastal lagoons and estuaries);• wetlands (which includes inland and

coastal wetlands).

If the different major habitat types areconsidered, it is clear that vegetated land willsuffer the most. However, data also shows theimpact of the planned road network onforests and semi-natural habitats.

10.3. Sensitive Zones Mapping

A second option for refining an indicatorbased on land use would be by definingsensitive zones on the basis of more criteriathan land cover type only. Various countriescommonly use “sensitive zone mapping” as atool in strategic environmental assessment.However, given the scale of the TEN and thedata limitations at EU scale, only a verysimplified version of these national orregional approaches can be applied.

Sensitive zone mapping derives from the so-called land-suitability analysis, an approachcommonly used in the USA since the early1970s. Its main characteristics are that: i) itfocuses principally on natural characteristicsof the land, that is, it can be spatially refer-enced and reported; ii) expert judgements(based on scientific knowledge) play acentral role in the method; and iii) impactsare implied but not explicitly reported.

Variants :1) existing TEN network;2) existing TEN network +14 projects of specificinterest;3) existing TEN network +all planned rail projects;4) existing TEN network +all planned inlandwaterway projects;5) existing TEN network +all planned road projects;6) complete multi-modalTEN. The sum of thecontributions for variant 1adds up to 100.

Based on a quantitative approach, thismethod uses categories or impact groupselection, ratings or assignment of prioritiesand can also yield a large index calculation.Map overlay techniques are the most suitableto perform this methodology, which was aprecursor for early GIS developments.

10.3.1. Methodological ApproachThe present identification of sensitive zonesis based on a methodological approachwhich, generally combines the use of GISand the use of indices, and which integratesexisting environmental and non-environ-mental information into general potentialimpact maps. The aim is to give an indica-tion of the sensitivity of an environmentalsystem (e.g. local community, ecosystems)according to the potential impacts of acertain activity (such as transport). Themethod consists of the following basic steps:

• The selection and mapping of areas thatare considered sensitive for specificrelated impacts, in this case sensitive totransport infrastructure. This is done onthe basis of an analysis of the geographic,demographic and environmental charac-teristics of the area and by aggregating alimited number of spatial/ecologicalindicators. This is carried out from theperspective of different “impact groups”;

• An appraisal of the expected sensitivityof the area using simple qualitativeindices such as “not sensitive”, “moder-ately sensitive” or “highly sensitive”.

amo

unt

of

land

140.0

vegetated-strongly artificialvegetated-less artificialsemi-naturalartificialforestswater bodieswetland

120.0

100.0

80.0

60.0

40.0

20.0

0.01 2 3 54 6

variant

Figure 10.4Land take by land cover type for TEN variants


An overlay of the infrastructure outline planswith the sensitivity maps enables the identifi-cation of areas where conflicts can arise. Theobjective of infrastructure planning shouldbe to avoid, where possible, crossing throughhighly sensitive zones, or, when this is una-voidable, to ensure that proper mitigationmeasures are taken.

The mapping of sensitive areas is doneaccording to “impact groups” or “sensitivitygroups” using existing environmental,physical and social baseline data.

For the present report, three differentsensitivity groups were distinguished:

• Group ISensitivity based on the general environ-mental physical nature of the area: water

Figure 10.5 Sensitivity thresholds attributed to baseline data sets

Sensitivity group Baseline data Low sensitivity Moderate sensitivity High sensitivity

I water patterns <10 km /cell 10-20 km/cell >20km/cell

semi-natural <25% SN area 25 -75% SN area >75% SN area

mountainous* class 1 dominant class 2 dominant class 3 dominant

II coastal zone** class 1 present class 2 present class 3 present

designated area <2% DG area 2-10% DG area >10% DG area

III population <50 inh/km2 50 - 250 inh/km2 >250 inh/km2

5 - 20 % BU area

built up area <5 % BU area >20% BU area

* class 1: lowland; class 2: hilly land;

class 3 : mountainous.** class 1: within 1 km

from the coast;class 2: between 1 and

5 km from the coast;class 3: between 5 and10 km from the coast.

