Geomorphology 66 (2005) 95 – 117 www.elsevier.com/locate/geomorph
An approach for quantifying geomorphological impacts for EIA of transportation infrastructures: a case study in northern Spain Jaime Bonacheaa,*, Viola Maria Bruschia, Juan Remondoa, Alberto Gonza´lez-Dı´eza, Luis Salasa, Jurjen Bertensa, Antonio Cendreroa, Ce´sar Oteroa, Cecilia Giustib, Andrea Fabbric, Jose´ Ramo´n Gonza´lez-Lastrad, Jose´ Marı´a Aramburue a
b
Universidad de Cantabria, Santander, Spain Universita` degli Studi di Modena e Reggio Emilia, Modena, Italy c ITC, Enschede, The Netherlands d Tecnologı´a de la Naturaleza S.L. (Tecna), Madrid, Spain e Diputacio´n Foral de Guipu´zcoa, San Sebastia´n, Spain
Received 30 January 2002; received in revised form 4 December 2003; accepted 2 September 2004 Available online 26 November 2004
Abstract A methodological proposal for the assessment of impacts due to linear infrastructures such as motorways, railways, etc. is presented. The approach proposed includes a series of specific issues to be addressed for each geomorphological feature analysed—both dstaticT and ddynamicT—as well as a series of steps to be followed in the process. Geomorphic characteristics potentially affected were initially identified on the basis of a conceptual activities/impacts model that helps to single out geomorphic impacts related to environmental concerns for the area. The following issues were addressed for each individual impact: nature of potential effects; indicators that can be used to measure impacts; criteria of dgeomorphologic performanceT; procedure for measurement/prediction of changes; translation of geomorphologic impacts into significant terms from the viewpoint of human concerns; possible mitigation and/or compensation measures. The procedure has been applied to a case study corresponding to a new motorway in the Basque Country, northern Spain. Geomorphological impacts considered in this analysis included: (1) consumable resources; (2) sites of geomorphological interest; (3) land units with high potential for use, high productivity or value for conservation; (4) visual landscape; (5) slope instability processes. The procedure has been designed for implementation in a Geographic Information System (GIS) environment. Details are given on the application of the method to each individual impact analysed and results are presented in both numerical and map form. Impacts assessed were initially expressed by means of heterogeneous magnitudes, depending on the geomorphological feature considered. Those geomorphological impacts were then translated into significant terms and homogeneous magnitudes. Integration was carried out on the basis of impact values thus obtained. Final integrated results were also expressed in numerical and map form.
* Corresponding author. Tel.: +34 942 201512; fax: +34 942 201402. E-mail address:
[email protected] (J. Bonachea). 0169-555X/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.geomorph.2004.09.008
96
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
The method proposed enables comparison of alternatives as well as dpredictionT and assessment of impacts in terms directly related to geomorphic characteristics. It also facilitates the expression of those impacts in terms that allow integration with other types of environmental impacts. D 2004 Elsevier B.V. All rights reserved. Keywords: Environmental Impact Assessment (EIA); Geomorphologic assets; Geomorphologic resources; Geomorphic impact; GIS; Northern Spain
1. Introduction Construction of motorways and other linear structures produces a variety of environmental impacts, many of which are directly or indirectly determined by geomorphologic characteristics such as relief, landforms, surface materials, active processes, etc. (Gonza´lez Alonso, 1989; Cavallin et al., 1994). These impacts can be particularly relevant in mountain areas, where geomorphic assets are often valuable and geomorphic processes more active. Environmental Impact Assessment (EIA) is a decision support tool that helps to incorporate scientific analyses and assessments into the decisionmaking process (Wathern, 1990; Ramos, 1993; Go´mez Orea, 1999). The incorporation of geomorphologic impacts into the general EIA process must address two main issues. On the one hand, measurement and prediction of impacts on geomorphic features; this is obviously a task for specialists. On the other hand, assessment and evaluation must be presented in terms meaningful for non-specialists, and in such a way that comparisons with other impacts as well as with environmental concerns existing in the region are possible (Beinat et al., 1999). General conceptual approaches for identifying and measuring impacts on geomorphologic features have been proposed (Panizza et al., 1996; Boyer et al., 1998; Marchetti and Rivas, 2001) but, to our knowledge, no detailed methodological schemes for measurement, evaluation and integration of those impacts within a general EIA framework have been developed. In particular, there is a need to find means to translate geomorphic impacts into terms that are significant from the point of view of society in general. The application of transparent procedures and criteria to transform geomorphologic assessments into, for instance, monetary terms would facilitate their incorporation into the EIA process. Development
of Geographic Information System (GIS) procedures for the implementation of those schemes would also be useful for their incorporation into EIA practices (Patrono et al., 2001; Barbieri et al., 2002). The scheme proposed in this work includes the following steps: general description of the project; identification of goals and concerns; establishment of methodological approach based on a general conceptual model (relationships and interactions between the proposed transportation infrastructure and different environmental components and processes); identification of the main interactions (relationships with concerns); identification of potential impacts on geomorphological components and processes, irrespective of their relative importance (relationships with concerns); proposal of possible impact indicators; establishment of dgeomorphologic performanceT criteria; proposal of specific procedures for the analysis and evaluation of each geomorphologic impact.
2. Study area and project description The study area where the method proposed has been applied is part of the Deva valley, northern Spain (Fig. 1); it has approximately 190 km2 and 80,000 inhabitants and an economy based on manufacturing industries and services (EUSTAT, 2000). Altitude ranges between 110 m and 700 m a.s.l.; the climate is oceanic, mild and humid (average temperature and rainfall are 14 8C and 1400 mm). Episodes with over 50 mm/day occur several times per year. Flood events have a return period of 5–10 years (DFG, 1997). The area (EVE, 1991; Tame´s et al., 1991) is moderately folded and faulted, with WNW–ESE as the main structural trend (Fig. 1). Main bedrock materials are Upper Cretaceous flysch facies (both calcareous and terrigenous–calcareous); some basaltic pillow lavas appear within the sedimentary sequence.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
97
Fig. 1. Location and simplified geological map of study area. 1: marly limestones and marls; 2: sandstones, clays, limestones and marls; 3: volcanic rocks; 4: quaternary deposits; F: faults; N: anticline; S: syncline; A and B: motorway alternatives.
