Natural Hazards 20: 279–294, 1999. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.

279

GIS Techniques for Mapping Groundwater Contamination Risk ? DANIELA DUCCI Dipartimento di Ingegneria Geotecnica – Sez. di Geologia Applicata, Università di Napoli “Federico II” Piazzale Tecchio, 80 80125 Napoli, Italy, e-mail: [email protected] (Received: 1 February 1998) Abstract. The groundwater contamination risk map of a sample alluvial area was produced by using the Ilwis Geographical Information System (GIS) to construct and to overlay thematic maps. The risk map has been derived from the vulnerability map, the hazard map, where the potential contaminating sources were identified, and the socio-economic value of the groundwater resource, represented by the wells. The groundwater quality map allowed the reliability of hazard and risk maps to be tested. The final map shows interesting results and stresses the need for the GIS to test and improve on the groundwater contamination risk assessment methods. Key words: risk assessment, groundwater contamination, vulnerability, GIS, hazard, economic value.

1. Introduction The Environmental Protection Agency, in the United States, has identified in 1993 more than 200 chemical compounds in groundwater (EPA, 1993), some of these are extremely hazardous to human health. Pollutants reach the aquifers from ordinary human activities (agriculture, farming, industry) or from occasional sources. In consequence the groundwater protection is a prior environmental concern in many countries, for example in Europe, where more than 50% of the water supply is obtained from groundwater (Lobo-Ferreira, 1997). The correct way to solve this problem is to prevent the contamination assessing and managing the groundwater pollution risk at regional scale. In this paper the groundwater contamination risk map of a sample alluvial area was produced by using the Ilwis Geographical Information System to construct and overlay thematic maps. The adopted groundwater contamination risk assessment method follows the approach of previous studies published in Italy (Civita, 1995; Civita and De Maio, 1997; Corniello and Ducci, 1997) and is optimised by developing new analytical procedures. Moreover, the use of the GIS has allowed: high spatial resolution; numeric checking of calculated parameters and easy matrix operations. ? Publication No. 1956 of the CNR-GNDCI Theme 4.

280

DANIELA DUCCI

This work is part of a research project on aquifer vulnerability and risk assessment, supported by the CNR-GNDCI, Italian National Research Council – National Group for the Defence against Hydrogeological Disasters (Research Line N. 4- Prof. M. Civita; U.O. N. 4.22; Prof. A. Corniello). 2. Hydrogeological Setting of the Sample Area The Caserta sample area is approximately 100 km2 and is located in the eastern part of the Piana Campana, in Southern Italy (Figure 1) (Corniello et al., 1995a, 1996; Corniello and Ducci, 1997). This area is characterised in the SW by an almost flat morphology, crossed by the Regi Lagni drainage system and in the NE by a gently inclined surface at the base of the limestone massifs. In fact the Piana Campana is a large Quaternary basin delimited by limestone Mesozoic mountains overthrusted during Miocene–Pliocene and disrupted during Pleistocene by distensive-tectonic activity. The plain, characterised by a high level of urbanisation, is filled with alluvial deposits interbedded with pyroclastics derived from Vesuvius and the Phlegrean Fields activities. The lithology of the study area can be described by a succession of three units (from the top): (1) less than 10 m of loose pyroclastic deposits (clay sands); (2) tuffs (the ‘Campanian Ignimbrite’) – thickness between 30 and 20 m; (3) about 80–100 m thick sequence of pyroclastic and/or sedimentary deposits of variable grain size, which represent the main aquifer. The hydrogeological setting is strongly related to thickness and physical characteristics (lithification, granulometry, amount of scoria, etc.) of the Campanian Ignimbrite, which plays the role of either confining (NE sector) or semi-confining (SW sector) bed. Groundwater circulation takes place mainly in the levels with gravel size grain, but the aquifer can be considered as a single continuous body. The aquifer is fed by limestone mountains and by rainfall; the global flow is directed towards the South West. 3. Concept of Groundwater Risk of Contamination Assessment The pollution vulnerability can be defined as ‘the sensitivity of groundwater quality to an imposed contaminant load’ depending only on the ‘proper characteristic of the aquifer’. The vulnerability is relatively static, excluding some variation on the time due, i.e., to the piezometric level changes. The pollution risk considers the impact of human activity on groundwater quality, depending not only on vulnerability but also on the presence of potential contamination sources, which are dynamic factors. The severity of the impact depends on: – the aquifer vulnerability to pollution – the importance of the contamination occurrence, and – the value of the groundwater resource (Lobo-Ferreira, 1997; Lobo-Ferreira and Oliveira, 1997). The term ‘risk assessment’ summarizes the structured process of analyzing risks, as defined below. ‘Risk management’ deals with measures to eliminate, reduce, mitigate, transfer or simply learn to live with risks. The evaluation of risk