Figure 10.6 Assignment rules for determining sensitivity per group

Combination of two risk data set (group II and III)

Original celle score Resultingsensitivity

high + high high

high + moderate high

moderate + moderate moderate

high + low moderate

moderate + low low

low + low low

Combination of three risk data set (group I)

Original celle score Resultingsensitivity

high + high + high high

high + high + moderate high

high + moderate + moderate high

high + moderate + low moderate

high + low + low moderate

moderate + moderate + moderate moderate

moderate + moderate + low moderate

moderate + low + low low

low + low + low low

patterns, occurrence of semi-naturalland, and mainly mountainous character-istics (possible environmental impactsinclude habitat fragmentation, loss ofbiodiversity, and visual intrusion oflandscapes);

• Group IISensitivity based on environmentallegal status (biodiversity, habitat frag-mentation, coastal zones);

• Group IIISensitivity based on territorial demo-graphic characteristics, such as popula-tion density or share of built up area(possible environmental impacts arenoise and accidents).

The present overall assessment scheme useswater patterns, semi-natural land, andmountainous areas (derived from DEM

Land Resources 41

Figure 10.7Sensitivity map based on Group I criteria (physical-environmental criteria)

Note: surface waterdata for Finland andnorthern Sweden ismissing

altitude and slope data-base) as GROUP Ibase-line data; coastal zone boundaries anddesignated areas as GROUP II base-line data;and population and built-up area as GROUPIII base-line data. GROUP II impact groupincludes coastal zones, as they constitute anirreplaceable environmental heritage withunique ecological, cultural and economicresources subject to increasing pressures. Asan example, approximately a third of theUnion’s wetlands are located on the coast, aswell as more than 30% of the SPAs.

Using expert judgement, simple scores forsensitivity indices are attributed to the areas.A three-class criteria or sensitivity thresholdwas attributed to the 7 baseline data sets. As aresult of this operation, each baseline dataset will provide a sensitivity distribution mapcontaining three impact classes. The thresh-olds are applied to square territorial units orcells measuring 10 km by 10 km. Dataanalysis was conducted on a raster format.

To combine the above indices into an impactgroup, assignment rules were used. Sensitiv-


ity maps were then built, based on these datasets, and according to defined assignmentrules. Semi-natural areas, mountainous areasand the density of surface water patternswere combined and overlaid to estimate theenvironmental/physical sensitivity of thearea. Secondly, legally defined nature sitesand coastal zones combined to form anenvironmental/legal sensitivity map. Lastly,built up areas derived from the major landcover types database and population density,gave rise to a territorial sensitivity map.

Figure 10.8 Sensitivity map based on Group II criteria (environmental legal status criteria)

Figure 10.5 illustrates the criteria or sensitivitythresholds for the 7 baseline data sets. As aresult of this operation, each baseline dataset will provide a sensitivity distribution mapcontaining three impact classes (low sensitiv-ity, moderate sensitivity and high sensitivity).The thresholds are applied to square territo-rial units or cells (“grids”) measuring 10 by10 kilometres. The thresholds can only beconsidered, in this exercise, as rules ofthumb. They are commonly, methodologi-cally based on expert judgement, which is

Land Resources 43

Figure 10.9Sensitivity map based on Group III criteria (demographic criteria)

generally derived from the application ofsome expert judgement assessment rule,which was in fact not the case.

The next step was to combine the aboveindices to identify a general measure ofsensitivity per group based on agreed assign-ment rules. (Figure 10.6)

Overlaying the multi-modal TEN networkwith the three sensitivity maps enables theidentification of high risk cells (“red flags”)

according to each impact group - that arecrossed by infrastructure links.

10.3.2. The ResultsApplying the described methodology, sensi-tivity maps were processed for each impactor sensitivity group. These maps are pre-sented in Figures 10.7, 10.8 and 10.9.

Network assessment was then performed byidentifying those links that cross the sensitiveareas, using spatial overlay techniques. Each


Figure 10.10 Overlay of the complete multimodal TEN network with sensitivity maps.Identification of high sensitivity cells crossed by infrastructure links.

