The country of Spain is representative of land change processes in Mediterranean member states of the European Union (EU). These land change processes are often triggered by European, national and sub-national policies and include widespread land abandonment and urbanisation trends, as well as an increase in land use intensities accompanied by strong exploitation of water resources. The Mediterranean is part of the dryland ecoregion, which is particularly vulnerable to ecosystem degradation. While remote sensing data permit the characterisation of the temporal dimension of land surface processes, the syndrome-based approach aims to integrate this with information on local/regional socio-economic and physical frameworks. In this study, we incorporated two major drivers of land change, climatic boundary conditions and population density change, to understand the patterns of the assessed land cover changes.
We used the Mediterranean Extended Daily One Km AVHRR Data Set (MEDOKADS), which represents the Mediterranean Basin at a resolution of 1 km2 for the period of 1989–2004 at 10-day time intervals. The long-term vegetation trend could be characterised based on a linear regression analysis of Normalised Difference Vegetation Index (NDVI) values. Further descriptors of phenology were calculated, such as the amplitude of annual cycles or the date of occurrence of the annual maximum of vegetation cover. Subsequently, trend analysis was applied to these parameters to map phenology shifts representative of changes in land use/cover.
Next, a rule-based classification was employed to integrate the individual results and assign syndromes representing specific land change process patterns. For natural and semi-natural areas, locations characterised by increasing biomass are associated with areas of population loss due to migration to urban centres. By contrast, many fertile agricultural areas are associated with intensification processes, including the expansion of irrigation schemes. Furthermore, areas that are dominated by inter-annual variability in precipitation were identified by employing the results of an analysis of NDVI–rainfall relationships. The resulting data set identifies major land change syndromes for the whole of Spain and links the apparent results of change to driving factors and causes, thus serving as an excellent data source to inform policy makers of effects caused by legislative and management actions.