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

281

Figure 1. Study area.

assessment separately from risk management limits the utility of risk planning in environmental decision making, but allows the scientist to keep away from the political issues in determining the degree of risk (Gough, 1996). Three different risk assessment approaches can be adopted and are considered in this paper (Gough, 1996): the health risk, where the aim is to estimate the possible harm causing substances; the environmental risk, this refers to the integrity of the whole ecosystem, including plant and animal life; and the engineering-based risk assessment, based on the statistical analysis of past events and extrapolation to the future. The engineering-based risk assessment in Earth Sciences (earthquakes, eruptions, floods, landslides, etc.) is generally defined as the product between the vulnerability, the hazard and the value. Moreover, in groundwater contamination risk assessment the object of the study for the hydrogeologist is the element at risk (the aquifer or the catchments) and the aquifer vulnerability, rather than the phenomenon itself (i.e., seismologists study the earthquakes). In this paper the parametric method of assessing the pollution risk in Italy proposed by Civita (1995), and modified by Civita and De Maio (1997), is applied to the above described area. The method outlined is suitable at regional scale for the individuation of areas of different degree of risk of NonPoint Source pollution (derived from all kind of NPS pollutants); it is greatly different with respect to the analysis of the risk derived from a specific point-source pollutant at detailed scale or from the geographical criteria for groundwater catchment protection zoning. The following three layers have been created to assess the groundwater contamination risk using the geographical information system Ilwis 2.1. (ITC, 1997):

282

DANIELA DUCCI

− the vulnerability map; − the hazard map; − the value map. The risk map is obtained by crossing the basic thematic maps with pixel size of 30 × 30 m using the GIS. A criteria for the linkages of the different GIS layers is established, using cross-tables (Civita and De Maio, 1997). In a previous paper (Corniello and Ducci, 1997) the risk of pollution was calculated in the same area as in the present study. The procedures used to evaluate the risk have been now further improved. 4. The Vulnerability Map Several techniques and simulation models have been developed to assess groundwater vulnerability to pollution on a regional scale. Moreover, with the advent of GIS, the inventory, archival, retrieval and display of spatial data and the link to numerical rating systems and simulation models have become easier (Lynch et al., 1994). Three categories of deterministic models have been coupled to GIS to simulate NPS pollution: regression models, overlay and index models, and transient-state solute transport models (Corwin et al., 1997). Regression models use multiple linear regression to relate various causative factors to the pollution. Overlay and index deterministic models (or point count system methods) refer to those models that compute an index of NPS pollutant mobility from either a simple functional model of steady-state solute transport or a steady-state mechanistic model. Two types of overlay and index models have been developed: property-based and process based. Property-based index models are based on the hydrogeologic setting (e.g. DRASTIC). Transient-state solute transport models describe the processes involved in solute transport: water flow, solute transport, chemical reactions, dispersion and diffusion, etc., (Corwin et al., 1997). In this study the DRASTIC vulnerability method, developed by the USEPA (Aller et al., 1987), was employed, for the following considerations: • The hydrogeologic features of the sample area (presence of the semi-confining bed, depth of the groundwater level) (Corniello et al., 1995, 1996; Napolitano, 1995) limit the influence of the root zone and render the use of solute transport models for the vadose zone unreliable (Corwin and Loague, 1996); • Although same chemical leaching assessments based on modeling approaches are very simple (Corwin and Wagenet, 1996), the available data (especially the distribution of the hydrodynamic parameters) only permitted the use of a property-based index model;