Figure 10.11 Sensitivity index per impact group and TEN variant

Impact

Alternative Group I Group II Group III


Existing TEN network + 14 specific projects 111 112 111


Existing TEN network + all planned inland waterway projects 101 101 101


Land Resources 45

Figure 10.12Sensitivity index graphs per impact group and TEN variant

link is subsequently classified according tothe highest sensitivity category it crosses(Figure 10.10), thus forming an averagesensitivity index for each variant (Figure10.11 and Figure 10.12).

Figures 10.11 and 10.12 show that the specificprojects are primarily located in built up andpopulated areas and that they, compared tooverall planned road and rail networks, haveless impact on the more “natural” areas.

An investigation of Figures 10.7 to 10.9demonstrates the following:

1. As it is not always feasible to avoid bothareas of nature conservation interest andareas of dense population (as these areto a certain extent complementary),choices must be made with respect to theprioritisation of sensitive zones. However,the present method does not attemptthis “weighting”, as such decisions shouldbe taken on a broader policy level (e.g.in the context of existing national orregional policies and circumstances),and in consultation between decision-makers and stakeholders. The only aimof the present method is to clearly bring

forward the potential risks vis-à-vis allthree sensitivity categories.

2. The incomplete water pattern baselinedata set illustrates the importance ofhigh quality source material. Data gapsand data heterogeneity complicate therisk assessment process and are clearlydistinguished in the final results.

3. In this exercise, all baseline data wereconsidered equivalent. If some kind ofweighting were applied to accentuate orattenuate some vulnerability factors, fine-tuning could be within reach. However,irresponsible or biased use of the weightfactors could easily undermine the goalof this procedure, which is to obtain anobjective and general vulnerabilityassessment of the territory.

Finally, the present method could be signifi-cantly improved once estimates becomeavailable on the predicted magnitude oftraffic impacts on the links (i.e. trafficvolume, emissions of pollutants and noise).This new information could take into ac-count the expected environmental pressures,which can differ amongst modes andamongst links.

incr

ease 120

115110105100 95

Group I V

ar. 1

Var

. 2

Var

. 3

Var

. 4

Var

. 5

125

130

incr

ease 120

115110105100 95

Group II

Var

. 1

Var

. 2

Var

. 3

Var

. 4

Var

. 5

125

130

incr

ease 120

115110105100 95

Group III

Var

. 1

Var

. 2

Var

. 3

Var

. 4

Var

. 5

125

130


11. Evaluation and Conclusions

11.1. Indicator Overview

As stated before, one of the main aims of theworking group was to identify possibleindicators for explaining spatial and ecologi-cal impacts of the TEN, considering bothCommunity-level environmental objectivesand targets, as well as data availability. Thescope of the ecological and spatial impactssuggested the following impact categories:

• Ecological impacts of infrastructure, i.e.impacts that are directly related toecosystems, habitats and species, as wellas the degree of biodiversity;

• Functional impacts like those derivedfrom barriers, as a division of functionalland units and the reduction of theviability of land units;

• Impacts with a spatial dimension likenoise, which can be evaluated by estimat-ing for instance the number of peopleliving in the vicinity of the infrastructureor by defining noise sensitive zones.

The chosen indicators are based on the levelof planning (in this case the EU level), as wellas data availability, as selected indicators haveto be used Europe-wide. For more local-levelassessments (like corridor or project assess-ments), other, more detailed (and dataconstrained) indicators can be used, whichaccount for more specific characteristics ofthe immediate environment. In short, theinitial selection of spatial and ecologicalindicators was based on the following criteria:

• relevance in regard to the Communityenvironmental objectives and targets;

• applicability on EU scale;• feasibility in terms of the available data.

A group of seven indicators was chosen,namely:

• vicinity to sites of nature conservationinterest;

• partitioning of land;• water crossings;• population in surrounding of infrastruc-

ture;• crossing of tranquil zones;• land take by land cover type;• sensitive zones analysis

The use of these indicators showed that theyare quite different in concept, scope, com-prehensiveness, usefulness for decision-making and communication to a broaderpublic, as well as in the degree of uncertaintythat is linked to them, taking into accountthe quality of the currently available data.

The Working Group felt that some assess-ment procedure should be applied to thechosen and tested set of indicators, keepingin mind further development of the exercise,namely additional research and consultation,as well as impact evaluation on a strategiclevel of TEN alternatives.

