In this study, an assessment of the landslide susceptibility using GIS and bivariate statistical analysis in the Calaggio Stream basin (Campanian Apennines) has been performed. Lithology, land-use, slope and aspect have been assumed as causal factors.The lithological, geomorphological and land-use features of the study area have been defined both in the field and by means of aerial photos and topographic maps analysis and reported on thematic maps. From the geomorphological map, a landslide inventory map has also been produced. A slope map and an aspect map have been automatically derived from a 20 m per pixel DEM. The landslide inventory map has then been combined with each thematic map to determine the density of the landslides detachment surfaces (LDS) within each class. This has allowed us to calculate the weighting values for each class of the considered causal factors, using the formula proposed by van Westen (1993), slightly modified: in which: Wi=weighting value of the class i; NpixDSi=number of pixels of the LDS within the class i; NpixNi=number of pixels within the class i; NpixDSi=total number of pixels of the LDS within the whole study area; NpixNi=total number of pixels of the whole study area map.The thematic maps have been then reclassified on the basis of the weighting values, which have been summed for each point of the study area by means of an overlay process. The obtained range of values has been subdivided into four intervals, corresponding to four susceptibility classes (van Westen, 1993). As verification, the produced susceptibility map has been first compared with the geomorphological map and, then, combined with the landslide inventory map.The total area of the detected landslides is 23,4 kmq, which is 7.5% of the study area. Landslides mainly consists of soil slips, mainly occurring in soils with vertic properties, flows and rotational slides. From a lithological standpoint, seven complexes have been recognized. The highest weighting value occurs in the marly-clayey-calcareous complex (0.44). Among the land-use classes, most part of the LDS occurs on surfaces covered by shrubby and herbaceous vegetation (weighting value: 0.61). With respect to slope and aspect, the highest weighting values occur in the class 16°-20° (2.33) and on slopes facing from south to west (0.37), respectively.The landslide susceptibility is high or very high in the 50% of the study area, moderately high in the 17%, moderately low in the 4% and low or very low in the 29%. A good correspondence has been highlighted between the areas characterized by the lowest susceptibility and the most stable landsurfaces detected during the field survey (top surfaces of river terraces and remnants of old erosional landscapes); similarly, most part of the unstable slopes falls within the highest susceptibility areas. The areas characterized by the highest susceptibility also display the highest percentage of surface effectively affected by landslides; such a percentage regularly decreases with the degree of susceptibility, as expected.In conclusion, this study has allowed us to assess the landslide susceptibility in the Calaggio Stream basin (Campanian Apennines). In the study area, 50% of the landsurfaces display a high or very high landslide susceptibility. Such a result could represent a first but fundamental step in set up safety measures and in land use planning.
Landslide susceptibility assessment using GIS and bivariate statistical analysis in the geomorphological framework of the Calaggio Stream basin (southern Italy)
MAGLIULO P
;RUSSO F.
2007-01-01
Abstract
In this study, an assessment of the landslide susceptibility using GIS and bivariate statistical analysis in the Calaggio Stream basin (Campanian Apennines) has been performed. Lithology, land-use, slope and aspect have been assumed as causal factors.The lithological, geomorphological and land-use features of the study area have been defined both in the field and by means of aerial photos and topographic maps analysis and reported on thematic maps. From the geomorphological map, a landslide inventory map has also been produced. A slope map and an aspect map have been automatically derived from a 20 m per pixel DEM. The landslide inventory map has then been combined with each thematic map to determine the density of the landslides detachment surfaces (LDS) within each class. This has allowed us to calculate the weighting values for each class of the considered causal factors, using the formula proposed by van Westen (1993), slightly modified: in which: Wi=weighting value of the class i; NpixDSi=number of pixels of the LDS within the class i; NpixNi=number of pixels within the class i; NpixDSi=total number of pixels of the LDS within the whole study area; NpixNi=total number of pixels of the whole study area map.The thematic maps have been then reclassified on the basis of the weighting values, which have been summed for each point of the study area by means of an overlay process. The obtained range of values has been subdivided into four intervals, corresponding to four susceptibility classes (van Westen, 1993). As verification, the produced susceptibility map has been first compared with the geomorphological map and, then, combined with the landslide inventory map.The total area of the detected landslides is 23,4 kmq, which is 7.5% of the study area. Landslides mainly consists of soil slips, mainly occurring in soils with vertic properties, flows and rotational slides. From a lithological standpoint, seven complexes have been recognized. The highest weighting value occurs in the marly-clayey-calcareous complex (0.44). Among the land-use classes, most part of the LDS occurs on surfaces covered by shrubby and herbaceous vegetation (weighting value: 0.61). With respect to slope and aspect, the highest weighting values occur in the class 16°-20° (2.33) and on slopes facing from south to west (0.37), respectively.The landslide susceptibility is high or very high in the 50% of the study area, moderately high in the 17%, moderately low in the 4% and low or very low in the 29%. A good correspondence has been highlighted between the areas characterized by the lowest susceptibility and the most stable landsurfaces detected during the field survey (top surfaces of river terraces and remnants of old erosional landscapes); similarly, most part of the unstable slopes falls within the highest susceptibility areas. The areas characterized by the highest susceptibility also display the highest percentage of surface effectively affected by landslides; such a percentage regularly decreases with the degree of susceptibility, as expected.In conclusion, this study has allowed us to assess the landslide susceptibility in the Calaggio Stream basin (Campanian Apennines). In the study area, 50% of the landsurfaces display a high or very high landslide susceptibility. Such a result could represent a first but fundamental step in set up safety measures and in land use planning.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.