Challenges to developing resilience to post-earthquake landslides

T.C. Hales

Xuanmei Fan, Oliver Francis, Brian MacGillivray, Guillem Domenech, Jing Ran, Dan Hobley, Yi Gong

Beichuan, Sichuan, China

May 12, 2008- Earthquake shaking

12 min post-earthquake-landsliding

June 10, 2008-landslide dam collapse

Sept 24, 2008- 1st post earthquake debris flow

Hazard

con

centratin

g from

a wid

e

spatial area to

river valleys at d

ecreasin

g frequ

en

cy

The 7.8 Wenchuan Earthquake

What are the spatial and temporal scales that are appropriate for enhancing resilience?

Town A

Town B

Administrative boundary

Town C

Research Question • How do we develop

resilience to post-earthquake debris flow hazards? • How does hazard vary

with time?

• At what spatial and temporal scale does resilience develop?

Post earthquake debris flow - 2010

2010 Qingping Debris Flow

100

10

1

0.1

Mea

n r

ain

fall

inte

nsi

ty (

mm

/h)

1 10 100

Rainfall duration (h)

Rainfall triggering DF 2008

Rainfall triggering DF 2009-2013

Threshold rainfall for DF triggering increases with time

Guo et al. (2016) Geomorphology

Rainfall conditions under which debris flows are

triggered

100

10

1

0.1

Mea

n r

ain

fall

inte

nsi

ty (

mm

/h)

1 10 100

Rainfall duration (h)

Rainfall without DF 2008-2013

Rainfall triggering DF 2008

Rainfall triggering DF 2009-2013

Guo et al. (2016) Geomorphology

Variable rainfall thresholds suggest that changes in

material affect debris flow triggering through time

y = 119.83x0.4143 R² = 0.1782

1000

10000

100000

1000000

10000000

10000 100000 1000000 10000000 100000000

De

bri

s Fl

ow

Vo

lum

e (

m3)

Coseismic Landslide Volume (m3)

Poor correlation between debris flows and availability of debris

Study area: Mianjiang River

2008

2014

We mapped the “activity” levels of coseismic landslides since 2008

Total coseismic debris volume = 1.1 km3

Volume remaining in 2011 = 1.0 km3

Topographic drivers of debris flows

Den

sity

0 10 20 30 40 50

Slope (degrees)

“Active” landslides

“Inactive” landslides

No difference in the slopes of active and inactive slides

Den

sity

0 2 4 6 8

Drainage area (log m2)

Active landslides found in topographically wetter locations

From Richard Iverson

Why is there so much uncertainty in understanding where debris flows initiate?

What are the spatial and temporal scales that are appropriate for enhancing resilience?

Town A

Town B

Administrative boundary

Town C

Rel

ativ

e H

azar

d

Res

ilien

ce

Rel

ativ

e H

azar

d

Res

ilien

ce

Town A & C

Time

Shaking and coseismic landsliding

Shaking and post-earthquake debris flows

EAR

THQ

UA

KE

Town B

Modifying effect of social vulnerability , social capital etc.

• Despite an apparent reduction in debris flow hazard through time, most of the coseismic debris that creates debris flows remains in catchments.

• Simple metrics of debris flow hazard are elusive, as hazard depends on the ability of a flow to bulk by entraining wet sediment.

• The earthquake hazard chain leads to heightened susceptibility along river channels potentially leading to a resilience chain.

Spatial measurements of resilience

Changes in post-earthquake population growth rates