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

283

• To construct specific NPS vulnerability mapping (pesticide, nitrates, etc.) was not required for the conclusive aim, which was to create a nonspecific pollution risk map at regional scale; • DRASTIC, largely used in the U.S.A. and in other countries (Engel et al., 1996; Lynch et al., 1994; Navulur and Engel, 1997), is a simple enough method, and useful in constructing comparable different vulnerability maps. For this purpose the groundwater group of the European Community Commission (EEC) in 1991 decided to standardize the criteria and procedures for mapping the groundwater vulnerability to pollution (Lobo-Ferreira, 1997; Lobo-Ferreira and Oliveira, 1997); • The case-study satisfies the 4 major assumptions on which DRASTIC is based: the contaminant is introduced at the surface of the earth, the contaminant is flushed into the groundwater by precipitation, the contaminant has the mobility of water, and the area evaluated is 0.4 km2 or larger; • In the area examined Napolitano and Fabbri (1996) had already performed a sensitivity analysis for this method to determine the relative importance of the input parameters; the single-parameter sensitivity analysis allowed the effective real weight of the ‘net recharge’ and ‘the impact of the vadose zone’ to be established; • A comparison between three property-based index models for vulnerability assessment, the AVI (Van Stempvoort et al., 1993), DRASTIC and SINTACS (Civita, 1994) methods, in the ‘Piana Campana’ demonstrated that they were all considered suitable, as they all point out the same areas as areas of ‘high vulnerability’ (Corniello et al., 1997). Nevertheless, it is important to point out that this method only allows a preliminary zonation of the area examined: the DRASTIC index is uncorrelated with the movement of pesticides into and through the soil and neglects differences in adsorption, solubility and degradation of each pesticide under different climatic and soil management regimes (Corwin and Loague, 1996). DRASTIC, evaluates vertical vulnerability by using the following seven parameters: Depth to water, net Recharge, Aquifer media, Soil media, Topography, Impact of the vadose zone, and hydraulic Conductivity. Each mapped factor is classified either into ranges (for continuous variables) or into significant media types (for thematic data) which have an impact on pollution potential. Weight multipliers are then used for each factor to balance and enhance their importance. The final vulnerability index (DI ) is a weighted sum of the seven factors and can be computed by using DI = Dr Dw + Rr Rw + Ar Aw + Sr Sw + Tr Tw + Ir Iw + Cr Cw , where D, R, A, S, T , I , and C are the seven parameters, r is the rating value of the analysed sub-area, and w is the weight associated with each parameter. DRASTIC provides two weight classifications, one for normal conditions and the other one

284

DANIELA DUCCI

for conditions in areas where agricultural activity is high. In the sample area the second weight classification was adopted. The resulting map for the Caserta area, constructed by using a GIS (Napolitano, 1995; Corniello et al., 1997) has been modified in relation to the flow rate and direction (as suggested, i.e., in the GNHS vulnerability method – Zaporozec, 1985). Instead of the velocity map the i/nu map was calculated, where i is the hydraulic gradient and nu the cinematic porosity, because the hydraulic conductivity, which is the other factor in the Darcy law (the aquifer is isotropic and has the same permeability in each direction), is already evaluated as a layer of the DRASTIC method and therefore it should not be taken into account twice (Aronoff, 1989). For the evaluation of the flow direction Civita and De Maio (1997) adopted a specific filter available in the ARC/INFO GIS. This application is always feasible by using the neighbour operations in all the GIS environments (Meijerink et al., 1994). In the case study, where the global flow direction is towards the SW, varying in a small range, a simplification has been performed. The total piezometric head was calculated at each point (pixel). It was represented by the difference between the piezometric head in the pixel and the maximum piezometric value in the whole groundwater basin (down from the groundwater divide). In conclusion, the modified vulnerability map (Figure 2) was obtained by multiplying in each pixel the DRASTIC indexes by a coefficient C = 0.9+2∗(Hmax − Hi ) ∗ i/nu , and then by classifying the map. The coefficient C varies in the study area from 0.9 and 1.6. It is interesting to observe that in the Regi Lagni River zone (SW sector of the map) the maximum vulnerability classes are in agreement with the results obtained with other vulnerability methods in this area, i.e., SINTACS (Corniello et al., 1995a; 1997), which considers the drainage from rivers.