In 1993, the OECD defined the followingissues as criteria for indicator selection:policy relevance and utility for users, analyti-cal soundness, and measurability (OECD,1993, OECD - Core Set of Indicators forEnvironmental Performance Reviews - Asynthesis report by the group on the state ofthe environment, Environment MonographsNo. 83, Paris).

As stated in the report, these criteria, whichwill not all be met in practice, describe an“ideal indicator” . Nevertheless, accordingto these criteria, the assessment of indica-tors can highlight their characteristics, andshow their usefulness for certain specificobjectives. The same report specified theissues within these criteria that should applyto each indicator. Concerning Policy Rel-evance and Utility for Users, each indicatorshould:

• provide a representative picture of environ-mental conditions, pressures on theenvironment or society’s responses;

• be simple, easy to interpret and able toshow trends over time;

• be responsive to changes in the environ-ment and to related human activities;

• provide a basis for international compari-sons;

• be either national in scope or applicable toregional environmental issues of nationalsignificance;

• have a threshold or reference value againstwhich to compare, so that users are ableto assess the significance of the valuesassociated with it.

Evaluation and Conclusions 47

Policy relevance and Analytical soundness Measurabilityutility for users

Vicinity to sites of Relates to defined Buffer-zones around Site boundaries lackingnature conservation environmental objectives network should dependinterest and targets on habitat/species type, Databases do not

for which information is cover all countriesIdentifies potential risks lackingas problematic links and Overlap of designatedsites (“red-flag” approach) Allows (and requires) areas

“sample surveys” atEasy to understand regional/local level Definition of sites

depend on nationallegislation

Characteristics andlocation of new linksmissing

Partitioning of land Relates to defined environ- Based on several assumptions, Ignores land units lessmental objectives and difficult to agree upon than 1000km2targets (quality evaluation, thresholds)

Infrastructure other thanEasy to understand Most applicable at regional level TEN not considered

(relatively detailed maps required)Difficult to interpret Characteristics and location

of new links missing

Crossing of surface Relates to defined environmental Not only water proximity may Characteristics and locationwaters (density) objectives and targets lead to environmental impacts of new links, information on

catchment areas andEasy to understand and interpret vulnerability of surface and

groundwater is missingToo simplistic

Allows for “red-flag” approach

People in vicinity of Not directly related to defined Gives only an indication of the Characteristics and location ofinfrastructure environmental objectives and potential population affected, new links, traffic forecasts and

targets but doesn’t apply to impact noise modelling missing.magnitude

Easy to understand Dwellings, working places,special infrastructure (hospitals,

Little added value, as it is an schools) not available“implied” indicator

Crossing of tranquil Not directly related to defined Characteristics and locationzones environmental objectives and of new links missing

targets

Easy to understandand interpret

More useful for issues otherthan noise (e.g. biodiversity)

Land take (by land Not directly related to defined Characteristics and locationcover type) environmental objectives and of new links missing

targets

Easy to understand

Little added value

Sensitive zones Not directly related to defined Data weighting not Characteristics and locationmapping environmental objectives and introduced of new links missing

targetsSuited to raise awareness

Easy to understand and at strategic levelto interpret

Requires wide expertBased on an extremely consultationsimplified model

Useful as a general indicator

Figure 11.1Summary of indicator characteristics


Concerning Analytical soundness, each indica-tor should :

• be theoretically well founded in technicaland scientific terms;

• be based on international standards andinternational consensus about its validity;

• lend itself to being linked to economicmodels, forecasting and informationsystems.

Concerning Measurability, each indicatorshould:

• be readily available or made available at areasonable cost/benefit ratio;

• be adequately documented and of knownquality;

• be updated at regular intervals in accord-ance with reliable procedures.

Figure 11.1 summarises the weaknesses andstrengths of the indicators identified in thepresent study and Figure 11.2 presents anevaluation of the indicators according to theOECD criteria. The assessment was con-ducted combining the results of an “expertspanel”, where each expert was asked to dothe ranking according to presented scores.The evaluation was conducted in conjunc-tion with the presented criteria, indicatorconcept, required environmental data andinformation display. Lack of data on infra-structure was also taken into account, as itseems more important for some indicators(population affected by noise) than forothers (e.g. crossing of surface waters).

