Geological and Atmospheric Sciences Publications Geological and Atmospheric Sciences 10-20-2000 Acute Sensitivity of Landslide Rates to Initial Soil Porosity Richard M. Iverson United States Geological Survey M. E. Reid United States Geological Survey Neal R. Iverson Iowa State University, [email protected] R. G. LaHusen United States Geological Survey M. Logan United States Geological Survey See next page for additional authors Follow this and additional works at: http://lib.dr.iastate.edu/ge_at_pubs Part of the Hydrology Commons, and the Sedimentology Commons The ompc lete bibliographic information for this item can be found at http://lib.dr.iastate.edu/ ge_at_pubs/136. For information on how to cite this item, please visit http://lib.dr.iastate.edu/ howtocite.html. This Article is brought to you for free and open access by the Geological and Atmospheric Sciences at Iowa State University Digital Repository. It has been accepted for inclusion in Geological and Atmospheric Sciences Publications by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. Authors Richard M. Iverson, M. E. Reid, Neal R. Iverson, R. G. LaHusen, M. Logan, J. E. Mann, and D. L. Brien This article is available at Iowa State University Digital Repository: http://lib.dr.iastate.edu/ge_at_pubs/136 R EPORTS 19. C. P. McKay, S. C. Martin, C. A. Griffith, R. M. Keller following laboratory simulations (29) in studies of must grow to millimeter sizes to reduce several Icarus 129, 498 (1997). clouds in Earth’s tropics (30). Observations of terres- percent supersaturation to saturation conditions. 20. J. R. Holton, An Introduction to Dynamical Meteorol- trial clouds indicate a value of ␣ϭ0.2 and range of 26. R. M. Goody, Y. L. Yung, Atmospheric Radiation, The- ogy (Academic Press, San Diego, CA, 1992). plume radii of r ϭ 0.6 to 4 km. oretical Basis (Oxford Univ. Press, Oxford, 1989). 21. E. Lellouch et al., Icarus 79, 328 (1989). 25. The presence of rain agrees with the paucity of 27. R. Courtin et al., Icarus 114, 144 (1995). 22. The LFC for supersaturated conditions in the mid to nucleation sites available in Titan’s atmosphere 28. R. E. Samuelson et al., Planet. Space Sci. 45, 959 upper troposphere, as suggested by Courtin et al. (Table 1). In the terrestrial atmosphere, the initial (1997). (27) and Samuelson et al. (28), significantly exceeds size of a raindrop is roughly determined by the 29. H. Stommel, J. Meteorol. 4, 91 (1947). 30. J. Simpson, G. W. Brier, R. H. Simpson, J. Atmos. Sci. those for saturated conditions. At 140% ambient number of drops and the mass needed to lower a 24, 508 (1967). supersaturation, the LFC lies at 13 km altitude for a supersaturated atmosphere back to saturation. The existence of only a few nucleation sites brings 31. O. B. Toon, C. P. McKay, R. Courtin, T. P. Ackerman, parcel of 60% humidity starting at the surface. about few cloud particles and large particle sizes. Icarus 75, 255 (1988). 23. M. Awal, J. L. Lunine, Geophys. Res. Lett. 21, 2491 The only known source of nucleation sites on Titan 32. P. Gierasch, R. Goody, P. Stone, Geophys. Fluid Dyn. 1, (1994). is its haze, the number density of which is several 1 (1970) 24. The entrainment of local air breaks the conservation orders of magnitude smaller than typical terrestrial 33. C.A.G. and J.L.H. are supported by the NASA Plane- ␪ of e for the rising plume. The increasing mass of a values. Considering the low number density of tary Astronomy Program under grant NAG5-6790. 1 dm ␣ nucleation sites compared to the high methane plume parcel, m, is parametrized as ϭ m dz r abundance, Toon et al. (31) realized that particles 18 July 2000; accepted 21 September 2000 various depths and measuring their volumes, Acute Sensitivity of Landslide masses, and water contents (15). No system- atic variations of porosity with depth were Rates to Initial Soil Porosity detected. Our suite of landslide experiments includ- R. M. Iverson,1 M. E. Reid,2 N. R. Iverson,3 R. G. LaHusen,1 ed individual tests with initial porosities rang- M. Logan,1 J. E. Mann,3 D. L. Brien2 ing from 0.39 Ϯ 0.03 to 0.55 Ϯ 0.01 (Ϯ1SD sampling error for an individual experiment). Some landslides move imperceptibly downslope, whereas others accelerate Ancillary tests of the same soil in a ring-shear catastrophically. Experimental landslides triggered by rising pore water pressure device and triaxial cell produced dilative moved at sharply contrasting rates due to small differences in initial porosity. shear failure when initial porosity was Յ0.41 Wet sandy soil with porosity of about 0.5 contracted during slope failure, and contractive shear failure when initial po- partially liquefied, and accelerated within 1 second to speeds over 1 meter per rosity was Ն0.46 (Fig. 2 and Table 1). Land- second. The same soil with porosity of about 0.4 dilated during failure and slides with initial