Geomorphology is determined by V-shaped valleys with limited development of alluvial plains; slopes have gradients around 20–25% and are covered by 0.5–3 m of regolith. Soil and regolith erodibility is high (when vegetation cover disappears); contacts between regolith and bedrock often act as rupture surfaces for shallow landslides. Land cover is intensely transformed by human activity. Natural vegetation barely covers 16%, and only parts of it are autochthonous forests. Areas reforested with pine trees amount to about 60%. Builtup areas concentrate along the narrow alluvial plain and adjacent slopes. Prominent environmental problems are related to intense transformation of the natural environment and presence of manufacturing activities. Well-preserved natural areas are extremely scarce. River courses are highly degraded by both works along the margins and pollution. Population and activities are concentrated in a narrow strip along the valley, where the best soils are normally present. These areas are subject to floods that produce important damages. Shallow landslides occur throughout the area, partly favoured by land cover and land-use changes. The project considered is a motorway linking the towns of Vitoria–Gasteiz and Eibar. The goal of this
project is obviously improving traffic capacity and safety. Environmental concerns related to geomorphic characteristics are: (a) visual impact of the new infrastructure, in rural areas with landscapes of a fairly high scenic quality; (b) land occupation or degradation in a valley where land with high potential for use is scarce; (c) damage to areas of high natural value or productivity, in a zone where the extent of well-preserved natural ecosystems is limited; (d) damage to the quality or integrity of geomorphologic sites; (e) interference with natural processes and hazards, which could result in increased risks for human life or property. Two alternative routes have been analysed: A, 15 km long of which 6.5 km correspond to tunnels, and B, 17 km long and consisting mainly of cuts and viaducts.
3. Methodology 3.1. Conceptual model Relationships between specific actions and significant environmental qualities, through chains of direct and indirect effects, for both the construction and operational phase, were identified and expressed by
98
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
means of a conceptual model (Fig. 2). Several dlevelsT in the chain of impacts are considered: (1) specific actions during construction and operation; (2) environmental components or elements affected by actions; (3) processes and environmental elements involved in secondary chains of effects; (4) consequences for significant qualities of the environment; (5) human concerns, that is, aspects related to productive sectors or human well-being that could be affected. Relevant geomorphologic interactions related to concerns have been highlighted. No further analysis of non-geomorphologic impacts will be undertaken here. 3.2. Potential geomorphologic impacts Impacts identified may affect geomorphic resources and assets as well as geomorphic processes. The former are essentially dstaticT from the viewpoint of EIA; they do not respond to human actions. Processes and hazards, on the other hand, are ddynamicT; human modifications induce new changes in the process (area affected, intensity, rate, etc.) and
may trigger a chain reaction and a cascade of additional impacts (Brunsden, 1996). Geomorphologic resources and assets (and concerns related to them) include: consumable resources (resource base, productive sectors); sites of geomorphologic interest (heritage), high productivity or high natural value units (resource base, heritage, productive sectors); land units with high potential for use (resource base, productive sectors); visual landscape (heritage, environmental perception). Processes and hazards concern human and material safety; they include terrain instability (landslides, collapse, subsidence) and infiltration/runoff related processes (groundwater recharge and pollution, soil erosion, stream flow, channel erosion, sediment load, water quality, silting, water logging, flooding). Indicators proposed as a basis for assessing impacts on the geomorphic features considered here are: consumable resources: volume of material affected; sites of geomorphological interest: number of sites affected and extent (degree) of damage; units with high potential for use: area lost; high value or
Fig. 2. Conceptual diagram for identification of potential impacts (only part of the activities corresponding to the construction phase are represented). Levels shown in circles. 1: activities; 2: direct effects on environmental components; 3: secondary effects (processes bold; human actions and influences in italics; elements underlined); 4: significant qualities (impacts analysed are shaded); 5: relationships with concerns.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
productivity units: area or productivity loss; visual landscape: magnitude (area, number of people) and intensity (degree of modification of visual quality) of visual intrusion; slope instability: absolute or relative change in landslide probability. Indicators related to water-driven processes are not considered here, because sufficient data is not available at this stage.
99
Departamento de Medio Ambiente, Departamento de Obras Pu´blicas) for the Basque Government. Several of those maps were originally made on paper format using standard field and air photo methods. Original digital maps were transformed into vector format, imported into ILWIS V.2.2 and then converted into raster format using a grid cell (bpixelQ) of 1010 m.
3.3. Performance and significance Indicators listed above can be determined objectively (with some limitations; see Section 4.4) and provide a measure of the degree of change of geomorphic conditions. The magnitude of those changes can be used as a basis to assess dgeomorphic performanceT or departure from existing or natural conditions. That is, if only a small proportion of a certain geomorphological unit is affected, or if the change in the rate of a process is marginal, impacts can be considered as acceptable and within the range of natural system’s variations. However, impacts expressed in geomorphologic terms (essentially objective) are normally not meaningful for non-specialists and do not allow integration with other impacts; therefore, they should be translated into significant terms to allow evaluation and comparison (more subjective). 3.4. Database for impact measurement and prediction Procedures for parameter measurement, geomorphic impact prediction and evaluation are presented below. The procedures proposed were implemented through a digital database and use of Geographic Information System (GIS) tools. The database used is composed of 13 thematic layers: topography (Digital Elevation Model, DEM), population centres, road network, motorway alternatives, plan for the prevention of floods, geology, geomorphology, geotechnical conditions, soils and soil capability, surface deposits, regolith thickness, land cover, sites of geomorphological interest. Each layer corresponds to a map; the last one was derived from an existing inventory (Tame´s et al., 1991). Original base maps were prepared by different companies and organisations (Ente Vasco de la Energı´a; Compan˜´ıa General de Sondeos; Tecnologı´a de la Naturaleza; Instituto Geogra´fico Nacional;
4. Implementation and results 4.1. Consumable resources Construction on top of deposits (sand, gravel, clay, peat) or introduction of constraints for their exploitation implies a total or partial loss of such materials. If these resources are scarce alternative, distant sources might be necessary. As sand and gravel are widely used materials, cost increases would affect all future construction activities and impact could be high. Impact measurement was carried out using dmotorway alternativesT, dsurface depositsT dregolith thicknessT and dpopulation centresT maps. Nonexploitable deposits were eliminated by overlay of deposits and population centres maps. Volume affected in each remaining deposit was calculated by overlay of motorway routes (a 100-m strip, including motorway and buffer was considered; tunnel sectors were excluded) on dsurface depositsT and dregolith thicknessd maps. Percentage of exploitable material in deposits (classified according to origin) was calculated using Casagrande’s classification (Casagrande, 1948). Local market prices were used to transform volume into monetary value. Each individual deposit and resource type are thus characterised in terms of area, volume and monetary value (Fig. 3, Table 1). Performance from the geomorphic point of view can be expressed using the ratio: lost resource/reserves; the lower its value the greater sustainability (in the sense that use of the resource could be more extended in time). Relevance of impacts can be assessed on the basis of the ratio: lost resource/consumption rate (equivalent to years supply), or else monetary value of material. Results expressed in terms of absolute geomorphic impact (m3 deposit affected) and significant impact (potential monetary loss) are shown in Table
100
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
Fig. 3. Exploitable resource deposits, SGI and motorway routes. a: sorted gravel; b: poorly sorted gravel; c: silty sand; d: clayey sand. A and B: motorway alternatives; t: tunnel sectors. 1–15 SGI (see Table 2).