Some conclusions can be derived fromassessing the indicators. In terms of their

Figure 11.2 Assessment of used indicators according to OECD criteria

Policy relevance and Analytical soundness Measurabilityutility for users

Vicinity to sites of nature + +/- +conservation interest

Partitioning of land + + -

Crossing of surface waters +/- +/- +

Population in surrounding + +/- +/-of infrastructure

Crossing of tranquil zones ++ +/- +/-

Land take (by land-cover type) +/- + +

Sensitive zones mapping ++ - -

++ very good+ good+/- fair- poor

-- very poor

policy relevance and utility for users, they areall considered as good or very good, exceptthe crossing of surface waters and land takeby land cover type, which is mainly due tothe “plain” characteristics of these directindicators.

Concerning analytical soundness, it generallyscores under the pointed relevance or utility,which means that further research on theirtheoretical validity as well as technical andscientific consensus building must be con-ducted. Although this is a broad conclusionover the general scope of the indicators,considering its importance for the SEA ofthe TEN, it points to further consultationbetween Commission services, MemberStates, the EEA and the scientific communityon the most feasible and appropriate meth-ods and indicators.

Some problems should be raised concerningthe measurability criteria, which relates todata availability, quality and updating. Directindicators such as the crossing of surfacewaters and land take, as well as the vicinity tosites of nature conservation were consideredas the most reliable, considering the datadriven issues. The importance of accurateand up-to-date data on land cover must bestressed, as it serves as baseline data for mostof the indicators used.

Considering biodiversity indicators, vicinity tosites of nature conservation interest seemsmore meaningful and appropriate to becarried out at the corridor level. A necessaryindicator refinement would be the improve-ment of the designated sites database and aqualitative analysis establishing the relation


between sites/habitats/species and respec-tive impact vulnerability, performed byspecific “zoom analysis”. This work would bebest linked, at a first stage, to the corridoranalysis already under way, and throughconsultation with the Member States. Specialattention must be given to potential risk sitesidentified by this exercise (“red-flag” analy-sis), but also to other specific situationswhich, because of an inconsistency in defini-tion, or information available (the case ofnationally designated sites) have not beenconsidered.

Concerning the partitioning of land indica-tors, in spite of being easy to understand as aconcept, it is difficult to interpret. This ispartly because it is based on many assump-tions, which are difficult to agree upon,particularly at a European level. It is bestused at a regional level (where smaller unitscould be considered in the analysis), and itshould be refined to consider infrastructureother than existing TEN. The impact evalua-tion also requires information on the charac-teristics and location of new links.

The Water resources indicator is simplistic andeasy to understand, but its usefulness is quitelimited. That is partly because environmentalimpacts are not just a result of proximitybetween water surfaces and infrastructures .An analysis based on water catchment areas,relating hydrology and quality of surface andgroundwater, lacked accurate data at anEuropean level. Again, at a strategic level,this indicator can be useful for scoping morein-depth analyses at a regional or corridorassessment.

The indicators used for noise assessment arean example of the need for more detailedinfrastructure data, as their relation to noiseis indirect and inferred. They do not con-sider noise levels and their spatial distribu-tion (critical noise contours), which can onlybe calculated based on traffic estimates andphysical infrastructure characteristics (pave-ments). They are both relatively easy tounderstand, but their added value for noiseimpact evaluation is very small. The first oneis more a “social-indicator” as it considers thepotential number of affected people (bynoise but also by other effects such as “bar-rier effect”). As a “social indicator,” it isimportant to note that detailed informationon location of special equipment and infra-structure (hospitals, schools, and residentialareas) is missing. The second one seems,also, to be more useful for issues other thannoise, e.g. for biodiversity and social impacts.

Land resources indicators, like land take, isstraightforward, easy to understand, anduseful as a general indicator for raisingawareness at the strategic level. Sensitivezone mapping seems to be very useful fordecision-makers at policy and strategic level.The specific characteristics of the method,based on data combination through mapoverlay, make the quality of data a significantissue. Again, the method could be signifi-cantly improved once estimates becomeavailable on the predicted magnitude oftraffic impacts on the links (i.e. trafficvolume, emissions of pollutants and noise).This would allow better accounting for theexpected environmental pressures, whichcan differ amongst modes and amongstlinks.

