HR: 14:10h
AN: H42J-03    [PDF]
TI: Landslides, Earthquakes, and Erosion
AU: * Malamud, B D
EM: bruce@malamud.com
AF: Environmental Monitoring and Modeling Research Group, Dept of Geography, King's College London, Strand, London, WC2R 2LS United Kingdom
AU: Turcotte, D L
EM: turcotte@geology.ucdavis.edu
AF: Department of Geology, University of California at Davis, One Shields Avenue, Davis, CA 95616 United States
AU: Guzzetti, F
EM: fausto.guzzetti@irpi.cnr.it
AF: CNR-IRPI Perugia, via della Madonna Alta 126, Perugia, 06128 Italy
AU: Reichenbach, P
EM: paola.reichenbach@irpi.cnr.it
AF: CNR-IRPI Perugia, via della Madonna Alta 126, Perugia, 06128 Italy
AB: In this paper, we quantify the role of landslides in erosion and in particular, the role of earthquakes to landslides to erosion. A landslide event consists of single to tens of thousands of landslides associated with a trigger, such as an earthquake, sudden snowmelt, or extended precipitation. We have previously introduced a general landslide distribution for landslide events and also a corresponding landslide-event magnitude scale, {\it M$_{L}$} = log({\it N$_{LT}$}), with {\it N$_{LT}$} the total number of landslides associated with the landslide event. The magnitude {\it m$_{L}$} can be determined from both substantially `complete' inventories and partial inventories of the largest landslides in the event. Equivalent magnitude values can also be obtained for historical landslide inventories, which represent the sum of many landslide events over periods of one to many thousands of years or more. We believe that our landslide distribution does a rough quantification of the total area (and volume) of landslides that occur for landslide events of different magnitudes, and use this distribution to relate our landslide magnitude to total landslide volume. Using estimated recurrence intervals for three landslide event inventories, we have inferred regional erosion rates due to landslides for the Northridge earthquake-triggered, Umbria (Italy) snowmelt-triggered, and Guatemala heavy-rain triggered events, as 0.1, 0.4, and 2.5 mm yr$^{-1}$, respectively. Comparing historical inventories to our general landslide event distribution, we have made extrapolations and estimated total landslide volumes associated with two historical inventories in Italy and Japan. Using estimates for time intervals over which these historical landslides accumulated, the associated long-term erosion rates were found to be 1.0 mm yr$^{-1}$ in Umbria, Italy and 2.2 mm yr$^{-1}$ in Japan. We then use an empirical relationship between total landslide volume and earthquake magnitude to determine analytic relationships between earthquake magnitude and the associated landslide event magnitude, area and volume of the largest landslide triggered, and total area of all landslides triggered. Using the Gutenberg--Richter frequency-magnitude relation for regional seismicity, we analytically relate the seismically-induced erosion rate to the regional seismic intensity and the moment magnitude of the largest regional earthquake. We find that typical seismically-induced erosion rates in very active subduction zones is $\approx$ 0.2--2 mm yr$^{-1}$ and adjacent to plate boundary strike-slip zones $\approx$ 0.01--0.2 mm yr$^{-1}$.
DE: 1815 Erosion and sedimentation
DE: 1824 Geomorphology (1625)
DE: 3200 MATHEMATICAL GEOPHYSICS (New field)
DE: 7223 Seismic hazard assessment and prediction
SC: Hydrology [H]
MN: 2003 Fall Meeting