Initiation and propagation of submarine sediment failure 14 July 2009 Schlanger Fellowship presentation at USAC summer meeting Robert Viesca Advisor: Prof.

Initiation and propagation of submarine sediment failure

14 July 2009Schlanger Fellowship presentation

at USAC summer meeting

Robert ViescaAdvisor: Prof. James R. Rice

Harvard University

What have been some remarkable events?

Offshore Morocco, within 200kyr (Talling et al., 2007) debris flow deposit extending 1500km carrying ten-times annual river sediment output to ocean

Papua New Guinea 1999 (e.g., Tappin et al., 2008) submarine landslide implicated in tsunami following earthquakeGulf of Mexico (e.g., Hurricane Ike, 2008) submarine landslides disrupting pipeline connections

(Modified from MBARI bathymetry perspective)

What do submarine landslides look like?

Santa Barbara Basin: Gaviota slide and Goleta landslide complex

~5km~10km

For such shallow slopes, how can these slides exist?

(Total “Know-How” Series, 2007)

I’ll focus on passive margins:

Is the origin from complex 2D fluid flow?

Expedition 308 (Gulf of Mexico): Sedimentation model

Color contour: pore pressure,Line contour: effective stress

0 4P (MPa) (Modified from Behrmann et al. 2006)

012

5 4 30.5

km

5 km

Site U1323

Site U1324

Here, low permeability lens brings high pore pressure near seafloor.

strength there reduced below local shear stress

σ , σ = σ − p (strength ~ effective stress )

total stress pore pressure

Force

Blo

ck s

ize

(sed

imen

t de

pth

)

Fx

Fz T

Under rapid sedimentation, gravity overcomes strength at some depth.

This is possible because strength becomes constant.

In “dry friction” (or submerged & no flow) strength grows with depth.

no flowrapidsedimentation

Or can failure occur in a simple 1D model?

More specifically…

zq

Depth of failure predicted by 1D model

k =koexp(−σ /σ*)

ko

Permeability, k(σ )

Effective stress, σ

With a strongly reducing permeability, trapped fluid creates overpressure in underlying sediments with steady flow q

zx depends on

α =μq

′γ ko

fluid flux

seafloor permeability

fluid viscosity

~sediment weight

0.01–10mm/yr

10–(11–14) m2

σ * perm. stressdependence

′x

Presume some sediment shears:sediment compacts flow barrier.

Flow barrier weakens underlying sediment further shear.

Further shear compacts more sediment more weakening. Weakening (pore pressure increase)

How can movement start & grow?

z

Δp ~ Dμq

k

D

Precisely how does weakening (pore pressure increase) enlarge a shear zone?

Will that shear zone become unstable?

Force weakening with variable curvature K and magnitude W.

Two regions in frictional contact.

Frictional strength decays with slip.

Use fracture mechanics concepts to estimate critical length and loading:

When does movement become unstable?f

fpf = fp−wδ

w1

δ

W =0

W =1

WK

Firstmovement

Total loss of shear strength

′x

Weakening

′x

Cra

ck le

ngth

: a

/ L

W (=Rtfpτo

)

K =10 (=κL2fpτo

)1 5

L =Gτo

fpw

Unstable

lengths

Firstmovement

Total loss of shear strength

τo

σo

slipping region

Potential slip surfaceH

Which direction will instability propagate?

Slight downslope preference for rupture propagation.

y

x'

′x / H

Δσ ( ′x , t)

σ o

Summary

Here we explain how shallow slopes can slide.

If movement begins locally, that localization can grow.

If that grows to a certain size, the slope catastrophically slides.

Catastrophic enlargement proceeds preferentially downslope.