porosities that bracketed the slipped episodically at rates averaging 0.002 meter per second. Repeated slip range from 0.41 to 0.46 were therefore of episodes were induced by gradually rising pore water pressure and were ar- greatest interest. rested by pore dilation and attendant pore pressure decline. Landslide motion was measured with two ground-surface extensometers and 17 or 18 In popular metaphor, landslide processes be- strength (8–10). Positive feedback between subsurface tiltmeters arranged at depth incre- gin spontaneously and gain momentum as frictional strength reduction and soil contrac- ments of ϳ7 cm in two vertical nests (16). they proceed, but what determines how real tion may cause some landslides to transform Pore water pressures were measured with 12 landslides move? Can small differences in into liquefied high-speed flows (11–13). tensiometers and 12 dynamic piezometers ar- initial conditions cause some landslides to To isolate the effect of initial soil porosity ranged in three vertical nests at depth incre- accelerate catastrophically and others to on landslide style and rate, we conducted ments of ϳ20 cm (17) (Fig. 1). Data from creep intermittently downslope? The distinc- large-scale experiments under closely con- each sensor were logged digitally at 20 Hz for tion is important because rapid landslides trolled conditions. In each of nine landslide the duration of each experiment. pose lethal threats, whereas slow landslides experiments, we placed a 65-cm-thick, 6-m3 To induce landsliding, soil prisms were damage property but seldom cause fatalities rectangular prism of loamy sand soil (Table watered with surface sprinklers and through (1). 1) on a planar concrete bed inclined 31° from subsurface channels that introduced simulat- A longstanding hypothesis holds that horizontal and bounded laterally by vertical ed groundwater (Fig. 1). Rising water tables landslide behavior may depend on initial soil concrete walls 2 m apart (Fig. 1). The down- were kept nearly parallel to the impermeable porosity, because soils approach specific crit- slope end of each soil prism was restrained by bed by adjusting discharge from a drain at the ical-state porosities during shear deformation a rigid wall, which ensured that deformation base of the retaining wall. Although prelim- (2–4). Tests on small soil specimens indicate occurred at least partly within the soil mass inary experiments indicated that different that dense soils (initially less porous than (rather than along the bed) and that landslid- styles and rates of water application influ- critical) dilate as they begin to shear, whereas ing included a rotational component. enced the onset of slope failure, this influence loose soils (initially more porous than criti- Different methods of soil placement yield- became negligible as failure occurred and cal) contract (5–7). Dilation can reduce pore ed different initial porosities. The highest instigated changes in soil porosity (18, 19). water pressures and thereby retard continued porosities (Ͼ0.5) were attained by dumping Landslides with differing porosities dis- deformation by increasing normal stresses the soil in 0.5-m3 loads and raking it into played sharply contrasting dynamics (compare and frictional strength at grain contacts, position, without otherwise touching its sur- Figs. 3 and 4). Each of four landslides with whereas contraction can increase pore water face. Lower porosities resulted from placing initial porosities Ͼ0.5 failed abruptly and ac- pressures and thereby reduce frictional the soil in 10-cm layers parallel to the bed celerated within1stospeeds Ͼ1 m/s. The and compacting each layer with either foot surfaces of these landslides appeared fluid and traffic or 16-Hz mechanical vibrations that smooth, and data from dynamic piezometers 1U.S. Geological Survey, 5400 MacArthur Boulevard, delivered impulsive loads of ϳ2 kPa at confirmed that pore water pressures rose rapid- 2 Vancouver, WA 98661, USA. U.S. Geological Survey, depths of 10 cm (14). After placement of ly during failure and reached levels nearly suf- 345 Middlefield Road, Menlo Park, CA 94025, USA. 3Department of Geological and Atmospheric Sciences, each soil prism, we determined porosities by ficient to balance total normal stresses and liq- Iowa State University, Ames, IA 50011, USA. excavating four to nine ϳ1-kg samples at uefy the soil (Fig. 3). Three landslides with www.sciencemag.org SCIENCE VOL 290 20 OCTOBER 2000 513 R EPORTS initial porosities indistinguishable from the crit- 2781 s (Fig. 3), when tiltmeters at all depths zero (atmospheric) values to hydrostatic val- ical porosity (0.44 Ϯ 0.03, 0.44 Ϯ 0.03, and began to rotate slightly upslope and pore ues (ϳ30 cm) as soil contraction forced the 0.42 Ϯ 0.03) displayed inconsistent behavior, water pressure heads below the water table water table upward.