1. Lost resource/reserves ratio is 0.381 for alternative A and 0.427 for alternative B. It is clear that A represents a more sustainable option, although the difference is small. Annual consumption of this type of materials in the area is about 2 m3/person/year. The amount sterilised in the case of alternative A is equivalent to 2.49 years consumption in the area, and 2.79 years for B. Mitigation/compensation measures include removal of deposits prior to construction or improvement of accessibility to other deposits in the area (thus
reducing exploitation costs and increasing their exploitation potential). 4.2. Sites of geomorphologic interest Sites of geomorphologic interest (SGI) are an important part of the natural and cultural heritage, valuable for educational, cultural, recreational or tourism use (Bertacchini et al., 1999). These sites can suffer total or partial destruction by road construction. Indirect damage to the site’s value can
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
101
Table 1 Area, volume and monetary value of exploitable deposits and comparison between alternatives Type
Area (m2)
Volume (m3)
% Price Value (o) Exploitable (o/m3)
Alternative A Area
Sorted gravel (a) 1,594,600 5,028,550 Poorly sorted 295,400 590,800 gravel (b) Silty sand (c) 743,200 1,344,950 Clayey sand (d) 2,246,900 3,516,600 Total 4,880,100 10,480,900
Alternative B
Volume Value 837,829 0
Volume Value
3.3 3.3
15,764,504 1,852,158
88 88
3.9 5.3
4,615,868 66,100 132,200 453,710 69,300 131,650 451,823 16,401,422 0 0 0 18,800 28,200 131,525 38,633,952 133,100 399,450 1,291,539 161,600 447,550 1,485,288
also be produced by constructions next to it, reducing access or visibility. Indicators for assessing impacts on SGI refer to the number and overall value of sites affected. Evaluation of a site’s quality or value is a process that necessarily implies some subjectivity. The procedure described here is an improvement of a former proposal; a more detailed description can be found in Rivas et al. (1997). The method uses clearly stated criteria based on observable characteristics, but nevertheless is less objective than the ones proposed for other features considered. Geomorphic performance can be expressed in relation to the ability of SGI within the area to continue providing their services (register of scientific information, vehicle for training and education, support for recreational activities, trigger of tourism activities). If those services are performed satisfactorily by the ensemble of sites in the area, the situation can be considered as sustainable. The proportion of damage to those functions (or sites that can perform them) can be used as measure of performance. For instance, if sites damaged represent 1% of those that can be used for education and the remaining ones have enough capacity to absorb users well above any projected increase, sustainability would not be significantly affected. On the other hand if one out of only two peat deposits in the area were destroyed, a significant part of that type of scientific record would be lost. Relevance of damage to SGI has been established by comparison with well-known and socially appreciated geomorphologic landmarks that are considered to represent a standard of doptimum conditionsT. Losses in SGI value can be expressed as a proportion of the total loss of that optimum. Data layers used for the assessment include: map of SGI (and associated database) for the area of
67,000 267,250 0 0
Area
95 95
73,500 287,700 0 0
901,940 0
reference, information on roads and buildings, population data. Value of each SGI was defined as: VSGIi ¼ Ci ð2Qi þ Pi Þ=48 where C i =degree of preservation (0–4); Q i =intrinsic quality (0–4); P i =potential for use (0–4); V SGIi =value of site (0–1). Calculation of index terms was made using and improving the procedure described by Rivas et al. (1997). Changes introduced in the method are small and will not be discussed here. Table 2 and Fig. 3 present a list of SGI in the study area (Tame´s et al., 1991). Values of the different factors considered are shown; those values were used to calculate Q and P (Rivas et al., 1997) and V SGIi . Total value of SGI in the study area before motorway construction is the sum of individual values for the n sites considered (8.27). To assess impacts three types of areas of influence of the motorway have been defined, corresponding to 50, 100 and 150 m from motorway axis. Damage to SGI within the central band is total; decreasing percentages of damage have been considered for successive bands (50–100 m=50%; 100–150 m=25%; N150 m=0). Table 2 shows the values of SGI corresponding to the two motorway alternatives considered (V SGI (post)). Total impact on SGIs is defined as the difference between value of sites after (post) and before (pre) project construction; those impacts are 0.35 for alternative A and 0.87 for alternative B. If total value of SGI in the area is accepted as a measure of their potential to provide the services mentioned above, reduction of that value provides a basis for assessing geomorphic performance (degree of change suffered by these geomorphologic assets).
102
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
Table 2 SGI in the study area Site no.
Name, SGI
A
E
K
Ex D
Ac O
S
H
Acc C
Q
P
V SGI
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Total value
Megaturbidites, San Lorenzo Cretaceous Flysch, Aristiburu Danian marls, San Lorenzo Pillow lavas, Argate Volcanic rocks, Arzabaleta Volcanic bodies, Igarate Trachyte flows, Malzaga Quarry, Malzaga Fold, Urko Folds, Eibar Pillow breccias, La Ascensio´n Tectonic breccia, Ugarriaga Cretaceous section, Arane Pillow breccias, Placencia Outcrop, Iturbe
0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.8 0.8 0.6 0.8
0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.2 0.2 0.2 0.2 0.2
0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4
0.0 0.0 0.0 0.2 0.0 0.2 0.0 0.4 0.2 0.2 0.2 0.2 0.2 0.2 0.0
0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8
0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
0.8 0.8 0.8 0.2 0.0 0.0 0.8 0.8 0.4 0.4 0.4 0.4 0.4 0.4 0.4
2.0 2.0 2.0 2.2 2.2 2.0 2.4 2.0 1.8 2.4 2.2 2.0 2.0 2.2 1.8
3.0 3.0 3.0 1.8 1.4 2.0 3.0 3.4 2.4 2.4 2.4 2.4 2.4 2.4 2.2
0.58 0.58 0.58 0.52 0.48 0.50 0.65 0.61 0.50 0.60 0.56 0.53 0.53 0.57 0.48 8.27
0.0 0.0 0.0 0.2 0.2 0.0 0.4 0.0 0.0 0.6 0.6 0.2 0.0 0.6 0.0
0.8 0.8 0.8 0.0 0.0 0.4 0.8 0.8 0.4 0.4 0.4 0.4 0.4 0.4 0.4
4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0
(pre)
V SGI
(post A)
0.58 0.58 0.58 0.52 0.48 0.50 0.30 0.65 0.50 0.60 0.50 0.53 0.53 0.57 0.48 7.90
V SGI
(post B)
0.58 0.58 0.58 0.00 0.48 0.50 0.30 0.65 0.50 0.60 0.50 0.53 0.53 0.57 0.48 7.38
A: relative abundance; E: extent; K: degree of knowledge; Ex: good example of processes; D: diversity; Ac: activities; O: observation conditions; S: availability of services; H: number of inhabitants; Acc: accessibility; C: state of conservation; Q: quality of SGI; P: potential for use; V SGI: value of SGI. In bold, values that change.
