VN u Journal of Science, Larth Sciences 25 (2009) 1-9
H eavy rainíall induced landslides in Bac Kan and Binh Dinh p ro v in ces
Do Minh Duc*
C ollege o f Science, VNU
Rcccivcđ 09 Jaiiuary 2009; rcccivcd in rcviscd form 3 Fcbruary 2009
A b stract. Landslidcs is onc o f the most scvcrc hazard in Vietnani, iimong that a m ajor numbcr o f laiìdslides arc induced by hcavy raiiifall The papcr dcals with hcavy rainfall induccs shallow laiìdslidcs at tlic cxcavatcd slopcs in Bac Kan provincc and at the natural slopcs in Binh Dinh provincc. Landslidcs in Bac Kaii can liappcn whcn Uic amount o f rainíall is 180111111 conlinuously.
Landslides o f natural slopcs in Binh Dinh can only occur whcn raiiifall niakcs tlic slopes almost saturated. This can happcn w h ai thc amount o f raiiifall rcach ovcr I260mm. Landslides induccd by hcavy rainfalỉ also d cp aid on ữie initial water contcnt (i.e. antcccdcnt rainfall), gcological scttings, and plant cover. Thcrcforc rainĩall paUi and critical rainfall can vary in a givcn rangc in a placc, and can vary widcly ữom pỉace to place.
Keynvrds: Hcavy rainfall; Slopc; Landslide; Unsaturatcd soiỉs.
1. I n tr o d u c tio n
L an d slid cs and flo o d in g arc thc m o st scvcrc g eo h azard s in V ietnam . A n n u ally , landslides alo n e c a u s c a d a m a g e o f n c a r l y 100 m illio n s U S D |6 ). T h e serious h a z a rd o ften ta k es placc d u rin g sto rm s o r tro p ical dep ressio n s. Storm s a n d la n d slid es h ad d estro y ed 4 4 8 b rid g es, 789 cu lv erts, stro n g ly d a m a g e d 6 m illio n s m 2 o f h ig hw ay su ría c e , 2 .7 m illio n s m 3 o f so ils and rocks h ad to b c clc a n e d d u rin g th e p e rio d from 1990 to 1995. M ore serio u sly , lan d slid es along w ith d eb ris flow s c a n c a u s e sev ere fatality. A s sh o w in g in tab le 1 th e h aza rd s o f lan d slid es and debris flo w s are v ery sev ere a n d c a n o c c u r in all
m o u n tain o u s a re a s o f V ie tn a m A rem ark able s itu a tio n is th a t all o f th e m on ly o ccu rred d u rin g h c a v y ra in fa ll. T h e o n g o in g clim atc c h an g es h a v e the p o tc n tia l to sig n iíican tly in crcase th c c o u n try ’s ris k ex p o su re to g eo h az ard s. E x trem e w e a th e r ev e n ts in d u c cd by clim ate c h a n g e s are h a p p e n in g m o re frequ ently an d w ith h ig h e r in te n sitie s. H e a v y rain fa ll and c lo u đ b u rsts a re le a d in g to m o re floo ds and lan d slid es. L o c a l liv e lih o o d s ex p erien ce set- backs. T h e ev en ts are cou n terp rcx ỉu ctiv e a n d are threatening th e su stain ed econ om ic developm ent o f V ie tn a m It m ay b e as a co n seq u en c e o f clim aíe c h a n g e la n d slid e s h a v e recently o ccu rred w ith h ig h e r fre q u e n c y a n d intensity.
'T cl.: 84-912042804.
E-mail: ducdm@vnu.edu.vn
2 D M . D u c/ VNU Ịournal of Science, Earth Sàences 25 (2009) 1-9
Tablc 1. Reccnt rccordcd scverc landslides & debris ũow in Vietnam
Location Date Typc No. o f deaths
& missing
Damages
Nam Cuong, Cho Don (Bac 23 Jul. 1986 Dcbris flow & 07 120 ha ricc íields, 20 km o f
Kan province) laiidslidcs roads
Lai Chau tovvn 27 Jun. 1990 Dcbris flow & Over 100 607 houses, 5 bridges, 10km2
Landslide o f town dcmolished
Nani Muc & Muong Lay (Lai 17 Jul. 1994 Debris Flow & 20 -
Chau province) Landslides
Muong Lay town (Lai Chau 17 Aug. 1996 Dcbris flow & 55 The conimuiie had to move to
province) Landslide anothcr placc
HighvvayNo. 27 (Lani Dong lO O ct. 2000 Debris flow & - 37 scvere landslidcs in 55 kin.
