2005 Salt Lake City Annual Meeting (October 16–19, 2005)

Paper No. 2
Presentation Time: 3:55 PM


TURCOTTE, Donald L.1, MALAMUD, Bruce D.2, GUZZETTI, Fausto3 and REICHENBACH, Paola3, (1)Department of Geology, University of California, Davis, 1 Shields Avenue, 174 Physics/Geology Building, Davis, CA 95616, (2)Environmental Monitoring and Modeling Research Group, Department of Geography, King's College London, Strand, London, WC2R 2LS, United Kingdom, (3)CNR - IRPI Perugia, via della Madonna Alta, 126, Perugia, 06128, Italy, turcotte@geology.ucdavis.edu

Seismogenic zones generally have metastable regions subject to landslides when triggered by an earthquake. A triggered landslide event can consist of a single to millions of landslides. There is accumulating evidence that the frequency-size distribution of landslides in an event satisfy a general probability distribution. One consequence of this generality is that a landslide event magnitude, mL, can be defined by the relation mL = logNLT where NLT is the total number of landslides in the landslide event. The general landslide distribution can also be used to infer a landslide event magnitude from incomplete landslide inventories, for example, if evidence for the existence of only the largest landslides has been preserved. The primary question addressed in this paper is whether the landslide magnitude can be related to the earthquake magnitude. We have previously considered this question (Earth and Planetary Science Letters, v. 229, p. 45, 2004) and will update the correlations. It is recognized that there is a minimum earthquake magnitude ME = 4.3 ± 0.4 necessary to generate landslides. We believe that a useful correlation between the two magnitudes is given by mL = 1.27ME – 5.45(± 0.46). We certainly recognize that there is considerable variability in the association of the two magnitudes. These include earthquake considerations such as location and depth as well as geomorphic considerations such as topography, rock and soil type, and moisture content. There is also the question of the extent of metastability present at the time of the earthquake. The landslide caused by a mainshock will reduce the metastability so that aftershocks are likely to produce few landslides. We believe that the basic scale invariance of the metastability leads to the applicability of the general landslide distribution independent of earthquake magnitude and of other considerations. Although recognizing the considerable variability of correlations between the earthquake magnitude, ME, and the landslide event magnitude, mL, we believe our proposed correlation has considerable use in assessing the landslide hazard and the contributions of earthquake induced landslides to erosion.