Reduction is 4.23% and 10.51% for alternatives A and B, respectively; the latter appears as quite significant. To assess relevance, comparison with the theoretical destruction of a well-known and socially appreciated geomorphologic landmark in the region has been made. The SGI chosen for comparison is the bRato´ n de GuetariaQ (Guetaria Mouse; Fig. 4), although outside the study perimeter, it probably is the most emblematic coastal landform in the region. The value obtained for that SGI is 0.86. The dabsoluteT loss of V SGI of alternatives A and B is 0.35 and 0.87, respectively. The latter would be
equivalent to the total loss of the bRato´n de GuetariaQ, whereas the former would represent a loss slightly greater than one third of such SGI. Values obtained indicate that alternative B implies a significant reduction of the ability of the environment to fulfil the kind of services provided by SGI. Mitigation measures in the case of important SGI imply the need to avoid destruction by route selection and/or design details. If SGIs affected are not of exceptional interest, compensation measures are a lower-cost, reasonable alternative. Road works can be used to uncover or enhance certain SGI. Improve-
Fig. 4. The dRato´n de GuetariaT (Guetaria Mouse), prominent geomorphological landmark in the region.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
ments on other SGI can be made to improve state of conservation or visibility (sites 4, 6 and 14), as well as accessibility (sites 5, 12 and 13) (Table 2). The criterion should be that the dtotal sumT of SGI value (intrinsic quality and potential for use) in the area be maintained or increased. 4.3. High value land units These non-consumable land resources include units with high potential for use (UHPU), high value for conservation (UHVC) or high potential productivity (UHPP). The first have attributes that make them attractive for residential, commercial, industrial or other uses; UHPP have microclimate and soils favourable for farming; UHVC have well-preserved natural land cover, valuable species or high biodiversity. The value of the latter is due mainly to nongeomorphologic features, although such features are often a necessary condition for their existence (for instance, green oak woods on karstic massifs). All such units represent scarce assets in the study area; if they are physically occupied or close to the motorway they will be destroyed or seriously degraded. The main impact indicator used is total area of each type of unit occupied and/or degraded by the motorway. Performance can be expressed in terms of percentage of the unit type affected. Comparison with the rates of land dconsumptionT for different uses, market value of land units consumed or degraded, or potential productivity loss are measures of relevance. Maps used for identifying and mapping UHPU are: slope gradient (derived from DEM using 1010 m pixel), geotechnical conditions (reclassified in four categories: favourable, acceptable, unfavourable, very unfavourable), flood hazard (return periods represented: 10, 100 and 1000 years), population centres (includes all consolidated nuclei), road network (includes all practicable roads). Criteria for defining UHPU were: gradient b20%, no geotechnical limitations (favourable and acceptable), return period N1000 years, b1000 m from population centres, b200 m from roads. High potential units are those meeting all criteria established (Fig. 5); the rest were considered as having low potential. Criteria were applied using a stepwise procedure, eliminating on each map units not meeting the criterion.
103
Impact was determined by overlay of motorway alternatives over the UHPU map obtained. A road width of 100 m within which total loss is assumed was used. A buffer of 50 m on either side was also considered; within this buffer a correction factor of 0.5 was applied. Tunnel sectors were excluded from the analysis. Geomorphic impact is expressed as area of UHPU lost. Performance of each alternative is assessed using the ratio: UHPU affected/total UHPU area. Relevance or translation into significant terms was assessed comparing area lost with the rate of UHPU consumption for other activities in the area; also computing the value of UHPU lost. The reference value used corresponds to a 2000m2 plot, flat and hazard free, less than 100 m away from a road and 500 m from the town of Vergara (160 o/m2). Table 3 presents the results obtained for the alternatives analysed. Performance ratios are 0.027 and 0.018 for alternatives A and B, respectively. Clearly, alternative B is better from this point of view. Rates of dland consumptionT for urban, infrastructure and industrial activities in northern Spain (Rivas, personal communication) have been estimated between 5 and 10 m2/person/year. This represents 400,000–800,000 m2/year in the study area, mostly in UHPU. Taking the lower figure as a reference, the effect of the motorway would be equivalent to about one third of normal UHPU use rates in the Deva valley. Attributes of some UHPU can be reproduced; if destruction of certain UHPU cannot be avoided project design should include dconstructionT of similar units near the new road. In the case considered, compensation could be achieved constructing some 140,000 and 90,000 m2 of UHPU for alternatives A and B, respectively. UHVC and UHPP were treated together. The former were derived directly from the land cover map. Alnus copses and Quercus woods on acid
Table 3 Impact on UHPU (total UHPU area: 5.4106 m2) UHPU
Alternative A 2
(m )
Alternative B
% Total (m2)
Area affected 97,600 1.81 Area in buffer 42,300 0.78 Monetary value (o) 19,000,000
% Total
64,900 1.20 21,800 0.40 12,128,000
104
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
Fig. 5. Units of high potential for use (UHPU); high potential productivity (UHPP) and high value for conservation (UHVC); A and B: motorway alternatives; t: tunnel sectors.
terrain represent the two types of valuable natural units within the study perimeter. Managed land units of high potential productivity (UHPP) were defined on the basis of the following maps and criteria: soil capability (reclassified into five categories; A: very high, B: high, C: moderate, D: low, E: very low; the first two are of high potential); altitude, (obtained from DEM; areas below 300 m have been considered); solar radiation, derived from DEM and parameters related to solar position during the day and throughout the year (Pickering,
1990; Felicı´simo, 1994; Ilwis, 1997; U.S. Naval Observatory, 2001). Altitude and solar radiation provide information on microclimatic conditions. Solar radiation was computed for the 21st of each month for year 2000, for each hour between sunrise and sunset. Average of those values has been used to characterise each pixel. The model obtained shows relative radiation in 255 classes. The upper 30% (N179) was considered as high potential. Criteria were established according to practices in the area.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
The final map obtained by overlay has therefore three classes: natural land units with high value for conservation, high productivity managed land units and the rest. The latter includes both units with high potential for use and others (Fig. 5). If a given unit was both UHPU and UHVC or UHPP, the former was not considered. Impact was measured on the basis of area of valuable or productive units affected by the alternatives considered (Table 4). Performance of alternatives was assessed on the basis of percentage of unit type destroyed; it is worst for managed productive units and for alternative B. Relevance was expressed in terms of potential biomass loss. Data obtained for most common crops in the area (Gobierno Vasco, 1999) are as follows: cultivated prairies, 30,000 kg/ha/year; cereals, 4200 kg/ha/year; potatoes, 17,000 kg/ha/year. No determinations of productivity of forest units analysed here are available for the study area, but 2.5 m3/ha/year is a reasonable value for oak woods according to local timber companies. Crop productivity can be expressed in monetary terms, as it refers to products with a market value; that is not the case of natural productivity. For comparison purposes only, it has been assumed that the deconomic valueT of natural biomass is that of wood production. This is obviously a simplification and does not take into account the value of other services provided by the unit. Productivity of managed units (both biomass and monetary value) has been taken as the average for the above-mentioned crops. Potential annual losses obtained using those assumptions are about 3400 o/ ha/year for high productivity managed units and about 1900 o/ha/year for natural valuable units (minimum value). Those figures have been used to calculate impacts in significant terms (Table 4). Results obtained show that alternative A has a lower impact. Partial mitigation of impacts on productive managed units can be achieved by removing topsoil from