provincc) laiidslides 500 m highw ay fully desưoycd
Ialy hydropowcr plant (Kon since 2002 Landslide - causing darnagc o f billions o f
Tum province) VND each ycar
Du Tien & Du Gia, Yen 19 Jul. 2004 Dcbris flow & 48 33 houses, 627 ha ricc fíclds
Minh (Ha Giang province) laiidslides
Sung Hoang (Phin Ngan, ^ at 13 Scp. 2004 Landslidcs 23 4 liouscs dcstroycd Xat, Lao Cai province)
Nghia Lo, Van Chan (Yen 28 Sep. 2005 Flash FIood & 42 Cat Thinh com mun c & Nghia
Bai province) Landslidcs Lo tơn \vere scvcrely damagcd
Bat Xat, Sa Pa, Bao Thang, 09 Aug. 2008 Flash Fỉood & 65 M any com m unes wcrc
Bao Ycn (Lao Cai provincc) Landslidcs dcstroycđ
2. H eav y r a in ía lỉ in d u c e d la n d s lid e s in B ac K a n p ro v in c e
B ac K an - a n o rth e a st m o untainous p ro vince belongs to th e n o rth e a st-ío ld in g reg io n w ith c o n p lc x g co lo g ical settings. T h e geological activities, esp ecially neotectonic m ovem ents h av e c re a te d the v ariab le and com plex to p o g rap h ical ch aracteristics in B ac K an province. In this stu d y , th e C au riv er b asin is m ainly íbcused. It stretch es fro m C ho D on to C h o M oi district (a lo n g natio n al highvvays No.
3 an d N o. 2 5 7 ) in th e le n g th o f 1 0 3 k ra T h e area o f study reg io n is a b o u t llO k m . T h e geo m orpho lo gy in the reg io n is ch aracterized by tw o m ain types: ero siv e re lie f a t hills and m ountains a n d acc u n iu la tiv e re lie f alo n g C au riv er and stream s [2]. M o st o f stud y area is characterized b y ero siv e re lie f w ith the elevation o f 2 0 0 -5 0 0 m an d th e com m o n slope angles o f 35-40°. T h e a re a o f o v er 5 0 0 m h igh is very limited.
In the region, b e d ro ck s a re d iv e rsified w ith 5 ío rm atio n s (P h u N g u - O3- S1 pn , B ac Biui - D ibby M ia Le - D ]/;//, S ong H ien - T i.2? h and Ha C oi - J |.2hc) an d Q u a te m a ry sedim ents.
A m o n g th at, land slid es o fte n o ccu r in the vveathering c ru st an d h ig h ly crack ed bedrocks o f Phu N g u an d M ia Le fo rm atio n s. The w eath erin g criist o n P h u N g u ío rm atio n is v aria b le in tlìickness, ra n g in g fro m 0 to over lO m T he p o p u la r thick ness is 1 .5 -2 .5 m The M ia Le ío rm atio n is c h a ra c te riz e d by clay sericite shale, silty san d sto n e, ch e rty schist, and thin beds o f lim estone. 6 0 0 -7 0 0 m thick. The
\veathering c ru st thickn ess is sm a lle r th an lm , landslides on ly o ccu r alo n g w e a k su ríace s in crack ed rock m asses. T he tecto n ic activities, especially the đ eep -seat ía u lt alo n g highw ay N o. 3 h av e in ten siv ely \v e ak en ed the stren gth o f rock niasses an d ma de in te n siíĩc a tio n o f w eath crin g pro cess. T h e b e d ro c k s w cre strongly cracked cau sin g so m e ro ck slid es w ith slip suríaces o f opposite direction to b ed d in g suríaces.
