A nalyse the fluctuation and water level trend in Saigon - Dong Nai river system

VVith the analyzed data series from 1980 to 2007, the resuỉls show thai after supplementing the non-tidal elements, the water leveỉ simulation quality is increased, the wi[r]

VNU Ji>umal of Science, Earth Sciences 25 (2009) 223-231

A nalyse the íluctuation and w ater level trend

in Saigon - Dong Nai river system

Luong Van Viet*, Pham Dang Manh Hong Luan, Le A nh Tuan

Sub-Institute o f H ydro- M etcoroỉogy an d Environm ent o f S o u th Vietnam

Received 17 November 2009

A bstract Beside the eíTcct o f gravitational íòrces, the vvater level in Sai Gon-Dong Nai River

System is also aíĩected by non gravitational ĩorces such a wind, air pressure, rainfall and sea ỉevel

risc The purpose o f this papcr is to simulate the vvater ỉevel o f these river based on the tidal and

non-tidal constituents, from that assess the roỉe o f non-tidaỉ constituents to w ater variation In this

study, the none-tidal data used includes the wind fíeld at surface, sea ỉevel prcssure and the rainíaỉl

in Sai Gon - Dong Nai Basin VVith the analyzed data series from 1980 to 2007, the resuỉls show

thai after supplementing the non-tidal elements, the water leveỉ simulation quality is increased, the

wind and rainíaỉl have an important role lo water ỉevel The results also show that near river

mouths, the vvater level raised 13 cm from 1980-2007 duc to the global clim ate change

Keywords: tidaỉ harmonic analysis, non-tidal constituents, sea lcvel rise

1 C e n e r a liz a tio n

T h e D ong N ai - Sai G o n riv er system

vvhosc w atersh ed area is ap p ro x im ately 45,0 00

k m is the seco n d largest riv er sy stem in the

S o u th ern p ro v in ces T h ese riv er system

c o m p rise s o f m a in riv ers in clu d in g La N ga, Be,

D o n g N ai, Sai G o n , N ha Be and rivers ending

u p in G anh R ai bay It is th e w atershed

c o v e rin g m ain eco n o m ic zo n e in the South

w h ich has m a n y activ ities relaíed to w ater

reso u rces, th e d o w n sư e a m o f th e area is low

an d v u ln e ra b le to sea level rise eíĩe c ts

T h e re ío re , sh o rt - term and lo ng - term forecast

o f w ater level in this area is a sig n iíìcan t

d em an d

• -Corresponding author Tel.: 84-4-38275971

E-mail: lgviet@yahoo.com

T id e reg im e o f riv e r sy stem in the research ed a re a is se m i-d iu m a l ty p e w ith tide

a m p litu d e p re tty h ig h A p a rt fro m th e eíTect o f tid e, w ater level v a ria tio n in th e area is also aíTected by ra in fa ll, w in d íìeld and a ir pressure

T h e y e a rly a v e ra g e rainfall o f this vvatershed is 2 0 2 8 m m an d co ncen trates m ainly from M ay to N o v e m b e r w ith ab ove 90% [1]

T h e rain fall w h ich is n o t e q u ally distrib u ted and

m a in ly c o n c e n tra te s in rain y season caused sig n iíìc a n t e íĩe c t o n w a te r level In addition, dry season a n d lo n g - lastin g h eav y rain s are also íacto rs th a t g re a tly affect w a te r level variatio n

T h e effect o f w in d o n w ater level rise in the area c an b e c le a rly seen in d ry season In this

p eriod, th e vvind a t th e sea su rface has m ain

d irection from E a st to S o u th East D ue to the

224 L.V Viet et al. / VN U Ịournaỉ o f Science, Earth Sciences 25 (2009) 223-231

land - sea interactio n, th ere is sh arp d aily

ch ang e o f d aily w in d directio n F rom 5 am to

12 am , the p rev alen t w in d d irection is usually

betw een E ast and E ast N o rth E ast w h ich n early

m atches th e vvind d irection on th e sea From 2

pm to 2 am , d u e to th e e íĩe c t o f sea - land

w ind, the p rev alen t w ind d irectio n is S outh E ast

[1] T his vvind d irection m atches the directio n

o f riv er m o u th s in G anh Rai bay, co n trib u tin g

to th e tran sp o rtin g o f salinity d eep in the land as

vvell as raisin g w a te r level C o m p ared w ith

daily av erag e w in d sp eed , the S outh East w ind

speed can in crease from 1.5 to 2 tim es [2] A t

areas near th e sea, the w ater level can in crease

sig n iíìca n tly w hcn th is w in d zo n e is strong

D ue to the íre q u e n t a p p e a ra n c e o f w ind

w ith directio n froirì N E to S E in dry season s (fro m the end o f O c to b e r an d la stin g till the end