105
units to be destroyed, replacing it at the end of the operation or placing it elsewhere. Soil conservation and regeneration actions in other parts of the area are possible compensation measures. Mitigation or compensation for damage to UHVC requires long periods. A suitable compensation measure can be restoration of degraded ecosystems in the area. The criterion should be to maintain or increase the total area of such systems or their natural productivity. 4.4. Visual landscape Visual quality (intrinsic merit of a unit from a perceptual point of view) of landscape is determined mainly by three groups of factors: geomorphology (relief, landforms, rock type), land cover, and land use (especially presence of impacting elements such as prominent constructions). The introduction of new large structures such as a motorway represents a visual intrusion that affects visual quality. Effects on visual quality or visual impact are related, on the one hand, to the degree of visibility—also dependent on relief and landforms—of the new structure and, on the other hand, contrast between the structure and preexisting landscape. Several indicators have been used for characterising visual impacts. Visibility area (VA): area from which at least part of the structure can be seen; viewing population (VP): number of persons living within visibility area; magnitude of visual effect (Mve=VAVP); quality correction factor (Qcf): dimensionless coefficient directly related to the quality of visual unit in which the new structure is located. Visual impact (VI)=MveQcf. Performance in this case is a concept that does not have a clear meaning from the geomorphologic point of view. Possible criteria to apply are size ratio between the new landform created and the size of natural, linear landforms in the area, or else percent
Table 4 Impact on UHVC (total area: 6.6106 m2) and UHPP (total area: 6.1106 m2) Units
Valuable natural (UHVC) Productive managed (UHPP) Total
Alternative A
Alternative B
Area (103 m2)
% unit
Productivity (kg/year)
Pot. loss (o/year)
Area (103 m2)
% unit
Productivity (kg/year)
Pot. loss (o/year)
81.6 158.8
1.22 2.59
20,400 714,600
2,285 107,190 109,475
143.0 236.0
2.14 3.86
35,750 1,062,000
4,004 159,300 163,304
106
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
Table 5A Categories of visual quality for land use/cover and landform types Visual quality Land use-cover type (Luc) Urban areas Industrial Cultivated prairies and crops with hedges Terraced crops and cultivated prairies Shrubs Rock Afforestations Deciduous forests Crops/afforestations mosaic Crops/forest mosaic Forests/afforestations mosaic Mixed forests/shrubs mosaic
1 1 2 2 4 4 3 5 3 4 4 5
Landform type (Lf) Alluvial plain Valley floor and adjacent slope Valley slopes and gentle interfluves Abrupt summits and crests Network of small valleys and interfluves
1 2 3 4 5
(1=low; 5=high).
of the study area from which the new structure will be seen. The lower those values the more acceptable will be the degree of modification of natural landforms from the point of view of preserving visual heritage. Relevance of impacts can be established by comparison with prominent structures in the area; also estimating the effect of degraded views on property values in the units affected. This helps to determine the social importance of the impact. The operating procedure was based on the use of the following data layers: DEM; geomorphologic map; land cover-use map; population centres and buildings. Population data was also used. First landform and land use/cover units were combined to define visual units. Visual quality rank for land-use/ cover and landform types was assigned on the basis of a survey of expert opinion. Eleven experts from different fields and well acquainted with the region were interviewed using the DELPHI method (Balkey, 1968; Delbecq et al., 1975; Dı´az de Tera´n et al., 1992). Visual categories thus obtained for the different unit types are shown in Table 5A. Visual quality thus defined is obviously not an dobjective measurementT but provides a clearly defined classification of landform units from the scenic point of view that
reproduces quite well the opinions of local residents. Table 5B shows the combination to obtain visual quality of integrated units. The quality of units for each motorway sector or point can thus be easily identified (Fig. 6). Visibility and viewing population for a series of points taken at regular intervals (200 m) was then determined by means of GIS (ArcView) and Mve calculated. A linear value function was used to define the relationship between visual quality of units and quality correction factor (category 5=1; category 1=0). That is, the magnitude of impacts will decrease with decreasing quality of units affected. Table 6 shows results obtained (sums for all pixels affected) for the two motorway alternatives. Visibility, viewing population, magnitude of visual effect and visual impact are clearly greater for alternative B. Visual dperformanceT and relevance, of the new structure were assessed, respectively, by comparison with natural linear landforms (crests, valleys) and prominent artificial structures used as reference standards. The Deva valley itself and a pronounced NW–SE crest at the northeast corner of the study area, the two largest natural linear landforms in the study area, were used as reference standards for performance. Comparison was made on the basis of the average vertical section (lengthaverage heightdepth). Average vertical sections obtained are as follows: Deva valley, 15,20020 m (304,000 m2); crest, 490040 m (152,756 m2). Average vertical sections obtained for alternatives A and B are 842510 m (84,250 m2) and 12,62836 m (454,608 m2), respectively. The size ratios (Rs) obtained were: Rs(A)Valley=AVSA/AVSV= 0.28; Rs(A) Crest =AVS A /AVS C=0.55; Rs(B) Valley= AVSB/AVSV=1.5; Rs(B)Crest=AVSB/AVSC=3. That is, the dartificial landformT represented by alternative Table 5B Visual quality categories obtained by integration of landform (Lf) and land use/cover (Luc) Luc Lf
1 2 3 4 5
1
2
3
4
5
1 1 2 2 3
1 2 2 3 3
2 2 3 3 4
2 3 3 4 4
3 3 4 5 5
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
107
Fig. 6. Visual quality of integrated units and motorway alternatives. 1–4: visual quality class; P: population nuclei; A and B: motorway routes; t: tunnel sectors; V: visual reference points.
B represents a significant modification compared to natural linear landforms in the area, especially if crests (more comparable to the shape of a motorway) are considered.
Artificial structures used as reference standards for relevance were two large buildings, one in Eibar and another in Vergara (Fig. 6). VA, VP and Mve of both structures are quite similar (8.34 km2, 1898 persons;
Table 6 Visibility and visual impact of motorway alternatives 2
Visibility area (VA) (km ) Viewing population (VP) (no. of inhabitants) Magnitude of visual effect (Mve) (km2 inhabitants) Visual impact (VI) (km2 inhabitants)
Alternative A
Alternative B
34.8 15,581 542,218 108,443
54.9 20,427 1,103,058 303,092
108
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
15,829 km2 pers. and 10.4 km2, 1419 persons; 14,758 km2 pers.). The average was taken as the standard. Magnitude ratios are RMveA=35.4 and RMveB=72.12 if comparison is made with the whole motorway. A more appropriate comparison, however, is with the average visual effect of individual motorway sectors (200 m). In this case, the values obtained are RMveA=0.82 and RMveB=0.97. Magnitude of visual effect of motorway sectors is therefore slightly below that of existing large buildings, somewhat lower in the case of alternative A. If the cumulative effect of all sectors is considered the effect is obviously much greater.