D .M . D u c / V N U Ịouniaỉ o f Scieĩỉce, Earth Sàences 25 (2009) ỉ -9 3
Tablc 2. Gcotechnical propcrtics o f rcsidual soils in Bac Kan
Properties Value
D60 (nira) 0.46
Dio(mra) 0.01
W ater contcnt(% ) Wct density (KN/m3)
31 18.5
Void ratio 0.759
EtTcctivc aiiglc o f ữiction (dcg.) 28
EíTcctivc cohcsions (KPa) 5.5
Saturatcd cocíĩìcicnt o f permcabiiity (m/s)
2 X 10-6 T h e d atab ase is m a in ly a c h ie v e d from sev eral in vestigation s a lo n g so m e national hig hw ays N o . 3 (fto n i C h o M oi to B ac Kan tow n ), N o. 2 5 7 (B ac K an to w n - C h o D o n) and som e seg m en ts o f N o. 2 5 6 (B ac K an to w n - N a Ri) in 2001 and 2002. Ít co n tain s geolog ical settings o f the reg io n , lcvels o f \veathering, p h y sical m echan ical p ro p e rtie s o f so ils and rocks. T h e m ost im p ortant đ ata is a set o f 72 reco rd ed large land slid es. T h e detail in vestigatio n was c a rrie d o u t for each landslide and th e retriev ed d ata co n tain s th e location, dim ensio ns o f lan dslide, slo p c anglc, ch aracteristics o f so ils a n d ro cks, plant coverage a n d hu m an activ ities affectin g the landslide. D uring th e in v estig atio n , 4 0 d isturbed and 80 u n d istu rb ed sam p les o f resid u al soils and ro ck s w e re a lso tak en fo r íu rth e r analysis in the la b o ra to ry . G e o te c h n ic a l p ro p erties o f soils are s h o w in g in table 2 a n d Fig. 1.
Fig. 1. Soil vvater charactơistic curvc ofresidual soils in Bac Kan (dry curve).
M o n ito rin g đata o f daily rain íall in 2001 are re trie v e d fro m 9 stations. H ow ever, only 3 statio n s, in c lu d in g P h u o n g V ien, D ong Vien a n d B ac K an a re dealt w ith th e cu rren t research.
T h e research w as co n d u ctcđ by u sin g various m eth o d s. T h e rem o te se n sin g & G1S m eth od is a p p lie d to cla ssiíy the areas o f different heights, in clin ed ang les and assu m e the p o ten tial areas o f lan d slid es for íu rth er research. G eological m e th o d s p e rn iit elu cid atin g lithological co m p o sitio n o f rocks, th e ir ab ility o f \veakening đ u e to w eath e rin g and d e íin e th e cracked zones c au sed b y te cto n ic m ov em cn ts. T hen, the site in v e stig atio n w as deployed, including m e asu re m en t o f lan dslid e dim ensions, taking sam p les and fie ld tcst o f soil, ro c k m ass shear stren gth. T h e sam ples, su b scq uently , w ere a n a ly z e d in the la b o rato ry to d e íĩn e physical, m e ch an ic al p ro p erties o f so ils a n d rocks.
A fterw ard s, the ach icv ed d ata in com bination w ith tran sien t slope in filtratio n w ere used for s lo p e sta b ility assessm ent.
2.1. R ainfall in the sto n n N o. 7 (3 a n d 4 Ju ly 2001) B ac K an h as tw o distin ct seasons in te rm o f rainfall. T h e w et seaso n stretch es fro m A pril to O c to b e r o c cu p y in g 8 5 -9 0 % o f the total rainfall.
T h e re m a in in g p e rio d o f tim e (N o v em b er to M arch n e x t y ear) is dry season. T h e am ou nt o f rain fa ll a n d can b e a h u n d red tim es m ore o r less fro m y e a r to y ear. T h e av erag e rain fall in Bac K an is 1400-1800 mnVy. It reduces fro m w est to east an d fro m hig her location s to lovver ones.
L o n g -te rm m o n ito rin g o f ra in ía ll show s that ra in in ten sity is co n ce n trate d in 3 m onths fro m Ju n e to A u g u st, acco u n tin g for 75% o f the total (Fig. 2). H o w ever, rain fall is m ainly con trib uted b y sev eral ra in sto rm events. In a surge storm , r a in in ten sity can reach h u n d red s o f m illim eters p e r day. T h e 3 a n d 4 Ju ly 2001 rain sto rm event, ío c u se d in this stud y, is a typ ical e x a n p le .
