Integrated use of satellite remote sensing, GIS, and ground spectroscopy techniques for monitoring olive oil mill waste disposal areas on the island of Crete, Greece

Olive oil mill wastes (OOMW) constitute a major pollution factor in olive-growing regions and an important problem to be solved for the agricultural industry. Olive oil mill wastes are normally deposited in tanks, or directly into the soil or even on adjacent torrents, rivers, and lakes, posing a high risk of environmental pollution in regard to public health. This study aims to develop integrated satellite remote sensing, geographical information systems (GIS), and ground spectroscopy methodologies to detect and monitor OOMW disposal areas on the island of Crete, Greece in the Southeastern Mediterranean. More than 1000 disposal tanks were mapped through an extended global positioning system (GPS) survey that took place throughout the island. Satellite images of both high (IKONOS) and medium (Landsat 8 OLI (Operational Land Imager)) resolution were preprocessed and analysed by applying geometric, radiometric, and atmospheric corrections. A library with a spectral signature of OOMW including both different time periods and satellite sensors was developed. At the same time, ground spectroscopy campaigns were carried out and a complementary spectral signature library was developed. The narrow band reflectance of ground measurements was recalculated using the relative response filters of the corresponding satellite sensors. Both libraries were compared for their accuracy through statistical approaches and the optimum spectral range for detecting OOMW areas was estimated. Subsequently, further auxiliary image-processing techniques such as image fusion, linear spectral unmixing (LSU), false-colour composites (FCCs), image classification, and principal component analysis (PCA) were applied to satellite images to enhance OOMW patterns, and an innovative OOMW detection index for Landsat 8 was developed. In addition, several vegetation indices were applied and compared in regard to their efficiency in detecting waste ponds. Finally an integrated, semi-automatic methodology was developed in the GIS environment employing classification algorithms for the detection of waste ponds. This study highlights the potential of satellite remote sensing, GIS, and ground spectroscopy in the semi–automatic detection of OOMW disposal areas in the context of the Mediterranean landscape.