Title of article
Scaling of differentially eroded surfaces in the drainage network of the Ethiopian Plateau
Author/Authors
Beaulieu، نويسنده , , Alexandre and Gaonacʹh، نويسنده , , Hélène، نويسنده ,
Issue Information
روزنامه با شماره پیاپی سال 2002
Pages
12
From page
111
To page
122
Abstract
Differentially eroded regions selected over the Ethiopian Plateau, Northeast Africa, were statistically analyzed using satellite images of various electromagnetic spectrum regions and resolutions (Landsat TM and ERS-1). The power spectrum exponent β values for the Landsat TM2 (visible) and ERS-1 images over the same surface are associated to the intrinsic properties of the different sensor type and postprocessing of the data. Differences in the β values were observed between eroded area (Blue Nile Canyon, BNC) and relatively noneroded area (plateau, PLA) for all data sets. These differences are associated to the mechanical erosion of the plateau. The remotely sensed data fields show scaling from 35 m to 15 km, with no break at 1.5 km, and are highly multifractal. Analyses of the Landsat TM bands over each area demonstrated something particular: β values for bands in the shortwave infrared (SWIR) range differed from β values for bands in the visible spectrum range in the plateau area by about 0.48, while in the drainage area, this difference is around 0.13. Landsat TM SWIR bands are sensitive to spectral signature of clay minerals, while data in the visible spectrum range mostly depict topography gradients. Two concurrent processes are highlighted, mechanical erosion and chemical erosion/deposition, which interact to produce the observed differences. In the drainage area, where cliffs and steep slopes are present and mechanical erosion intensively occurs, the alteration minerals are remobilized quickly, whereas in the plateau area, mechanical erosion is low, and alteration mineral deposition is less disturbed. Such new statistical highlights of topographic versus chemical surfaces will have to be taken into account in landforming models.
Journal title
Remote Sensing of Environment
Serial Year
2002
Journal title
Remote Sensing of Environment
Record number
1574003
Link To Document