o f A pril), the w a te r level rises and salinity

in tru des to the land, w h ich c an b rin g on the

e íĩe c ts on ag ric u ltu re p ro d u ctio n so th at the vvind is also n a m e đ “C h u o n g W ind ” in tradition Hovvever, th e m o st sig n ific a n t w ind is still S E w ind b eca u se o f its h ig h v elo cily and

d u e to the fact th at it run s p arallcl to th e river

m ouths A parl from th e e d g e o f A sia

C o n tin e n ta l cold h ig h p ressu re sy stem , C h u o n g

W ind is also o rig in a te d from b rccz e an d ed g e o f

N orlh P aciiìc O cean h ig h p ressu re system

2-3m/s 3-4m/s 5-€m/s >6m/s

f l ị z

0% 10%

I—I—I—I—I—l—

50%

Fig 1 Wind rose o f February and July, Tan Son Nhat station

B eside th e e ffe c t o f tid e - fo rm in g forces,

w in d an d th e rain íall in th e area, th ere is also

hu g e im pact o f sea level rise due to clim ate

ch ang e on w a te r level T h e an aly sis o f the roles

o f n on-tidal factors o n w ater level is stro ng ly

n ecessary in w ater level íbrecasting

2 M eỉh o d o lo g y and d ata used

2 1 M ethodology

T h e an aly sis o f w ater level A uctuation and

trend is bascd on h am io n ic an aly zin g m cthod

su p p lem en ted w ith no n-tidal elem ents W ater

level at a certain tim e is co n sid cre d th e su m o f tidal vvater level an d the affected p art o f n o n - tidal elem en ts:

In w hich z , is th e sim ulatcd w ater level at

tim e t, z, is the tidal w ater level, z \ is v ariation

part o f the w ater lev el d u c to non-tidal factors

T h e tidal vvater level z t , a c c o rd in g to

harm o nic a n a ly z in g m cth o d , was:

n

z , = z n + c o s ( + ( K + u ) i - g i ) (2)

i - 1 n

F c b r u a r y

L.v Vict ct ứ/ / V N U Ịournưl o f Science, Earth Sciences 25 (2009) 223-23ĩ 225

H e re , Zois U c c o n s ta n t; H „ q „ g„ fh (Vo+u)i

are am p litu d e, >peed and initial phase angle,

am p litu d e s c a ln g íacto rs and phase ang le

o ffse ts o f ilh co n ititu en t, w ith i = 1, 2, n and

n is th e n u n ib e ro f con stitucnts T h e values o f f,

an d (V 0 + u), depend on tim e and can be look up

fro m ta b le o r calcu lated by astron om ical

functions

In th is study, co m p o n en t z \ includes vvater

level v ariatio n by w in d ( z * ), air p ressure

( z ,p ), rainfall (2 * ) in the w atershed and the

tren d o f w ater level rise due to g lobal clim ate

ch a n g e ( z ] r ):

7 1 = 7 R + 7 * * + 7 p ' + 7 Tr' n ì

C o m p o n c n ts in z \ are con sid ered the total

o f th e lin c a r im p a c ls a n d e q u a tio n (1 ) c a n be

so lv ed b y linear regression m ethod w ith the

foIlow in g step s o f assu m in g sub-unknow n:

B y a s s u m in g R l = f lH l a n d

c = g , ~ ( V 0 + u ị , z t w ill b e c o m e :

n

2 , = 2 Ơ + z R iCOs(q,t - c ẽ) (4 )