5. The Hazard/DCI Map In the groundwater contamination risk the phenomenon causing the damage is not usually natural and the statistical analysis of past events is not significant. In the classical risk formulation the hazard (H ) is a function of time (H = 1−(1−1/t)t ), but in this case the term ‘recurrence time’ is meaningless. In fact, it is possible to have a source of contamination just on the surface of the water body while the contamination never reaches the aquifer. However, contamination might reach the aquifer at regular intervals, if the aquifer is not monitored and the pollution source is not removed. In the mapping of the groundwater contamination risk, the hazard should be defined as the ‘source of danger’ and the ‘layer’ of the hazard is represented by the spatial distribution of the potential contamination sources. The hazard map is substituted by the DCI (Danger of Contamination Index) map. Activities affecting groundwater systems include: agriculture, general farming, industrial activities and other sources as landfills, waste disposal, etc. In the whole area the potential contaminating sources (point-source and NPS), classified

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

285

Figure 2. Aquifer pollution vulnerability map computed using modified DRASTIC method.

in Table I (Civita and De Maio, 1997), have been identified and digitized. The resulting DCI map is shown in Figure 3.

Figure 3. Hazard/DCI map of the Caserta area.

286

DANIELA DUCCI

Table I. Classification in DCI levels of the contamination sources (from Civita and De Maio, 1997) DCI

Industry

Cattle breeding

1

Other punctual sources

Agriculture

Mountain or hill dumps not intercepting the piezometric level

Grazing Cultivation that does not require treatments

2

Generic farm or wild livestock

Cultivation (fruit, vineyard or seed) with limited chemical support (CEE regulation2078/92)

3

Intensive cattle with more than 50 units

Heavy traffic highway

Intensive swine, breeding with less than 500 quintals of live breeding weight. Intensive sheep/goat breeding and horse breeding. Large pisciculturists Intensive swine, poultry, rabbit, breeding with more than 500 quintals of live breeding weight

Dumps intercepting the piezometric level

4

Food and kindred products

5

Various foodstuffs (drinks, tobacco, sugar, etc.)

6

Paper and allied products Textile products

Cultivation in nurseries or greenhouses Cultivation with high chemical support

7

Engineering and Metallurgical

USW (or assimilable) unlined disposals

8

Leather and Leather products Galvanic technology Chemicals and allied products Petroleum refining and related industries (storage in quantities of toxic and/or harmful substances)

Storage tanks of toxic and/or hazardous waste

9

Cultivation with limited chemical support and spreading of cattle manure

Cultivation with limited chemical support and with spreading of swine and rabbit manure

Toxic and/or hazardous unlined waste disposals

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

287

Table II. Social-economic value classes (modified from Civita and De Maio, 1997 and Corniello and Ducci, 1997) Class of the value

Description of the catchments

Very high High

Well, well field or spring supplying more than 50.000 inhabitants. Well, well field or spring supplying a population between 10.000 and 50.000 inhabitants or an industry with more than 100 workers. Well, well field or spring supplying a population between 1.000 and 10.000 inhabitants or an industry with a number of workers between 10 and 99. Well, well field or spring supplying less than 1.000 inhabitants or an industry with less than 10 workers.

Medium

Low

6. The Socio-Economic Value Map Groundwater represents a valuable natural resource: in the risk evaluation the value of the aquifer should be considered as the value of the water supply resource. Nevertheless, it is difficult to quantify the aquifers in terms of socio-economic value. Civita (1995) suggested linking the value with the catchments. The value map (Corniello and Ducci, 1997) has been created as follows, taking into consideration the lack of catchment data: – It was assumed that the area around the well extended from the catchments to a radius of 250 m: this distance is 4–5 times larger than in reality in order to make up for regional gradient and the lack of discharge data; – The value of the identified areas were related to the catchments according to Table II; – Where two or more areas had different value overlap, the largest value was chosen; – In the areas without any reported catchments, the lowest value was always assigned, because of the widespread presence of irrigation wells (90/100 per km2 ). The value map, drawn according to these criteria, is shown in Figure 4. 7. The Contamination Risk Map The contamination risk map was calculated (Figure 5) by making the three basic maps overlap (Vulnerability, Hazard and Value) through the Ilwis GIS by using the cross-tables indicated in Table III, derived from Civita and De Maio (1997). In the study area the risk ranges between very low and extremely high, but the majority of the map ranged from very low to low. The zones at higher risk coincide with the catchment areas. This is the greatest disadvantage of this map as it cre-

288

DANIELA DUCCI

Figure 4. Value map of the Caserta area.