T h e ev en t started lately o n 2 July 2001 w h en the su rg e sto rm N o. 2 lan đ ed on to the
4 D .M . D uc / UNU ịournaì o f Science, Eữrth SciiHCỀs 25 (2009) 1-9
m ainland. T he re c o rd e d d ata every 6 h a t 3 stations show th at th e early av erag e rainfall intensity is 2 m m /h fo r the first 18h. lt reached to the hig hest íig u re o f ab o u t 15mnVh a t the early o f 4 Ju ly (Fig. 3). T h e event w as then sto p p ed at the en d o f 4 July. F inally, the rain intensity during sto rm w as 2 6 1 .1 , 191.1 and 133.7 m m a t B ac K an, D o n g V ien, an d P huong V ien stations, respectively. T h e rainfall intensity w as the h ig h e st daily rain íall since 1996. In the s to rm N o . 2 in 20 0 1 , th ere w ere 11 large landslides w ith th e to tal v o lu m e o f about 16,000 m \ T h ese oncs lcd to th e in terru p ts in th e national h ig h w ay N o. 3 and m any other national highw ays for a long tim e. T he co st o f dam age w as over sev eral b illio n V N D . A lo n g the road No. 3, N o. 2 5 6 an d N o. 257 ov er 30 landslides w ere re c o rd e d to b e o ccu rred in this surge storm.
Fig. 2. Monthly rainfall Ũ1 Bac Kan province.
2.2. L a n d slid e P ro p erty
L andslides in B ac K an p ro v in c e are often occurred at th e slopes o f n atio n al highw ays, and only take place in ra in y seaso n s. T h ey can occu r either in the vveathering c ru st o r in the highly crack ed ro ck m asses. T h e slip surfaces h av e arc shape in the w e a th e rin g c ru st (Fig. 4a) and they co incide \vith b e d d in g su ríace s in high w eathered ro ck m asses. T he land slides w ith large volum es o ften h a v e c o n p le x slip su ríaces including both arc and íla t b ed d in g slip su ríaces (Fig. 4b). T h e vo lu m es ran g e fro m tens m 3 to
over 5 ,0 0 0 m \ am o n g th at th e v o lu m es o f 100- 50 0 m 3 are dom inant. S om e typical lan dslides in the s to rm N o . 2 are sh ow n in tab le 1. T h e 72 reco rd ed land slid es d istrib u te m ain ly in 7 regions (3 regions in C h o D o n d isứ ict, 2 in Bac K an tow n a n d 2 in C h o M oi district). T h e site in v estig atio n in d icate d th at th e lan d slid es only occu rred w h ere b e d ro ck s s tro n g ly in ílu en ce d b y ía u lts o r th e d irectio n o f b e d d in g surface coincides w ith in c lin cd d ire c tio n o f slo p es; the slopes are co n stitu ted b y resid u al so ils o r highly crack ed rocks; the th ick n ess o f \v eath erin g crust o f over 5 m u su ally m ct a t land slid es over 5 0 0 m 3. T he h ig h d en sity o f v e g e ta b le co v erag e is taken p la c e a t alm ost landslides. T herefore, the role o f vegetable co verage in slope instability is n o t sig n iíĩcan t. T he p rio r research has defĩned 4 zones o f d ifferent p o ssib ility o f landslide, w h ich are very higli, high, m e d iu m a n d low possibility , resp ectiv ely [4].
a) Rain fall intcnsity
Fig. 3. Intcnsity and accumulation oftlic rainstorm [3].
0 6 12 B 2 4 3 0 3 6 4 2 4 8 &) 6 0 66 E l a p s a d tim ® ( h. s lr v c e 1AM 2 J u l y )
D .M . Dỉtc / V N U Ịounm ỉ o f Science, Earth Sciences 25 (2009) 1-9 5
In sh o rt, the storm in th e 3 & 4 Ju ly 2001 c a u se d th e highest d a ily ra in ía ll in B ac Kan p ro v in ce d u rin g the p criod 1996-2001. It trig g e re d a b o u t 30 la n d slid es a lo n g the national hig hw ays N o .3, 256 and 2 57. Shallovv landslidcs can o ccu r w hen the rainỉall am ount is higlìer than
180m m and PW P is still in n eg ativ e v alu es [3].
a) Arc slip suríacc
h) Complex slip suríacc Fig. 4. Slip surfaces o f landslides.
