Tourism is the basic industry in Fuerteventura Island (Canary Islands, Spain), mostly due to the sunny conditions and excellent beaches. Probably the best well-known beaches are those located in the southeastern part of Jandía Peninsula, locally named Sotavento Beaches. These beaches are the largest and widest in the whole archipelago, and are up to 15 km long and 800 m wide in some points. Interest has focussed on these beaches not only from the tourism point of view, but also from a geomorphological perspective. Based on beach profile data and aerial photographs from 1963 until 1996 from the central part of these beaches, landward migration of the coastline has been identified. The width reduction of these beaches –to the point that in some areas the beach has completely disappeared- is a consequence of the reduction of sediment supply to the coast from inland sedimentary deposits. This reduction in sediment supply and consequent beach erosion is due to the development of tourism resorts and associated activities, such as road widening, sand mining and gardening along the roads.

The Coastal Vulnerability Index (CVI) has been widely applied around the world. This study provides a review of the suitability of the variables and mathematical expression of the CVI and proposes a new Integrated Coastal Vulnerability Index (ICVI), comparing both indices for 4 study areas in the southern Caribbean. The ICVI assesses vulnerability to sea level rise by integrating the Environmental Vulnerability Index (EVI) and the Socioeconomic Vulnerability Index (SVI). Regarding the variables that constitute the indices, it is noted that the CVI includes vulnerability and hazard variables and therefore it should be considered a risk index rather than a vulnerability index. The EVI includes geomorphological vulnerability variables, like the CVI, but also ecological ones. Regarding the mathematical expression, the use of the arithmetic mean versus the formula proposed for the CVI is discussed based on the comparison of the results obtained for EVI and SVI in the 4 study areas. In addition, the use of absolute (between 0 and 1) or relative (based on percentiles) limits in these indices, and the use of weights or not, are also discussed. The conclusion is that the use of relative thresholds necessarily forces the identification of very low to very high vulnerability zones for any study, and the use of weights on the variables increases the subjectivity of the assessment, all of which impedes the comparability of the index at a global level. Therefore, the ICVI, with the formula based on the arithmetic mean, with absolute limits between 0 and 1 and without variable weightings, is preferable to the CVI for use at the global level.