Figure 5. Groundwater contamination risk map built up crossing the maps of Figures 2, 3 and 4 (see Table III).

V→ Vr→ DCI↓ 1 2 3 4 5 6 7 8 9

Very low l m h

vl vl vl vl vl l l

vl vl vl vl l l m

vl vl vl l l l m h

vh

vl vl l m m h vh

Low l m

vl vl vl vl l l l

vl vl vl vl l l l m

h

vh

vl vl l l l m h vh

vl vl l l m m h vh eh

Medium l m h

vl vl vl vl l l l m

vl vl l l l m m h

vl vl l l l m h h vh

vh

vl l l m m h vh vh eh

High l m

h

vh

Very high l m h

vh

Extremely high l m h vh

vl vl vl l l l l m

vl l l m m h h vh eh

l m m h h vh vh eh eh

vl vl vl l l l l m m

m h h vh vh vh eh eh eh

vl l l l l l m m h

vl vl l l m m h vh

vl vl l m m h h vh eh

l m h h h vh vh vh eh

vl l m m h h vh vh eh

m m h h vh vh vh eh eh

h h vh vh vh vh eh eh eh

V: vulnerability degree; Vr: value class; DCI: danger contamination index: vl: very low; l: low; m: medium; h: high; vh: very high; eh: extremely high.

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

Table III. Cross-tables to evaluate the groundwater contamination risk

289

290

DANIELA DUCCI

Figure 6. Groundwater contamination risk map built up by the P.C. analysis applied to the maps of Figures 2, 3 and 4.

ates small areas which also appeared in the contamination risk map constructed in Corniello and Ducci (1997). The use of a larger range of classes in the vulnerability map and in the hazard/DCI map has not solved the problem. An attempt to solve the above mentioned problem and to avoid the subjectivity of the matrices was carried out by applying with the GIS the Principal Components Technique to the three basic maps (Figure 6). The result is interesting in that the map (PC1) is similar to the map of Figure 5. A correlation analysis for the six risk classes was carried out using the GIS Ilwis, which allows spatial statistics (Engel et al., 1996), and the comparison of the maps obtained by using the two methods. The result was a correlation coefficient of 0.8. This result shows the lack of influence of the matrixes and this is an important result because it indicates that research should be continued by improving and standardising the basic maps and rather than changing the method of overlapping them. 8. The Groundwater Quality Map The ground-water quality map (Figure 7) was created following the classification shown in Table IV (Civita et al., 1993), deriving from the Italian and the European legislation. This map makes it possible to evaluate the quality and to find out the sectors suitable for exploitation. The construction of the ground-water quality map was carried out in a GIS environment, by applying the method described in a previous paper (Corniello et al., 1995b). The method suggests codifying the water wells with numbers: 10, 20

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

291

Figure 7. Groundwater quality map (see table IV). Table IV. Groundwater quality classification (from Civita et al., 1993)

Quality

Group 1 Class DH Cond. ◦ ( F) (µS/cm)

Optimum A Medium B Poor C

SO4 (mg/l)

Cl (mg/l)

NO3 (mg/l)

< 30 < 1000 < 50 < 50 < 10 30–50 1000–2000 50–250 50–200 10–50 > 50 > 2000 > 250 > 200 > 50

Group 2 Fe Mn (mg/l) (mg/l)

NH4 (mg/l)

< 0.05 < 0.02 < 0.05 0.05–0.2 0.02–0.05 0.05–0.5 > 0.2 > 0.05 > 0.5

and 30 for the classes A, B and C of the group 1 (Table IV) plus 1, 2 and 3 for the classes A, B and C of the group 2 (Table IV). After codification (11 for the class A1 A2 , 12 for the A1 B2 , . . . 33 for the C1 C2 ) it is easy to zone the area, for example by using the Thiessen polygonals, or another interpolation method. The construction of the quality map through the overlapping of the thematic maps, where the ionic value is represented by isopleths, was discouraged. In fact, this determines, in the final quality map, the creation of areas with classes not existing in reality but only derived from the crossing. The groundwater quality basic map could be used to check the reliability of the hazard and risk maps produced, like the inventory landslides map in same methods for landslide risk assessment (Van Westen, 1993). The statistical comparison between the groundwater quality maps and the maps in Figures 5 and 6, was performed pixel by pixel by using the GIS Ilwis. The