3. H e a v y r a iiiĩa ll ỉnduccc! la n d s lid e s in B ỉnh D in h p r o v in c c
H eavy rain ỉall in B inh D inh p ro v in ce from 12 D ecem b er to 15 D ecem b er 2005 caused serio u s d am ag es. Lancislides cau sed traffic ja in s a t niany roads, 03 p erso n s vvere killed by ílo o d in g . A v a st la n d slid es o ccu rrcd at a m o u n tain o f C an h Lien co m m u n e, V an Canh d istrict, killed 4 hulls, íllle d up so m e rice field.
P o rtu n ately , th e re vvas n o t a n y fatality . N early the sam e at th e o p p o sitc site o f th e m ountain th erc w as a lso a series o f larg e landslides.
E sp ecially , th e la n d slid es o ccu rred w ith several loud e x p lo sio n s. T h e fact h a s lcad to serious n erv o u sn ess o f the re sid e n t population.
L an d slid es to o k placc a t th e n atu ral slo pes vvith slope a n g le s o f 28-31°. T h e v eg ctab le c o v e r vvas very loose.
T h e g eo lo g ical settin g s are very co m p licated . T h e reg io n d istrib u tes 3 fo rm atio n s sucli as X a L am C o (A R x /c ), M ang Y a n g (T 2iwy). and Ọ u atern a ry (Ọ ). M o st o f the a re a is co n stitu te d by ig n e o u s ro ck s th a t belong to 4 co m p lex es, in c lu d in g V an C anh (G - SG /T2v c ị C h a v a l (G b T 3 n cv ), D eo C a (G /Kd c ị C u M o n g (G b/E tvn). T h e fault system esp ecially th e se m i-lo n g itu d e fault leads to m any c ra c k in g b lo ck s o f th e bed rock s. T he h etero g en eo u s d istrib u tio n o f cra c k in g sy stem h as lead to th e diíTerent th ick n ess o f w eath erin g c ru st and lìiakes p o tetial slip su ría c e for large landlsides.
Table 3. Geotechnical propcrties ofresidual soils in Binh Dinh
Properties Value
Water content (%) Wet density (K N /nv)
19 18.7
Void ratio 0.75
Effective angle o f ữiction (deg.) 28.9
Effective coliesions (KPa) 15
Saturated cocfficicnt o f pcrmeability (m/s) (samples taken at 0.3-0.5m dccp)
5 X 10‘5
6 D .M . D uc ỉ V N U Ịoum al o f Science, Earth Sciences 25 (2009) 2-9
T he d atab ase is m a in ly ach iev ed from the inv estigation in V an C a n h d istrict fro m A ugust 2006 to June 2007. It contains geological settings o f th e reg io n , ch aracteristics o f w eathering, p h y sical m e ch an ie al p ro p erties o f soils an d rocks. T he d etail in v estig atio n w as carried o u t a t 05 large landslides. D uring the investigation, 60 d istu rb e d an d 5 0 un disturbed sam ples o f so ils and ro ck s w ere tak en for further analysis in the lab o ra to ry (T ab le 3).
M onitoring d ata o f d aily rain fa ll in 2005 is retrieved fro m th e V an C an h statio n vvhich is 20 k m aw ay fro m the slid in g site. In addition, the data o f m eteorolog y a n d ra in ía ll fro m 1977 to 2003 w as also tak en in to account.
3.1. R a in fa ll in the p e r io d o f S ep tem b e r - D ecem ber 2005
A s m en tio n ed ab o v e a series o f large landslides o ccu rred in th e m id d le o f D ecem ber 2005. So as to e lu cid ate a rela tio n betvveen
ra in ía ll a n d this p h en o m en o n th e ra in d ata ỉro m S e p te m b e r to D ec e m b e r 2005 is tak en into a c c o u n t. T h e to ta l rainfall in Sept. 200 5 was 2 8 7 .3 m m , a n o rm a l fig u re in c o m p a riso n to o th e r y e a rs. In O ct. 2005, the íĩg u re w as 1 01 6.8 m m w h ich is larg er th a n the n o rm alize d ra in fa ll w ith th e freq u en cy o f 5% (1015m m ).