Ì - /

B y assu n iin g x Ị = R lc o s(C l ) and

x ‘ = R ^ i n í C ị ) , (4 ) can b c perform cd as:

z, = z0 + [cos^C, M + sinrc, )x2, ] (5)

i-l

By assu m in g a l ị = c o s ( C ị) and

a 2 ị = sin (C ẳ) , eq u atio n (5) beco m es the

fo llow ing lin ear equation:

z, = z 0 + Ỳ X a l i x l i + a2i x2i } (6)

i-1

In this stu d y , co m p o n en ts o f z ’i in equation

(3 ) can be ex p ressed by the follow ing linear

function:

m

/=l

In this eq u atio n , m is the nu m ber o f

affectin g factors, a 3 j are co n stan ts, X3 is

affectin g íactors, j = 1, 2, m W ith rainfall, the aíTecting factors in clu d e to tal rain íall on the

w atersh ed accu m u lated from lh to 2 4 h , 25h - 96h, 97 h - 264 h an d 26 5h - 6 2 4 h b efo re tim e t

W ith w ind specd and air p ressu re, aíĩe c tin g íacto rs include w in d v ccto r 10 m , sea surĩace pressure T hese c o m p o n e n ts are av erag ed on

th e ch o sen arca w ith cu m u lativ e tim e rang es Oh

- lh , 2h - 5h, 6h - I4 h , 15h - 30h, 31h - 55h befo re tim e t C o m p o n e n ts ex p ressin g w ater levcl trend d r i v e n b y g lobal c lim ate is th e last

co m p o n en t o f z \ vvith v alu e o f (ỉ3mx 3 m in

w h ic h x3 m is e q u a l t

C o m b in e eq u atio n (6 ), (7), and (1 ), the

fu nction p e ríb rm in g Z t is:

Z, = z 0 + Ẻ ỉa / ‘ x / < + <>2 i x 2i ] + zi (8)

/»/

O r:

z , = Zo + Ỳ Ả aI‘xI‘ +a2ix2í] + Ỹ ềa3j x3j (9)

W ith th e se tran sfo rm a tio n s, eq u atio n (9) has

lin ear form w ith fo llo w in g v ariab les: x l Ễ , x2 ị

an d x 3 j T h is eq u a tio n w ill be solved by

m u ltip le linear reg ressio n m e th o d w h ich is

b ased on m inim um squ are m ethod Hovvever, in

o rd e r to ch o o se su ita b le tidal as vvell as non- tidal factor, in this stu d y , o n ly m u ltiple linear reg ressio n is ap plied T h e seq u en tial m u ltiple linear regression is sim ilar to m u ltip le linear regression in w h ich the eq u atio n qu ality gained

is th e b est by ro tatin g m atrix o f affectin g íactors

ch o sen from eq u atio n b u ild in g steps

A fter solving out variables x l ẵ and

x 2 ị th rough interm ediate steps, R, and £ á can be

226 L v Viel el al / V N U Ịournal o f Science, Earlh Sciences 25 (2009) 223-231

d eíìn ed and then H, and g, can also be deíined

By replacing these values in to (2), tidal w ater

level wil! be d eíìn ed at any g iven time

In o rd er to im p ro ve th e q u ality o f

z , m a n ip u latio n , the p ro cess to b u ild the

eq u atio n follow s tw o steps belovv:

- S tep 1: B uilding a3J factors throu gh

eq uation (9) for ev ery m o n th w ith w ater

level o b serv ed from 1980 to 2004

- Step 2: B uild in g v alu es o f H, an d g,

th ro ug h o n e - y e a r d ata o f o b serv ed

w ater level ch o sen ran d o m ly from data

series o f tim e perio d from 1980 to 2004

th rou gh eq u atio n (8) w h en the v alue o f

z ' f is s o l v c d o u t from v alu es o f a 3 j

T h e assessm cn t o f the qu ality o f the

eq u atio n is th ro u g h statistical ĩa cto rs includ ing

F ish er statistical index, th e d ev iatio n o f

reg ressiv e factors, d efin in g facto rs, m ean erro r

and m ax crror T h ese facto rs are also

d eterm in in g facto r for th e n u m b e r o f w av e to be

cho sen T he n u m b e r o f w av e w ill be d epen dent

on th e length o f an aly zin g series and the

featu res o f every area F or th e stu d ied area , if

the length o f d ata series is lo n g en o u g h , the

nu m b er o f w av e ch o sen w ill ra n g e from 18 for

on an d near- sea locations to 22 for lo catio n s lying d e e p in the land

B esides, th e qu ality o f sim u la tin g w ater level is also assessed by o b se rv e d w ater level data w hich is no t included to b u ild , H, and

g, factors T h is data is taken fro m 2005 to 2 0 07

2.2 D ata u sed

T h e d ata used in th is s tu d y in clu des d a ta o f

ob serv ed vvater level, rain fa ll, w in d a n d air pressure T im e rang e o f th e se d ata se rie s is from 1980 to 2007 T h e w a te r level d a ta is hou rly o b serv ed data T he ra in ía ll d a ta is d aily data, th e w in d an d air p re ssu re d a ta ty p e are taken 4 tim es a d ay In o rd e r to tra n síb rm th e se data to th e sam e tim e reso lu tio n w ith w ater level d ata, th cy w ill bc be in terp o lated by u sin g linear m ethod