292

DANIELA DUCCI

correlation coefficients are low, around 0.3, but it should be considered normal because of the total independence of the maps, the large number of pixels and the above mentioned impossibility of using probabilistic methods to predict the frequency of future groundwater contamination. 9. Conclusion The pollution risk map of the test area has been obtained by combining three basic thematic maps: vulnerability, hazard and value. A criteria for the linkages of the different GIS layers, proposed in previous papers (Civita and De Maio,1997; Corniello and Ducci, 1997), have been improved and tested by spatial statistics. The procedures used to evaluate the risk makes it possible to make the following suggestions: (a) The use of the GIS techniques is vital and it enables the testing and improvement of the groundwater contamination risk assessment methods; (b) The choice of the kind of matrix to combine the basic thematic maps has no influence on the final risk map; (c) The groundwater flow rate and direction is an important layer of the contamination vulnerability; (d) The groundwater quality map can be used to verify the hazard and the risk maps; (e) The economic value of the groundwater resource at medium-large scale (1 : 50.000 ÷ 1 : 100.000) should not be linked to the well or to the spring, but to the aquifer which feeds the catchments. This paper represents a contribution to the problem of groundwater contamination risk assessment, and it shows the need to continue the research in this direction in order to improve and to standardise methods for the construction of the basic thematic maps and the final map. References Aller, L., Bennet, T., Leher J. H., Petty, R. J., and Hackett, G.: 1987, DRASTIC: A Standardised System for Evaluating Ground Water Pollution Potential Using Hydrogeologic Settings, EPA 600/2-87-035. Aronoff, S.: 1989, Geographic Information Systems: A Management Perspective, WDL Publications, Ottawa, Canada. Civita, M.: 1994, Le carte della vulnerabilità degli acquiferi all’inquinamento. Teoria e pratica, Pitagora, Bologna, Italy. Civita, M.: 1995, Sul rischio di inquinamento delle risorse idriche sotterranee, Quad. Geol. Appl. Pitagora 1(4), 103–119. Civita, M., Dal Pra’, A., Francani, V., Giuliano, G., Oliviero, G., Pellegrini, M., and Zavatti, A.: 1993, Proposta di classificazione e mappatura della qualità delle acque sotterranee, Inquinamento 35, 8–17.