T h e re ío re th e ra in ía ll in O ct. 2005 w as abo ve th e íìg u re o c c u rre d o n ce e v e ry 20 years. T h e n u m b e r o f ra in y days in this m o n th w as 20 days co n tin u o u sly . E specially, th e rain fa ll in 3 days (2 3 -2 5 O c t.) re a c h e d to the to tal o f 5 6 6 .5m m (F ig. 5). H o w e v e r no larg e la n d slid es occuiTed d u rin g th a t m o n th . T h e ra in ía ll in N ov. and D ec. 2 0 0 5 w a s 6 27 .6 an d 829. Omni, re sp e c tiv e ly (F ig. 6). T he ra in ía ll in N ov. 2005 is also a n o rm al fact. B ut the one o f Dec. 2005 is v ery re m a rk a b le . It is a íĩg u re th at can only m e e t o n c e ev ery 5 0 years. T h e re ío re all o f the la rg e laiìd slid es o c c u rre d in a m o n th vvith a very h ig h a m o u n t o f raỉn íall.
1
1
I __________________________ _________________________ L
■ ^ H r i
l l 1 1 1 t f f r i i ỉ i i i m i i i n m i Ẩ J r J r J J l í L ẻ i Ì i r o m Ì u U
15 22 29 36 43 5 0 57 64 71 78 85 9 2 9 9 106 113 120 E la p s c d đ ay s fnm» 01 S c p te in b c r 2005
Fig 5. Rainĩall from 01 Scptember 2005 to 31 Deccmber 2005.
Fig 6. Monthly and accumulativc rainĩall.
D .M . D u c / V N U Ịournal o f Sciaice, Earth Sciences 25 (2009) 2-9 7 T h e ab o v e results sh o w th a t the landslides
o ccu rred after 3 m o n th s o f h e a v y ra in from S ept. to Dcc. 2005. T h e to tal ra in ía ll o f onc m o n th b c fo re the ía ilu re (1 7 N o v . 2 0 0 5 - 15 D ec. 2 0 0 5 ) was o v er 1 2 6 0 m n i T h e average d a ily rain fall was 4 3 .4 mnVday.
3.2. L a n d sỉid e p r o p e r tie s
T h e largest la n d slid e to o k p la c e in Lang C h o m w ith th e v o lu m e o f a b o u t 2 2 ,0 0 0 m 3 (Fig.
7). T h e landslide o c c u rre d w ith a serious exp lo sio n. Ít m ight d e riv e fro m th e reliev e o f co m p re ssiv c p ressu re in the slo pe. O th c r sounds w ere ind uced by the m o v e m e n t o f so il a n d rock m asses afte r sliding. T h e to p s lid in g su rface situates in the w e a th e rin g c riist h o w e v e r m ost o f slid in g su ría c c is the in te ría c e o f intact igneous rock s. R esid u al so ils a re w eath ered fro m ro ck s o f D eo C a c o m p le x \vith the th ickness o f 4 -6 m T h e slo p e is 65 h ig h , and slope a n g le o f 27ờ - 32°. L o ts o f ro c k m asses o f ab o u t 10m 3 vvere tran sp o rte d d o w n s lo p e along the d ista n ce o f h u n d red s m e te rs . T h e sliding m aterials then d e stro y e d a lo c a l ro a d segm ent and íille d up th e L a u s tre a m n e a r the slope cau sin g a n in crease o f 2 -3 m o f th c s tr e a m w ater level. F ortunately, th e re w as n o d e b ris flo w due this phen om en on .
A t th e sam c tim c in K a B u n g (th e o ppo site site o f th e m ountain) th e re w as a series o f large landslides (Fig. 8). T h e th ic k n e ss o f residual soils is 6 - 9 m T h e s lo p e a n g le is 27° - 32°. T h e av erag e vo lum e o f th e se la n d slid e s is 1 0 ,5 0 0 m \ T he la n d slid es tcx)k p la c e fa r fro m re s id e n t area and d id n o t cau se any íatality .