T h e w ater level m o n ito rin g sta tio n s th a t are used in the an aly sis in clude V u n g T a u , N h a B e, Phu A n, H oa A n, Ben Luc and T an A n T h e

lo catio n s o f these statio n s are illu stra te d in

F igure 2

Fig 2 Localion o f water level stations

L v Viet et al ỉ V N U lournal o f Science, Earth Sciences 25 (2009) 223-231 227

T h e w ind a n d a ir p re s s u re d a ta a re fro m

g lo b al re a m ly z e d d ata o f N C E P in the

h ttp ://w w w cJc.n o aa.g o v /d ata/g rid d c d /d ata.n ce

p rean aly sis.su rface.h tm l, these d ata h av e the reso lution o f 2.5 X 2.5 d egree T he ch o sen area from 102.5° to 112.5°E and 5° to I5°N is illustrated in Figure 3

Fig 3 Surface wind field o f the studied arca

T h e rainfall m o n ito rin g data is taken from

statio n s lying in the stu d ied area, in o rd e r to

im prov e the q u a lity o f sim ulatio n, som e statio ns

n earb y are also included T h e area for ch o o sin g

rainfall m o n ito rin g statio ns is from 106° to

108°E and 10° to 12°N T h ere are to tally 13

rain fall m o n ito rin g statio ns taking p art in the

calcu latio n in c lu d in g Bao Loc, B icn H oa, Dac

N o ng , D ong P h u, M oc H oa, Phuoc Long, So

S ao, T an A n, T a n Son H oa, T ay N in h , T ri A n,

V u n g T au , X u an L oc A part from these

statio n s, d ata fro m 12 rainfall g au g in g site in

the studied a re a are also used T h e d ata form to

be used is th e av erag e rain fall in pixel w ith

reso lutio n o f 0.5 X 0.5 degree

3 R e s u lts a n d D iscu ssio n

3 1 The q u a ỉity o f sim ulation o f w ater level with

the sup plem enta tion o f n o n -tid a l com ponents

A fter su p p le m c n tin g non - tidal

co m p on ents, th e v alu es o f am p titu te (H ,) and

initial p h ase (g.) o f tidal w ave in vvater level sim u latio n are chang ed T h e illu stration o f this

c h an g e at V un g T au statio n is shovvn in table 1, here, ch an ge m ain ly o ccu rs in tid al w ave ssa, sa

an d k l In ad d itio n , there ex ist ch ang e in

co m p o n en t Zo, in V u n g T au statio n , the values

o f Zo b efo re and afte r ad d in g no n - tidal

co m p o n en ts are -2 4 cm an d -22 cm respectively

T h e quality o f th e sim u latio n o f w ater level ílu ctu atio n w ith th e su p p lem en tatio n o f non- tid al co m p o n en ts is acc essed th ro u g h the

d eterm in e d c o c tĩic ie n t an d th e m ean error

T im e o f assessm en t is th e tim e ran g e th at w as

no t inclu d ed in b u ĩld in g the íacto rs a3j, H, and

g , , from 2005 to 2 00 7 T h ere are tw o ty p es o f assessm en t including:

- D etcrm in ing the qu ality o f th e sim ulation

o f tidal w ater level by eq uation (2 ) w ith values

o f H, and g, built fro m step 2

228 L v Viet ct aỉ. / V N U lournal o f Science, ía rth Saence> 25 <2iH)9> 223-2

- D eterm ining th e q u ality o f th e sim u lation calib rate the w a te r level v ariatio n d u e to

non-ta k in g in acco u n t co m p o n en t z \ used to tidal íactors

Table 1 The amplitude and initial phase values o f tidal wave in Vung Tau station