Mitigation of visual impacts can be achieved by standard procedures such as creation of visual barriers (earth, vegetation) or use of materials that will reduce visibility or contrast. Restoration of visually degraded units within the area can be used for compensation. 4.5. Slope Instability Processes Road construction can produce important impacts on slope instability, and therefore on hazards and risks. Interaction between a geomorphological process and a project can be both dpassiveT and dactiveT
Fig. 7. Landslide hazard/impact map. Numbers in scale indicate increase/decrease in landslide probability after road construction.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
109
Fig. 7 (continued).
(Cavallin et al., 1994). This results in modifications of either hazard level (potential destructiveness of the natural process), element exposure (presence of persons/structures) or vulnerability (degree of expected damage to human elements) or the three of them. Possible impact indicators include number of potentially unstable pixels affected by the new road, or increase in hazard and risk levels. Performance can be assessed on the basis of changes in probability of landslide occurrence in the future. Relevance can be assessed comparing expected future damages in preand post-project situations.
The procedure used has been based on the assessment of landslide susceptibility in the study area. Susceptibility was assessed using Spatial Data Analysis (SDA) techniques based on the application of several Favourability Functions (FF; Chung and Fabbri, 1993, 1999, 2001). Different susceptibility models were elaborated and validated using independent data sets. Temporal validation (comparison between predicted landslide distribution and actual landslide occurrence in periods after the ones used for model construction) was the basis for transforming susceptibility (relative spatial probability) maps into hazard maps, representing probability of new ruptures
110
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
in each pixel for a 30 year period. A detailed description of the method has been presented elsewhere (Remondo et al., 2003(a,b)). Results showed that the most satisfactory susceptibility/hazard models were obtained combining DEM-derived variables, land cover and lithology (bedrock and surface deposits). A Digital Elevation Model (DEM) incorporating modifications produced by motorway alternatives was elaborated. Susceptibility and hazard models for post-project situations were thus obtained and compared with pre-project ones. The difference between post- and pre-motor-
way models was calculated for each alternative and impact maps obtained (Fig. 7). The dsumT of probability differences in all pixels affected provides a numerical expression of such impact and also of geomorphic performance. The values obtained were 51.8 and 176.6 for alternatives A and B, respectively; that is, there would be a decrease in the overall probability of new landslides. Impact in this case is therefore positive, although only slightly. This is coherent with the actual behaviour of predominant slope movements in the area (shallow translational slides affecting the regolith).
Fig. 8. Significant visual impact (o/pixel). Pixels in population nuclei have been reclassified into 10 classes. Light background shading: total visibility area. Blank: area from which the motorway is not visible.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
111
Fig. 8 (continued).
Relevance was established by estimating the economic consequences (risk) of changes determined above; the procedure is described below.
5. Integration Impacts on individual geomorphic characteristics have been described and dmeasuredT by means of heterogeneous units that express degree of change with respect to pre-project situation. Changes have
been expressed in terms that reflect differences with respect to existing geomorphic conditions. Comparison or integration with other impacts is difficult in those terms because units used are different for each impact type. Moreover, their meaning and relevance from the human point of view is not clear in general. Translation into significant terms has therefore been attempted. This has a double objective: on the one hand, it makes understanding and interpretation of results easier; on the other hand, if homogeneous units are used, integration is much simpler.
112
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
An effort has been made to translate geomorphic impacts assessed into significant terms through the use of monetary units. In some cases, those units reflect actual market values, but in others, certain criteria and assumptions have been applied to express in monetary terms characteristics not subject to market exchanges. The latter, of course, are more questionable. If all impacts are expressed in monetary terms, integration is simple and immediate because impacts are additive. The overlay of individual impact maps (monetary units) results in an dintegrated, significant geomorphic impactT map including all characteristics considered here. It is obvious that integration with impacts on other environmental features can easily be performed if they can also be expressed in monetary terms. It must be stressed, however, that monetary values reflected in these assessments should not be taken as predictions of future losses. They must be understood as a reflection of potential or theoretical average costs with a relative rather than absolute meaning. Procedures and criteria used for the transformation of each dgeomorphologic impactT into dsignificant impactT (o value) are described below.
o12106 some 10 years ago. On the basis of those figures, an baverage valueQ of o9106 has been considered for a theoretical site with V SGI=1.00. The value of the reference site used, the bRato´n de GuetariaQ (V SGI=0.86) would be o7.79106. In other words, it is estimated that society would be willing to spend that sort of amount to protect the integrity of the reference site. The figure above was presented to several experts, well acquainted with the study area and the region, and they considered it as a reasonable value. Translation of impacts on SGI has been made multiplying total V SGI reduction by the standard defined. The dlossT would be o3.15 10 6 (I SGI =0.35) for alternative A and o7.83106 (I SGI =0.87) for alternative B (Table 2). Map representation is very simple in the case study analysed, as SGIs affected do not cover more than one pixel. Impact on each site would thus correspond to the pixel containing it. In the case of larger SGI, the total dlossT experienced by each SGI could be evenly distributed among all pixels affected, although this is not a totally satisfactory solution, because actual damage is not likely to be evenly distributed within a large site.
5.1. Geomorphologic resources
5.3. Land units with high potential for use (UHPU)
Resources considered have a well-known market value; translation of cubic meters (m3) into euros (o) is thus immediate. Total impacts on geomorphologic resources are equivalent to o1,291,539 for alternative A and o1,485,288 for alternative B (Table 1).
Market value of this type of land is about 160 o/ m2 for the study area. Significant impacts are o19,000,000 and o12,128,000 for alternatives A and B, respectively (Table 3).