T h e in itial resu lts s h o w th a t th e natural slopes in B inh D in h c a n b e in s ta b le w h e n the rainfall in a p e rio d o f tim e re a c h in g over 1260m m , an d thc h e a v y ra in fa ll o f 2 0 0 -3 0 0 m m in 2-3 d ay s occur at the e n d o f th is p e rio d . T he slopes w ith the in c lin e d a n g le o f 30° o r m ore steep h a v e vcry h ig h p o te n tia l o f s lid in g in such situation.
Fig. 7. Laiidslidc in Lang Chom.
Fig. 8. Landslides in Ka Bung.
4. D iscu ssio n s
H eav y rain íall th e re fo rc c a n trigger landslides at b o th ex cav ated and n o n -excavated (n atu ral) slopes. E m p irically , the critica l rain íall is 1 8 0 m m co n tin u o u sly in B ac K an province.
T his can m eet d u rin g th e ex trem e clim ate events su ch as storm s an d trop ical dqDressions [5]. T h e tim e o f rain in g is ab o u t 2-3 days w ith the hig h est in ten sity o f 15 mnVh. T h e threshold herein is co n sid ered as a d aily rain íall. T he ĩig u re is v e ry lo w in co m p ariso n to the situation in som e p la ces su ch as H o n g K o n g (75m m /h) [1] o r S in g ap o re (8 0 m m /h ) [7]. H cavy rain íall
8 D .M . D u c / V N U Ịournal o f Science, Earth Sãences 25 (2009) 1-9
also induces m a n y sh allo w land slid es along the roads in B in h D inh p rov in ee. Hovvever the thresh old can n o t estab lish b eca u se o f lim ited data.
R egarding the ex ca v ated slopes in B ac Kan, along w ith the ro le o f h eav y rain fa ll one reason sho uld be em p h asized is th e slo p e angle. The slopes herein are stab le w hen th e slo p e angles a re lovvcr than 35° [4]. B ut all o f slo p e angles a re m ore than 45°, an d m o st o f th e m o ver 60°.
S urprisingly, the fa c t is co m m o n n o t on ly a t the highvvays b u t also a t the resid en tial sites. This derives fro m th e effort to red u ce excavated volum e. T h a t leads to a quite lo w critical rainíall b eca u se the lan dslid es c an o ccu r w hen the po re w ater p ressiư e (P W P ) in the slopes is still negative.
T h e initial satu rated d eg ree (S) o f residual soils in slopes plays an im p o rtan t role. A s see in the Fig. 1 the residu al su ctio n o f soils in Bac K an p rov ince is abo ut 8 0 K P a an d a dry soil slope can on ly increase th e P W P to ab o u t -60 K Pa under the rain fa ll o f 1 8 0 m m M eanw hile the slopes can b e instable w hen the average PW P is -6 K P a [3]. Hovvever ra in can occu r every m onth a t the sam e tim e th e tem p eratu re is low in the dry (w in ter) seaso n in B ac K an province. T h ere ío re the satu rated deg ree o f soils is ahvays high. T he sam ples taken in D ecem ber have the av erag e
s
o f 0.75 -0 .8. A nd the initial P W P interpo lated fro m Fig. 1 is about -45 KPa. In the con trary , th e rain fa ll is very sm all and the ev ap o ratio n is re m a rk a b le in B inh D inh province. T hes
o f soils is only 0.3 -0 .4 in dry season. T he h eav y rainfall a t the b eg in n in ^ o f w et season c an o nly increase the \vater co ntent b u t c an not m ake the so ils m ostly saturated. In fact, the rain fall in O cto b er 2005 w as larger than in D ecem b er 20 05 b u t the landslides d id n o t o ccu r in O ctober. A n interesting fact is th at the effectiv e in tem al angle o f soils is a b o u t 2 8 - 3 1°(28.9° in average),an d th e slo p e an g les o f la n d slid es vvere a b o u t 28-30°. So the lan d slid es w ere trig g e re d w h en th e slo p es w ere alm o st saturated.
T h e in ílu e n c e o f g eo lo g ical settin g s is also rem arkable b e s id e tw o reaso n s o f slo p e an g le an d h eav y ra in ía ll. T h e b e d rocks o f sc h ist and clay shale in B a c K an p la y th e ro le o f an im perm eable la y er b e lo w resid u al soils (Fig.
