Links

Tools

Export citation

Search in Google Scholar

Using Palsar imagery for characterizing volcanic fissures (DYKES) in the Ethiopian Volcanic Province

Journal article published in 2009 by Daniel Mège ORCID
This paper is available in a repository.
This paper is available in a repository.

Full text: Download

Question mark in circle
Preprint: policy unknown
Question mark in circle
Postprint: policy unknown
Question mark in circle
Published version: policy unknown

Abstract

Palsar images are used to delineate and identify dykes in the western Ethiopian volcanic province. Dykes are vertical, fractured solidified magma sheets that play a major role in groundwater storage, and might be considered as potential drilling sites for water supply to the local inhabitants. On Palsar imagery, they can be discriminated from other linear patterns by roughness contrasts between the dyke and its debris slopes and between the debris slopes and the surrounding smooth plain. For an appropriate combination of incidence angle and polarization, dyke signal on Palsar imagery is therefore made of 3 parallel bands of distinct roughness, with the debris slopes appearing brighter than both the dyke and the surrounding plain. In comparison, multispectral images are good at identifying vegetated dyke portions, but not at tracing dyke continuity. Moreover, uncertainty frequently remains as to the discrimination between dykes and other lienear patterns. Envisat/Asar (band C) and SIR-C (bands C and L) images are very good at identifying dyke continuity, but not at identifying the 3 diagnostic roughness bands, more because of coarser resolution than for incidence angle, beam wavelength, or polarization issues. The ideal sensor for dyke identification will have a resolution higher than Palsar because still many dykes are too small to be observed by Palsar, a 4-pol mode because the dyke signal is always a trade-off between roughness, vegetation and moisture, and a combination of wavelength (C or L) and incidence angle (20-50°) that is flexible enough to image roughness at a scale of 5 to 100 cm.