[LANDSLIDE] 9-AM

Realtime Monitoring of Shallow-landslide Potential Area

Using Multi-parameter Radar

Ryohei MISUMI

National Research Institute for Earth Science and Disaster Prevention

Abstract

In this lecture, we introduce recent rainfall-induced disasters in Japan and our

researches for reducing them with the use of the multi-parameter radar system. As for

the heavy rainfall in Japan, precipitation extremely concentrates on narrow regions by

peculiar behavior of the convective clouds that is called "back-building". Forecasting of

such phenomena is still difficult at present, thus real-time monitoring of rainfall area

with meteorological radars and providing information are important for disaster

reduction. In National Research Institute for Earth Science and Disaster Prevention, the

multi-parameter radar system, which is able to estimate rainfall accurately with 500m

mesh, is operated to detect heavy rainfall area in real time and provide information on

the WEB. This technology is expected to reduce disasters caused by urban floods and

rainfall-induced landslides.

9-AM

1

REALTIME MONITORING OF SHALLOW-

LANDSLIDE POTENTIAL AREA USING

MULTI-PARAMETER RADAR

Ryohei Misumi

National Research Institute for Earth Science and Disaster Prevention, Japan

Recent rainfall disasters in Japan

9-AM

2

"Back-building" of convective cells

Formation

of new cell

"Organized" storm

Repeat

Dissipation

of cellHeavy rainfall

Loss of lives

476mm

9-AM

3

Shallow landslide（Tochio City）

Debris flow（Izumozaki Town）

9-AM

4

Typhoon Man-Yi on 14-15 July 2007

Isumi City

9-AM

5

Landslide（Isumi City）

A car carried by debris flow（Isumi city）

9-AM

6

防災白書より

WeightWeightWeightWeightBed rock

Soil layer

Slip surface

Sliding force

Shear resistance

Normal situationNormal situationNormal situationNormal situation

What is shallow

landslide ?

9-AM

7

Bed rock

Soil layer

Saturated ZoneSaturated ZoneSaturated ZoneSaturated Zone

Soil

particlesWater During heavy During heavy During heavy During heavy

rainfallrainfallrainfallrainfall

Increase of pore pressure

Decrease of normal stress

Rain

Why do rainfall induced disaster occur ?

�Problems in rainfall forecasting

�Difficulty in making appropriate warning

�Problems of "aging society" etc.

Monitoring heavy rainfall using

Multi-parameter radar

9-AM

8

Ordinary weather radars

Use of reflectivity

・・・・Large errors for strong rainfall・・・・Hard to distinguish between ground echo and rain

・・・・Strong attenuation

Multi-parameter radar

Use of ““““phase shift””””

・・・・Small errors for strong rainfall

・・・・Possible to estimate rainfall over

mountains

・・・・Not affected by attenuation

・・・・Small errors for strong rainfall

・・・・Possible to estimate rainfall over

mountains

・・・・Not affected by attenuation

Polarized waves

9-AM

9

RaindropVertically polarized wave

Horizontally polarized wave

Because the shape of raindrop is flat, some phase shift is

made between the back-scattered radio waves

9-AM

10

Estimation of rainfall intensity using the Estimation of rainfall intensity using the Estimation of rainfall intensity using the Estimation of rainfall intensity using the

specific differential phase (Kspecific differential phase (Kspecific differential phase (Kspecific differential phase (KDPDPDPDP) ) ) )

� not affected by rain attenuation

� immune to beam blockage

� less sensitive to beam filling and drop size distribution

＋＋

Z KDP

Disdrometer

+

R-KDP

R-Z

R-Z (corrected)

9-AM

11

MP-X radar

JMA radar (corrected with rain-gauges)JMA radar

Rainfall on 15 July 2006 (with hail）

青梅：22mm

13:23＠Tokyo

相模湖：0

府中：20

世田谷：32

X-band radar network (X-NET)

NUED1 NIED2 (2008-)

(To be installed) Chuo Univ

NDA

9-AM

12

Estimation of landslide area using radar-derived rainfall

"Effective" rainfall

⋅⋅⋅⋅+×+×+= 22110E RRRR αα

T/nn )

2

1(=α

T: Half-life period

(like radioactive decay)

t=0 t=-1 t=-2

0

10

20

30

40

50

60

70

0:00 3:00 6:00 9:00 12:00 15:00 18:00 21:00 0:00

Time (JST)

Rai

nfa

ll (m

m)

Rainfall

T=1.5 hours

T=72 hours

Strategy for landslide prediction

A realtime monitoring system for landslide

Radar-derived rainfall Observation at

slopes

WarningSoil moisture and slope movement

9-AM

13

6 hour rainfall (0000JST-0600JST)

Miura

Raingauge

MP-X

Slope failure (shallow landslide) at 0530 JST

9-AM

14

Effective rainfall (T=72 hours)

at 0500 JST on 12 September 2007

Slope failure

Deep-seated landslide

Effective rainfall (T=1.5 hours)

at 0500 JST on 12 September 2007

Slope failure

Deep-seated landslide

9-AM

15

Landslide Prediction Support System (LAPSUS)

http://lapsus.bosai.go.jp