11.2. Conclusions

The importance of improving certaindatabases has been clearly raised in theprevious chapters. In fact data shortage andinconsistencies, as well as all the assumptionsused to overcome them, limit the usefulnessof the chosen indicators. Their measurabilityand their soundness, in particular, cannot beassured.

One of the main problems encountered inthe study is the poor information on theplanned TEN-infrastructure. Special atten-tion should be paid to collecting additionaldata on the distinction between upgrade andnew constructions for the planned TEN linksand to the improvement of information onthe location of the planned new infrastruc-ture. Also estimates on the predicted magni-tude of traffic impacts on the links (i.e.traffic volume, emissions of pollutants andnoise) would enable a better understandingof the expected environmental pressures,which can differ amongst modes andamongst links.

Another priority action is the completion ofthe information on the NATURA 2000network, as this would provide a moreconsistent basis for the assessment of impactson habitats. Furthermore, this underlinesthe importance of developing a SEA systemfor the TEN in an ‘iterative’ manner, sincethis would allow the refining of the assess-ment techniques and scope as more accuratedata become available.

The models and parameters that were usedin this study have been applied in a ratherstraightforward manner, and using expert


judgement which inevitably involves acertain degree of subjectivity. Parametersand threshold values (such as the criteriaand indices used for identifying sensitivezones) have been entered in a pragmaticway, based on literatureexamples gatheredfrom various national, regional or localenvironmental assessments. The key assump-tions that have been made and their transfer-ability to EU scale have to be improved bybroad and multi-disciplinary consultation ofexperts. In particular the EU-wide definitionof sensitive zones, which inevitably relies to acertain extent on a subjective attribution ofsensitivity values, should be further re-searched.

In this study, it was also assumed that theimpact a transport infrastructure exerts whencrossing a sensitive environment is similarfor all types of infrastructure (e.g. motor-ways, rail or inland waterways). This may notnecessarily be the case. Also, it was notpossible to differentiate between varioustypes of motorways (e.g. number of lanes).In the future, attributing weights to varioustypes of infrastructure could increase thevalue of the assessment. This could be doneon the basis of, for instance, infrastructurecharacteristics (such as the number of lanesor tracks per link). Another solution wouldbe to integrate the results of the predictionof traffic-related impacts in the GIS (i.e.traffic volume, emissions of pollutants andnoise for each link). This would enable abetter comprehension of the various degreesof environmental pressures, and thus allow abetter weighting of the impacts according tomode and to links.

In order to ensure the compatibility of theoutputs of both parts of the assessment, theRTD project that the Commission is cur-rently launching on the assessment of globalimpacts should therefore be closely co-ordinated with the current work on spatialand ecological assessment.

The study also made clear that an assessmentat EU level is naturally limited in scope andlevel of detail, due to data limitations. Themethods that were tested allow the identifica-tion of potential conflicts with certainenvironmental concerns (“red-flag” ap-proach), but do not permit a prediction ofthe magnitude of the actual impacts (interms of losses in species or damage tohuman health, for example). This wouldrequire extremely detailed data on speciesand ecosystems and a qualitative investiga-tion into the relationship between habitat

types, species and their vulnerability toimpacts of infrastructure and traffic. At thisscale, it is not feasible to evaluate the im-pacts on small valuable areas. It is thereforeessential that the SEA of the TEN be comple-mented by more local-level assessments, i.e.assessments of national/regional networksand of corridors, and, possibly, pilot projectenvironmental impact assessments (EIA).Using such a “tiered” assessment process, theSEA of the TEN should be considered as afirst broadbrush screening instrument foridentifying those corridors or series ofcorridors that should be addressed as apriority.

During the April 1997 workshop on theSpatial and Ecological Assessment of theTEN, some issues were identified, whichhave not yet been addressed in the presentstudy, as they require in-depth research,including:

• development of indicators and methodsfor the spatial assessment of secondarydevelopment around TEN-nodes;

• identification of ecological corridors andthe use of such corridors in infrastruc-ture network and corridor assessments.