4b). W hen rain y vvater p en etrate in to the slo p e s an d reach to th a t layer it w íll b e a c c u m u la te d an d m ake a flo w d o w n slope. T h is leads to m o re ra p id in c re ase o f PW P in th e slop es. In B inh D inh p ro v in c e this layer is ta k en p la c e by so m e intact ig neous ro cks (F ig.9).
Fig. 9. Slip surface at ứie interface o f intact bcdrock in Lang Chom, Binh Dinh.
T he resu lts g iv e a re c o m m e n d a tio n fo r the early vvaming o f la n d slid es. In Bac K an p ro v in cc th e m o n ito rin g d a ta o f 6-hourly rain íall c an b e m a in ly used. T h e h o u rly d ata is also n ece ssary b e c a u se the in ten sity som etim es reaches ov er 15m m /h. T h e Nvaming sy stem has 3 levels for ex am p le level 1 - 125m m , level 2 - 150m m a n d lev el 3 - 1 7 5 m m T h e re ío re the
P l l l l i lặ É tm P
D .M . D u c / V N U lo u m a l ủ f Sciaice, Earth Sciences 25 (2009) 2-9 9
local au th o rity and resid ents can arrange tim e to rem o v e an d sav e th c ir property. F or natiư al slopes in B inh D inh pro vince th e daily rainfall is u sed and w hen the rain íall can reac h to the th resh o ld an early w a m in g m u st establish for the residen ts to go far avvay fro m the slopes w ith th e slo p e angles o f ovcr 30°.
5. C o n clu sio n s
1. H ea\ 7 ra in ía ll can cau se severe landslides at the ex cav ated o r n atu ral (non- excavated) slopes. H eav y rainfall induces shallow land slides o f sm all volum es a t the ex cav ated slopes. In natural slopes the volum e o f landslides induced b y h eavy ra in ía ll can be large o r very large.
2. L andslides in B ac K an c an happ cn w hen the am ount o f rainfall is 180m m continuoiisly.
PW P in the slopes is still n eg ativ e w hen the íailure occurs. T his fact can be early w am in g by the av ailable rain m o n ito rin g station s in the region. L andslides o f n atu ra l slopes in Binh Dinh can only occu r w h en rain íall niakes the slopes alm o st saturated.
3. L andslides depend on th e initial vvater content (i.e. an teced en t rain fall), geological settings, veg etab le cover. T h erc fo re rainfall path an d critical ra in ía ll in d u ce slo p e failure in a place c an v ary in a given range.
A c k n o w le d g e m e n ts
The p a p e r is su p p o rtcd by the V ietnam N ational U n iv ersity , H an o i’s Spccial Prọịcct cod ed Q G -08-16.
R e íe re n c e s
[1] H. Chen, C.F. Lee, Geohazarđs o f Slope Mass Movement and its Prcvention in Hong Kong, Enginceriĩỉg Geology 76 (2004) 3-25.
[2] N.M. Dong, H.N. Duong, L Minh, Report on Environmeìit o f Bac. Kan Province, 2000. (in Vietnamse)
[3] D.M. Duc, Latìdslides Triggered by the 3 and 4 Juìy 2001 Rainstorm Evení in Backan Province, Northeast Vietìiam, Hanoi GcoEngineering 2005, Hanoi, 25 November 2005.
[4] D.M. Duc, Numericaỉ Simulation fo r Slope Stability Analysis and Lamislide Prediction in Bac Kan Province, Northeast Vietnam, International Symposium on GeoInformatics for Spatial-Inírastructure Development in Earth and Allied Sciences, Hanoi, Victnam, September 2004.
[5] N.T. Huong (cditor in chieO* Characterừtỉcs o f MíỊteorology and Hydrology in Binh Dinh Province, 2004. (in Vietnamese)
[6] D.M. Tam, Flooding and Landslides at the Hìghways o f Vietnam, Inter. Workshop on
“Saving Our Water and Protecting Our Land”.
Hanoi, 20-22 Oct. 2001.
[7] H. Rahardjo, K.K. Aung, E C . H. Lcong, R.B.
Rczaur, Charactcristics o f residual soils in Singapore as íbrmed by vveathenng, Engineering Geològy 73 (2004) 157-169.
