INVESTIGADORES
BERTILLER monica Beatriz
artículos
Título:
Estimates of dryland degradation with Fourier signatures in low-altitude high resolution monochromatic images
Autor/es:
ARES, J.O.; BERTILLER, M.B; BISIGATO, A.J.
Revista:
LANDSCAPE ECOLOGY
Editorial:
SPRINGER
Referencias:
Lugar: Berlin; Año: 2003 vol. 16 p. 51 - 63
ISSN:
0921-2973
Resumen:
Most world drylands are used as grazing lands and undergo degradation of their vegetation canopy. The plantcanopy is typically structured in patched arrangements, inducing fertility islands critical to maintenance of ecosystem properties. The characteristics of patch structure (size of patches, connectivity-continuity of patch units,etc.) are indicators of the degree of dryland deterioration. We characterized changes in patch structure induced bysheep grazing at a landscape scale using monochromatic low-altitude imagery digitized to a spatial resolution ofabout 1 m with standard techniques of harmonic analysis applied to develop Fourier signatures. The signaturesdeveloped on image line transects were tested with ground samples and mathematical models of plant canopiesin several dryland fields where spatial deterioration gradients existed. The sensitivity and errors associated tolong-wave noise introduced by the geometry of the camera-field-Sun spatial arrangement and to high frequencynoise introduced by the digitizing process were evaluated by applying suitable filters in the frequency domain.Fourier signatures developed on monochromatic low-altitude imagery proved to be indicative of changes in thepatching arrangements of the plant canopy. It is concluded that adequately filtered, high spatial resolution monochromatic images can be used to evaluate the degree of deterioration of dryland landscapes through the computation of selected Fourier signatures in their frequency domain. At comparable cost, aerial photography allowsinspecting the landscape at a higher spatial resolution than those attainable with satellite imagery. Also, aerialphotos of many areas are available for earlier dates than images from remote sensors, which would allow betterinspection of long-term ecosystem changes.