Finally, in order for the SEA to become anintegrated decision-aiding tool within theoverall revision of the guidelines, it shouldbe conducted in parallel with the evaluationof the socio-economic impacts of the TENand of its impact on cohesion and accessibil-ity. The studies that are carried out on allthose aspects should be carefully co-ordinated, and should be based on the sameunderlying assumptions and data sets.

The present methodological study can beconsidered successful as it realised most of itsoriginal objectives:

• it provided a comprehensive compilationof a Union-wide spatially referenceddatabase, holding information on thestate of the TEN, as well as of a selectionof environmental and other data, whichis relevant for spatial and ecologicalimpact assessment;

• the project initiated the development,testing and application of a series ofimpact assessment methodologies suchas proximity analysis, impact predictionmodels and sensitivity mapping. Thesepreliminary results will form a practicalbasis for refinements and further meth-odological developments;

• the provisional results are stored in a


new database that can be made availableto all parties concerned in order tolaunch the discussions and raise aware-ness on the environmental impact of theTEN implementation;

• the importance of this SEA of the TENexercise in the screening for corridorassessment must also be highlighted.

The study furthermore demonstrated thetechnical feasibility, but also the limitations,of a Europe-wide SEA of the TEN, and re-affirms the usefulness of the multidiscipli-nary data collection at European scale.

The present exercise also pointed to issuesneeding further research and methodologi-cal improvements:

1. Optimisation of the indicators andmethods through a wide and multi-disciplinary consultation of experts,stakeholders and Member States. In thisrespect, the following actions can beenvisaged:• organisation of a technical workshop(involving the EEA, the Commissionservices, external experts and NGOs),which should allow consensus buildingon the methods and indicators.• presentation of the methods to theMember States in the next meeting ofthe Ad Hoc Group on the Environmen-tal Evaluation of the TEN.

2. Filling of the major data gaps on theTEN and on the environment.

3. The carrying out of the full spatial andecological assessment of the TEN,including the integration into the GIS ofthe results of the traffic-related impacts(emissions, noise, safety) and a compre-hensive multi-criteria analysis of thepredicted impacts. For this, sufficientresources (in terms of budget and staff)have to be made available.

4. Initiation of longer term research on theabove issues, in order to graduallyimprove methods, data and tools for theecological and spatial impact assessmentof transport plans.

5. Perform some “zoom analyses” or “sam-ple surveys”, enabling the testing of theinterrelation between different plan-ning/geographical levels and correspondent assessment techniques (rela-tion between strategic assessment andcorridor and project assessment). Forthis purpose a follow-up and feedback ofthe 5 corridor assessments currentlybeing carried out at national level withsupport from the Commission, should beperformed.

6. The selection of the alternatives, shouldbe a subject for further research, due toits strategic nature and assessmentimplications.


List of Acronyms

CBD - United Nations Convention on Biological DiversityCCFE - Commission des Chemins de Fer EuropéensCoE - Council of EuropeCORINE - Coordination of Information on the EnvironmentDEM - Digital Elevation ModelDGVII - European Comission Directorate-General for TransportDGIX - European Comission Directorate-General for Environment, Nuclear Safety and CivilProtection5EAP - 5th Environmental Action ProgrammeEEA - European Environment AgencyETC/LC - EEA European Topic Centre on Land CoverETC/NC - EEA European Topic Centre on Nature ConservationGIS. - Geographical Information SystemGISCO - EUROSTAT Directorate on Social and Regional Statistics and Geographical Infor-mation SystemsHSR - High Speed RailNATURA 2000 Network - Process related to the Habitats and Birds Directives, on the conser-vation of european natural resources for the year 2000 and beyondNGO - Non Governmental OrganizationNUTS - Nomenclature of Territorial Units for StatisticsOECD - Organisation for Economic Co-operation and DevelopmentRSPB - The Royal Society for the Protection of BirdsSEA - Strategic Environmental AssessmentSPA - Special Protected Areas under directive 79/409/EECTEN - Trans-European Transport NetworkWERD - Western European Road Directorates

Documents

Demonstration of Indicators and GIS Methods