5.2. Sites of geomorphologic interest
5.4. Units of high value for conservation or high productivity
This is the geomorphologic feature that presents the greatest difficulties for translation into significant terms. Assessment of SGI value implies certain subjectivity; also, both SGI value and impact are expressed by means of dimensionless indices. The criterion applied to establish the dmonetary valueT of impacts on SGI has been comparison with known restoration or protection actions on other sites in the region, one in the coastal area of Vizcaya and another in Cantabria. In the former area, with V SGI 0.5 about o3106 were devoted to restoration 4–5 years ago; the expense in the latter, with V SGI 0.9, was about
Potential productivity loss has been expressed in terms of kg biomass/m2/year. That biomass (crops in the case of managed units, wood in natural units) also has a market value that has the meaning of a dminimum environmental lossT as it does not include the value of other environmental services, particularly relevant in the case of natural units. dPotential productivity reductionQ has been estimated as 109,475 o/year and 163,304 o/year for alternatives A and B, respectively (Table 4). To make comparison and integration with other impacts possible, this annual figure has been transformed
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
into a total value, considering a 30-year period (a reasonable life-time for any infrastructure). Impacts would thus be o3,284,250 and o4,899,120 for A and B, respectively. As potential loss is calculated per unit area, map representation is simple and direct; impact value for each pixel is simply the sum of individual impacts affecting that pixel. 5.5. Visual landscape This type of impact also presents difficulties for translation into monetary terms. The criterion used was to consider that landscape degradation has an influence on the value of houses within the visibility area of the new infrastructure. It is a well-known fact that housing prices are influenced, among other things, by the quality of views. The difference between the prices of two similar houses, one with good and the other with poor views can be over 25% in the region. In this analysis, a conservative criterion has been used: house value will experience a 0.5% reduction for each motorway sector that can be seen from it. This means that a 10% dlossT will be experienced by houses from which most of the motorway is visible. Average value per house (including apartments), obtained from a sample of market prices, has been taken as o120,000. This represents a o600 reduction for a house viewing one point. Number of houses per pixel (h) for the different population nuclei was determined from data in building and population census (EUSTAT, 2000). Number of motorway points viewed from each pixel was determined by overlay of visibility areas of each point. The product Qcfh o600 corresponding to each overlay (motorway point) was calculated
113
for each pixel and total dlossT for all pixels obtained (Fig. 8). This provides an expression of visual impact in monetary terms. Total significant impacts were o2,816,218 and o4,415,707 for alternatives A and B. Total impact is higher in the case of alternative B because it affects a larger area; however, if maximum impact on individual pixels is considered, alternative A is worse. A further refinement can be included to take into account the effect of distance. Decrease of visual impact with the square of distance between object (motorway) and observer (pixel/house) can be incorporated by means of a simple function. This has not been done for the present analysis. 5.6. Slope instability Significance of impacts on slope instability processes was made through its expression in risk terms (potential damage increase/decrease in each pixel). This is directly used for integration. Average damage of past landslide events in the area, according to existing data, has been approximately 600 o/event. As probability change in each pixel is known, risk variation for all pixels can be calculated. Risk decrease derived from results above would be equivalent to o31,128 and o106,161 for alternatives A and B, respectively. It is clear that impacts on this process have a limited significance; this is coherent with the nature of slope movements in the area, mainly small, surface landslides that normally produce limited damages. It must be stressed that the values obtained are useful for comparing impacts on slope instability processes with other impacts, but should not be interpreted as a prediction of actual monetary losses, rather as
Table 7 Total impacts expressed in geomorphologic and significant terms Geomorphologic impact Resources (m3) SGIs (maximum value 8.27) UHPU (m2) UHVC/UHPP (m2) Visual (km2 inhabitants) Landslide hazard (AD probability) Total
Significant impact (o)
Alternative A
Alternative B
Alternative A
Alternative B
399,450 0.35 139,900 240,400 108,443 51.8
447,550 0.35+0.52 86,700 379,000 303,092 176.6
1,291,539 3,150,000 19,000,000 3,284,250 2,816,218 31,128 29,510,879
1,485,288 7,830,000 12,128,000 4,899,120 4,415,707 106,161 30,651,954
114
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
dpotential loss increaseT. The reliability of the approach proposed depends, obviously, on the amount and quality of data available. 5.7. Total geomorphic impact All impacts are now represented in terms of o/ pixel; integration can be carried out by simple addition. Table 7 and Fig. 9 show the results obtained for both alternatives. Impact values obtained have the meaning of dpotential lossesT and are presented as negative values. In the case of slope instability hazard, an improvement is expected and the figure is therefore
presented as positive (it must be pointed out that originals of Fig. 9, as well as Figs. 7 and 8, are in colour and therefore easier to visualise). As shown in the table, alternative A has a lower total impact, although not significantly. Impacts on individual factors are also lower except on units of high potential for use. This is logical, route A is closer to population centres and affects more valuable land. On the other hand, as it is shorter and has several tunnel sectors, its impact on other features considered is clearly less. It is interesting to make comparisons among the different significant impacts. In some cases, monetary values obtained are related to actual market values or
Fig. 9. Integrated impact map in significant terms.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
115
Fig. 9 (continued).
costs (resources, UPHU, productivity of managed units, landslide damage). In other cases, however, certain assumptions have been made and criteria applied to express in monetary terms the social value of assets not directly subject to market exchanges; this is particularly the case for SGI and visual landscape. Criteria used are based on actual expressions of the willingness to pay for such intangible values by local society. Several examples of actual expenses for valorisation of geomorphologic or other natural sites in the region show that the values obtained are not
unrealistic. In the case of visual impact, the figures used are comparable with other indicators of the effect of views on housing prices in the region (for instance, increase in the value of flats with height above ground). In short, significant impacts presented here do not have the meaning of precise measures of actual costs, but can be considered as reasonable expressions of the relative importance that society attaches to the geomorphic and related factors analysed. The value obtained for overall impact is approximately equivalent to the cost of 5 km of motorway in the region.
116
J. Bonachea et al. / Geomorphology 66 (2005) 95–117
From the results obtained, it appears that the main potential impact, by far, is the one due to the occupation of land valuable for other types of uses. Another important impact is the one on productivity of valuable managed and natural units. Both values are related to actual costs. Impacts on geomorphologic assets with a cultural, recreational or aesthetic value also appear as quite important. Finally, impacts on exploitable resources and slope movements are relatively minor in the area, particularly the latter.
6. Final comments The procedure described here facilitates the incorporation of different geomorphologic factors into the EIA process. It provides a basis for expressing very diverse impacts on a common basis as well as assessing their relative importance from the societal point of view. The method proposed has the following advantages: Individual geomorphologic impacts are expressed using specific magnitudes; changes introduced by the proposed action can thus be measured objectively. An exception to this are sites of geomorphologic interest; impacts in this case are represented by changes in a dimensionless, relative index. Nevertheless, this index is calculated applying clearly defined and transparent criteria and procedures, thus making results replicable. Individual geomorphologic impacts are also represented in map form; impact assessment and alternative comparison can thus be made numerically or visually. Criteria and procedures are also proposed for translating geomorphologic impacts into terms that have greater significance for non-specialists. That translation includes the transformation of initially heterogeneous impact magnitudes into homogeneous units that can be easily understood (monetary value), thus facilitating incorporation into decision making processes and communication with the general public. Transformation of geomorphologic impacts into dpotential monetary lossesT presents difficulties and criteria are disputable. This transformation makes integration simple and straightforward, as impacts thus expressed are additive. Integrated geomorphologic impacts can be represented as well in significant terms in both numerical and map form.
The whole procedure can be implemented using GIS tools, facilitating its incorporation into normal practice. Similar methods could be used to incorporate other types of impacts into the analysis. If all impacts assessed were transformed into monetary units they could be directly integrated using the procedure proposed here. The final result would obviously be different. However, the method must be improved and expanded to incorporate other geomorphic processes not considered here. Assessment of individual geomorphologic impacts, in particular those related to the prediction of changes in dynamic characteristics (processes) must be refined. Criteria for translating individual impacts into monetary values should also be improved and tested under different conditions, to make sure that such values reflect social perception appropriately. Alternatives to monetary units should also be sought for expressing impacts in significant terms and providing other means to summarise overall impact.
