Identification and mapping of landforms in geomorphology are based on geological and geomorphological survey and interpretation of topographic maps and aerial photos. Considerable enhancement for morphometric interpretation can be obtained through generation of a synthetic stereo pair, by means of the integration of spectral data with Digital Elevation Model derived by Shuttle Radar Topography Mission (SRTM). The global coverage and moderate spatial resolution (30 m) offered by Landsat 7 are a necessary complement to the SRTM imagery and the combined use of both systems would allow for greater accuracy than either could provide independently. Although earlier applications of previous Landsat missions have met with mixed success for landscape mapping, more sophisticated methodologies combined with advances in the ETM+ sensor will facilitate the mapping objective. In addition to an improved signal/noise ratio in the multispectral bands and higher spatial resolution (60 m) in the thermal band, the ETM+ sensor also provides a panchromatic band with 15 m spatial resolution. The multispectral and panchromatic bands can be combined using pan-sharpening algorithms to provide a more detailed view. Classification algorithms that accommodate spectral heterogeneity provide a way to discriminate heterogeneous built environments from more homogeneous natural environments. Spectral Mixture Analysis (SMA) classifies individual mixed pixels according to the distribution of spectrally pure end member fractions and provides a tool for discrimination and classification. Some landforms, as DSGSD (Deep Seated Gravitational Seated Deformations) have a well defined evidences recognizable by topographic surface observation and deriving from their morphologic characterization. Trenches, double ridges and counterslopes are the superficial evidences of the mass movements scarp edges. Sagging, cambering and a widespread landsliding are the evidences of compressional stress deformation in the lower part of the slope. Our methodology is direct towards the automatic analysis of the geomorphologic parameters which characterize the arrangement of the DSGSD (such as the slope, the curvature and the relief), starting from a DEM. The method classifies the landscape into geographic areas as function of complex interdependent parameters rather than a single parameters. To understand the behaviour of all the parameters pertaining to DSGSD and to process their automatic reduction in morphological unit we use the three bands (147) of Landsat ETM. Then we used PCA analysis in order to map them for separating homogenous areas into these morphological unit. In this case PCA helps us to groups the DSGSD areas bringing out the similarity in the morphometric characteristics. The multitemporal spectral mixture analysis will yield continuous change fraction maps as well as thematic classifications for each date. A class change analysis will yield the discrete change classifications. The result of the mapping will be continuous end member fraction maps and thematic classification maps for 1990 and 2000 as well as continuous and discrete change maps. These can be used to constrain the spatial distribution and type of DSGSD regions. On the basis of these spectra, we anticipate that it will be possible to determine which classes can and cannot be distinguished in the Landsat imagery.

A 3D approach using Landstat images types and SRTM data to map Deep Seated Gravitational Slope Deformations (DSGSD).

TARAMELLI, Andrea;MELELLI, Laura
2005

Abstract

Identification and mapping of landforms in geomorphology are based on geological and geomorphological survey and interpretation of topographic maps and aerial photos. Considerable enhancement for morphometric interpretation can be obtained through generation of a synthetic stereo pair, by means of the integration of spectral data with Digital Elevation Model derived by Shuttle Radar Topography Mission (SRTM). The global coverage and moderate spatial resolution (30 m) offered by Landsat 7 are a necessary complement to the SRTM imagery and the combined use of both systems would allow for greater accuracy than either could provide independently. Although earlier applications of previous Landsat missions have met with mixed success for landscape mapping, more sophisticated methodologies combined with advances in the ETM+ sensor will facilitate the mapping objective. In addition to an improved signal/noise ratio in the multispectral bands and higher spatial resolution (60 m) in the thermal band, the ETM+ sensor also provides a panchromatic band with 15 m spatial resolution. The multispectral and panchromatic bands can be combined using pan-sharpening algorithms to provide a more detailed view. Classification algorithms that accommodate spectral heterogeneity provide a way to discriminate heterogeneous built environments from more homogeneous natural environments. Spectral Mixture Analysis (SMA) classifies individual mixed pixels according to the distribution of spectrally pure end member fractions and provides a tool for discrimination and classification. Some landforms, as DSGSD (Deep Seated Gravitational Seated Deformations) have a well defined evidences recognizable by topographic surface observation and deriving from their morphologic characterization. Trenches, double ridges and counterslopes are the superficial evidences of the mass movements scarp edges. Sagging, cambering and a widespread landsliding are the evidences of compressional stress deformation in the lower part of the slope. Our methodology is direct towards the automatic analysis of the geomorphologic parameters which characterize the arrangement of the DSGSD (such as the slope, the curvature and the relief), starting from a DEM. The method classifies the landscape into geographic areas as function of complex interdependent parameters rather than a single parameters. To understand the behaviour of all the parameters pertaining to DSGSD and to process their automatic reduction in morphological unit we use the three bands (147) of Landsat ETM. Then we used PCA analysis in order to map them for separating homogenous areas into these morphological unit. In this case PCA helps us to groups the DSGSD areas bringing out the similarity in the morphometric characteristics. The multitemporal spectral mixture analysis will yield continuous change fraction maps as well as thematic classifications for each date. A class change analysis will yield the discrete change classifications. The result of the mapping will be continuous end member fraction maps and thematic classification maps for 1990 and 2000 as well as continuous and discrete change maps. These can be used to constrain the spatial distribution and type of DSGSD regions. On the basis of these spectra, we anticipate that it will be possible to determine which classes can and cannot be distinguished in the Landsat imagery.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11391/21090
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact