Combining digital elevation data (SRTM/ASTER), high resolution satellite imagery (Quickbird) and GIS for geomorphological mapping: A multi-component case study on Mediterranean karst in Central Crete

Abstract

Remote sensing data have become more and more popular for geomorphological investigations because their steadily increasing level of detail and accessibility opens up new potentials. In this context, this paper examines the application and quality of digital elevation models (SRTM and ASTER DEMs), high resolution satellite imagery (Quickbird) and GIS techniques for the detection and mapping of karst landforms (mainly enclosed depressions) at different scales in the Ida Mountains of Central Crete. Besides discussing methodological issues and evaluating suitability potentials, we conducted an exemplary case study based on spatial analysis of the regional karst morphology. Different input datasets and processing methods are applied (GIS-based analysis, land cover classification, raster calculations, etc.) in order to carry out an area-wide surveying and mapping of karst depressions. The findings are supported and validated by auxiliary field studies. Due to the level of detail and occasional data errors, an exclusive use of satellite imagery or digital elevation models for automatic karst landform detection performs insufficiently. Our results demonstrate that mapping karst features through remote sensing is significantly dependent on scale of interest, existing environmental conditions and data quality. A semi-automatic data integration approach on the basis of digital datasets generated by combined satellite image processing and DEM analysis yields the best results, especially when indirect karst-indicating variables like iron oxide-rich sediments are included as detection criteria. The multi-component application presented in this paper provides a time-saving and effective tool for meso- to macro-scale object detection and extensive study areas. However, the potential of fully automated karst feature mapping still needs to be explored in future work. Concerning the spatial dimension of karstification in Central Crete, the GIS-based results allow differentiating further geomorphological characteristics, e.g. by indicating a significant altitudinal change of karst forms within the study area. Size and shape of depressions (dolines, uvalas, and poljes) vary considerably due to the geological setting, climatic impacts, neotectonics and elevation.