GIS TECHNIQUES FOR MAPPING GROUNDWATER CONTAMINATION RISK

293

Civita, M. and De Maio, M.: 1997, Assessing groundwater contamination risk using ARC/INFO via GRID function, Proceedings of ESRI Conference 1997, San Diego, U.S.A. Corniello, A., de Riso, R., and Ducci, D.: 1990, Idrogeologia e idrogeochimica della Piana Campana, Mem. Soc. Geol. It. 45, 351–360. Corniello, A., de Riso, R., Ducci, D., Napolitano, P., Guarino, P. M., and Bellucci, F.: 1995a, Carta della vulnerabilita’ all’inquinamento dell’acquifero nella zona orientale della Piana Campana, Quad. Geol. Appl. Pitagora 1(3), 63–71. Corniello, A., de Riso, R., Ducci, D., Napolitano, P., Guarino, P. M., and Bellucci, F.: 1996, Pollution vulnerability map of the aquifer in the eastern part of Piana Campana, southern Italy, HydroGIS 96: Application of GIS in Hydrology and Water Resources Management, IAHS, Volume of Poster Paper, pp. 76–77. Corniello, A. and Ducci, D.: 1997, La carta del rischio di inquinamento delle risorse idriche sotterranee: un esempio in Piana Campana, Quad. Geol. Appl. Pitagora 1, 117–131. Corniello, A., Ducci, D., and Napolitano, P.: 1997, Comparison between parametric methods to evaluate aquifer pollution vulnerability using a GIS. An example in the ‘Piana Campana’, In: P. G. Marinos, G. C. Koukis, G. C. Tsiambaos and G. C. Stoumaras (eds), Engineering Geology and the Environment, Balkema, Rotterdam, The Netherlands, pp. 1721–1726. Corniello, A., Ducci, D., and Napolitano, P.: 1995b, La piana del medio corso del F. Volturno: Carta della qualità delle acque sotterranee 1/50.000, Quad. Geol. Appl. Pitagora 1(3), 499–505. Corwin, D. L. and Loague, K., (eds): 1996, Applications of GIS to the Modeling of Non-Point Source Pollutants in the Vadose Zone, Soil Science Society of America Special Publication, Vol. 48. Corwin, D. L., Vaughan, P. J., and Loague, K.: 1997, Modeling nonpoint source pollutants in the vadose zone with GIS, Environmental Science and Technology 31(8), 2157–2175. Corwin, D. L. and Wagenet, R. J.: 1996, Applications of GIS to the modeling of non-point source pollutants in the vadose zone: A conference overview, J. Environ. Qual. 25(3), 403–411. Engel, B., Navulur, K., Cooper, B., and Hahn, L.: 1996, Estimating groundwater vulnerability to nonpoint source pollution from nitrates and pesticides on a regional scale, HydroGIS 96: Application of GIS in Hydrology and Water Resources Management, IAHS Publ., Vol. 235, pp. 521–526. EPA Environmental Protection Agency: 1993, Wellhead Protection: A Guide for Small Communities, Seminar Publication, EPA/625/R-93/002, Washington, U.S.A. Gough, J. D.: 1996, Environmental decision making and risk management for groundwater systems, Proceedings of Tephra, New Zealand. ITC: 1997, ILWIS 2.1 for Windows. User’s, Reference and Applications Guides, Ilwis Department, ITC International Institute for Aerospace Survey & Earth Sciences, Enschede, The Netherlands. Lobo-Ferreira, J. P. and Oliveira, M. M.: 1997, Drastic groundwater vulnerability mapping of Portugal, In: Groundwater: An Endangered Resource, Proceedings of Theme C of the 27th Congress of the International Association for Hydraulic Research, pp. 132–137. Lobo-Ferreira, J. P.: 1997, Gis and mathematical modelling for the assessment of vulnerability and geographical zoning for groundwater management and protection, Proceedings of Nato Advanced Research Workshop Environment Contamination and Remediation Practices at Former and Present Military Bases. Lynch, S. D., Reynders, A. G., and. Schulze, R. E.: 1994, Preparing input data for a national-scale groundwater vulnerability map of Southern Africa, Water S.A. 20, 239–246. Meijerink, A. M. J., de Brouwer, H. A. M., Mannaerts, C. M., and Valenzuela, C.: 1994, Introduction to the use of Geographic Information Systems for Practical Hydrology, ITC Pub. N. 23, Enschede. Napolitano, P.: 1995, GIS for aquifer vulnerability assessment in the Piana Campana, Southern Italy, using the DRASTIC and SINTACS methods, MSc Thesis, ITC, Enschede, The Netherlands. Napolitano, P. and Fabbri, A. G.: 1996, Single-parameter sensitivity analysis for aquifer vulnerability assessment using DRASTIC and SINTACS, HydroGIS 96: Application of GIS in Hydrology and Water Resources Management, IAHS 235, 559–566.

294

DANIELA DUCCI

Navulur, K. and Engel, B. A.: 1997, Predicting spatial distribution of vulnerability of Indiana State Aquifer system to nitrate leaching using a GIS, Third International Conference/Workshop on Integrating GIS and Environmental Modeling, Santa Fe, U.S.A. Van Stempvoort, D., Evert, L., and Wassenaar, L.: 1993, Aquifer vulnerability index: a GIS compatible method for groundwater vulnerability mapping, Canad. Water Resour. J. 18, 25–37. Van Westen, C. J.: 1993, GISSIZ: Geographic Information Systems in Slope Instability Zonation, ITC Publication, 15, Enschede, The Netherlands. Zaporozec, A. (ed.): 1985, Groundwater protection principles and alternatives for Rock Country, Wisconsin Geol. Nat. Hyst. Survey, Spec. Rep. 8, 73.