Acknowledgements This work was carried out as part of project GETS (Contract ERBFMR CT97-0162, TMR Programme, European Commission). A. Cendrero had a fellowship from the DGICYT, Spain (Prog. de Saba´ticos) and V.M. Bruschi a grant from the Fundacio´n Marcelino Botı´n. DEM, air photos and other data were kindly provided by the Diputacio´n Foral de Guipu´zcoa, Departamento de Obras Pu´ blicas. Review and criticism by Dr. Reynard and one anonymous referee is gratefully acknowledged.
References Balkey, N.C., 1968. The Delphi method: an experimental study of group opinion. Rand Memorandum, vol. 5888. P.R. Rand, Santa Monica, CA. Barbieri, M., Bonachea, J., Corsini, A., Giusti, C., 2002. A GISbased model for geomorphological impact assessment of a railway track (province of Modena, Italy). Geografia Fisica e Dinamica Quaternaria 25, 11 – 22. Beinat, E., Bressan, M., Jones, M., Fabbri, K., 1999. Geographical information systems and environmental impact assessment. UNIGIS, Amsterdam.
J. Bonachea et al. / Geomorphology 66 (2005) 95–117 Bertacchini, M., Giusti, C., Marchetti, M., Panizza, M., Pellegrini, M., 1999. I Beni Geologici della Provincia di Modena. Artioli, Modena, Italy. Boyer, L., Fierz, S., Monbaron, M., 1998. Geomorphological heritage evaluation in karstic terrains: a methodological approach based on multicriteria analysis. Supplementi di Geografia Fisica e Dinamica Quaternaria 3, 103 – 113. Brunsden, D., 1996. EIA and geomorphologic processes. ITC Journal 4, 339. Casagrande, A., 1948. Classification and identification of soils. Transactions of the American Society of Civil Engineers 113, 901 – 991. Cavallin, A., Marchetti, M., Panizza, M., Soldati, M., 1994. The role of geomorphology in environmental impact assessment. Geomorphology 9, 143 – 153. Chung, C.J., Fabbri, A., 1993. The representation of geoscience information for data integration. Non-Renewable Resources 2, 122 – 139. Chung, C.J., Fabbri, A., 1999. Probabilistic prediction models for landslide hazard mapping. Photogrammetric Engineering and Remote Sensing 65, 1389 – 1399. Chung, C.J., Fabbri, A., 2001. Prediction models for landslide hazard zonation using a fuzzy set approach. In: Marchetti, M., Rivas, V. (Eds.), Geomorphology and Environmental Impact Assessment. Balkema, Lisse, The Netherlands, pp. 31 – 47. Delbecq, A.L., Van den Ven, A., Gustafson, D.H., 1975. Group Techniques for Program Planning: A Guide to Nominal Groups and Delphi Processes. Scott Foreman, New York. DFG. (Diputacio´n Foral de Guipu´zcoa), 1997. Los Rı´os de Guipu´zcoa. Estaciones de Aforo y de Calidad de las Aguas. D.F.G., San Sebastia´n. Dı´az de Tera´n, J.R., France´s, E., Duque, A., Cendrero, A., 1992. Indicators for the assessment of environmental impacts from small industrial installations; application to an asphalt agglomerate plant in northern Spain. In: Hermelin, M. (Ed.), Environmental Geology and Applied Geomorphology in Colombia. AGID-Univ. EAFIT, Medellı´n, pp. 121 – 134. EUSTAT, 2000. Informacio´n estadı´stica por temas. Instituto Vasco de Estadı´stica, Vitoria http://www.eustat.es/. EVE, 1991. Mapa Geolo´gico del Paı´s Vasco. E: 1/25,000: Hojas 63-III (Eibar) and 88-I (Vergara). Ente Vasco de la Energı´a, Bilbao. Felicı´simo, A.M., 1994. Modelos Digitales del Terreno. Introduccio´n y Aplicaciones en las Ciencias Ambientales. Pentalfa, Oviedo. 220 pp. Gobierno Vasco {Basque Government}, 1999. Anuario Estadı´stico del Sector Agroalimentario 1997–1998: CAPV. Departamento de Agricultura y Pesca, Vitoria.
117
Go´mez Orea, D., 1999. Evaluacio´n del Impacto Ambiental. Mundi Prensa, Madrid. Gonza´lez Alonso, S., 1989. Guı´a Metodolo´gica Para la Elaboracio´n de Estudios de Impacto Ambiental. 1. Carreteras y Ferrocarriles. Ministerio de Obras Pu´blicas y Urbanismo, Madrid. Ilwis, 1997. The Integrated Land and Water Information System, v 2.2. ITC, The Netherlands, http://www.itc.nl/ilwis. Marchetti, M., Rivas, V. (Eds.), 2001. Geomorphology and Environmental Impact Assessment. Balkema, Lisse, The Netherlands. Panizza, M., Fabbri, A., Marchetti, M., Patrono, A., 1996. Geomorphologic analysis and evaluation in environmental impact assessment. ITC Publication 32, 372. Patrono, A., de Francesch, F., Marchetti, M., Moltrer, A., 2001. GIS applications for environmental impact assessment. An analysis of the effects of a motorway project in an Alpine valley. In: Marchetti, M., Rivas, V. (Eds.), Geomorphology and Environmental Impact Assessment. Balkema, Lisse, The Netherlands, pp. 207 – 221. Pickering, R.P., 1990. Digital image analysis of SPOT multispectral data for topographic mapping. MSc thesis, ITC, Enschede, The Netherlands. Ramos, A. (Ed.), 1993. Guı´a para la Elaboracio´n de Estudios del Medio Fı´sico. Ministerio de Obras Pu´blicas y Transportes, Madrid. Remondo, J., Gonza´lez, A., Dı´az de Tera´n, J.R., Cendrero, A., 2003a. Landslide susceptibility models utilising spatial data analysis techniques; a case study in the lower deba valley, Guipu´zcoa (Spain). Natural Hazards 30, 267–279. Remondo, J., Gonza´lez, A., Dı´az de Tera´n, J.R., Cendrero, A., Fabbri, A., Chung, C.J.F., 2003b. Validation of landslide susceptibility maps; examples and applications from a case study in Northern Spain. Natural Hazards 30, 437–449. Rivas, V., Rix, K., France´s, E., Cendrero, A., Brunsden, D., 1997. Geomorphological indicators for environmental impact assessment; consumable and non-consumable geomorphological resources. Geomorphology 18, 169 – 182. Tame´s, P., Mendiola, I., Pe´rez, C., Ortega, I., Salazar, A., de Alba, S., Gallardo, J., Portero, G., Henar, M., Olive´, A., Dı´az de Tera´n, J.R., France´s, E., Cendrero, A., 1991. Geomorfologı´a y Edafologı´a de Guipu´zcoa. Diputacio´n Foral de Guipu´zcoa, San Sebastia´n. U.S. Naval Observatory, 2001. Software for Sun or Moon Altitude. Astronomical Applications Department, http://aa.usno.navy.mil/ data/docs/AltAz.html. Wathern, P. (Ed.), 1990. Environmental Impact Assessment; Theory and Practice. Routledge, London.