Karstic aquifer vulnerability assessment methods and results at a test site (Apulia, southern Italy)

• Published in: Natural hazards and earth system sciences 2009-01, Vol.9 (4), p.1461-1470
• Main Authors: Polemio, M, Casarano, D, Limoni, P P
• Format: Article
• Language: English
• ISSN: 1684-9981; 1561-8633; 1684-9981
• DOI: 10.5194/nhess-9-1461-2009

Karstic aquifers are well known for their vulnerability to groundwater contamination. This is due to characteristics such as thin soils and point recharge in dolines, shafts, and swallow holes. In karstic areas, groundwater is often the only freshwater source. This is the case of the Apulia region (south-eastern Italy), where a large and deep carbonate aquifer, affected by karstic and fracturing phenomena, is located. Several methods (GOD, DRASTIC, SINTACS, EPIK, PI, and COP) for the assessment of the intrinsic vulnerability (Iv) were selected and applied to an Apulian test site, for which a complete data set was set up. The intrinsic vulnerability maps, produced using a GIS approach, show vulnerability from low to very high. The maximum vulnerability is always due to karstic features. A comparison approach of the maps is proposed. The advantages and disadvantages of each method are discussed. In general terms, three groups can be distinguished. The GOD method is useful for mapping large areas with high vulnerability contrasts. DRASTIC and SINTACS are "any-type aquifer" methods that have some limitations in applications to karstic aquifers, especially in the case of DRASTIC. EPIK, PI, and COP, which were designed to be applied to carbonate or karstic aquifers, supply affordable results, highly coherent with karstic and hydrogeological features, and reliable procedures, especially in the case of PI and COP. The latter appears simpler to apply and more flexible in considering the role of climatic parameters. If Iv of each method is considered, the highest variability is observed in cells in the neighbourhood of karstic features. In these spatial domains, additional efforts to define more reliable and global methods are required.