13 mk3 4 4,0251729 2.82627 -120.98422 2.82923 -120.94674

15 op2 2 8.9019669 2.07748 -94.18789 2.08401 -94.22076

16 2n2 27.8953548 2.43837 14.16457 2.44413 14.12801

Time

Fig 4 The observed and simulated vvatcr level in Jan/2007 o f Vung Tau station

A ssu m in g R2|, E ĩ| a re d e fin in g co efficien t

and m ean error w hen sim u latin g th e tidal w ater

level b y equation (2 ), w h ich m ean s n o t taking

into acco u n t the v ariatio n b y w in d , a ir force and

rain o f w atcr level; R 22, Er2 are defining

co efficien t and m ean e rro r w hen s in iu la tin g the

tidal vvater lev el by equation (2 ) in case th e

v alu es o f H, an d g, alread y tak e in to a c c o u n t the

e íĩe c t o f w e a th e r elem en ts; R23, E r3 are

d e íìn in g c o e ffic ie n t and m ean e rro r w h en sim u latin g the w a te r level w ith the e ffe c t o f vveather elem en ts is already taken into a c c o u n t

L.V Viet et a i / V N U lournal o f Science, Earth Sciences 25 (2009> 223-231 2 2 9

(in equation (8)) T h e assessm en t resu lt is based

o n the d iíĩe re n c e b e íb re an d after

su p p lem en tin g th e aíĩe c tin g disp lay ed on T ab le 2

ía cto rs and

Table 2 Change o f defming coetĩicient and mean error

E rl-E r3 (cm) 2.2 4.8 4.1 3.5 3.4 3.4 3.2

R esult from T ab le 2 shovvs th at w ith the

su p p lem cn tin g o f w eath er elem en ts in b u ild in g

harm onic constants, the q u ality o f sim u latin g

the tidal level is im proved w ith the m ean error

red u ced ab o v e 1 cm from o n -riv e r stations T h e

m axim um m ean e rro r for all statio n s đ ecreases

5 cm w h ich is p retty m eaningfiil in bu ild ing

tid al w ater level tables T ab le 2 also su g g ests

that vvhen sim u latin g w ater level w ith the

su p p lcm cn tin g o f vveather ía cto rs, the q u ality o f

sin iu latio n is im proved sig n ifican tly E x cep t for

V ung T a u station w h ich has m ean error

d e c re a sin g 2.2 cm , the in riv e r statio n s h av e the

d e c re a sin g level o f 3 cm T he m ax im u m error

avcrag ed fro m statio ns d ecreases 14 cm

3.2 A nalyzing the ro le o f non-tidal com ponenís

T o see the e íĩe c t o f rain fall, w in d and air pressure on the v aria tio n o f gau ged w ater level, these co m p o n en ts are sep arated from

c o e íĩic ie n t o f s im u la tin g eq u atio n in general, the sim u latin g re su lt o f the w ater level variation

d u e to n on-tidal e lem en ts has p retty good relatio n sh ip w ith th e d ifference o f th e ob served

w ater level an d th e tid al vvater level T he nearer

th e station is to th e riv er m outh, th e b etter this relatio n sh ip is F ig u re 5 is th e illu stratin g resu lt for V un g T au sta tio n in 2005

£

o

■ *—'

- C

ơ>

03

X

Dỉfference o f the observed water level and the tidal water level VVater level variation due to wind and air pressure

Vung Tau statio n , 2005 '

1*^

§ 5

9 9

ỏ ỏ

V—

o

c

“ 5

O i c (TJ

“ĩ

£3

u

ế

5

■ 2

CN

3 k

to

2

c

—>

co c

—ì

CN 3

—>

0 cn 1 (N

Fig 5 The resuỉt o f simulating water level variation due to wind and air pressure

T h e ro le o f th e no n -tid al co m p o n en ts on

vvater level v ariatio n is sh ow ed th ro ug h

Standard d ev iatio n (S ) o f w ater level sim u latin g

d ata s e rie s o f these co m po nents T h e result is

sh ow ed in T a b le 3 T h e statistical resu lt

su g g ests that at sta tio n s on th e sea (V u n g Tau station), s in dry season has v alu es g reater than

in rainy season T h is íe a tu re is due to the

230 L.v Vict et aỉ / V N U ịoum al o f Science, Earth Sciences 25 (2009) 223-231

d iíĩe re n c e o f N o rth E ast and S outh W est

seasonal w in d in term o f w in d v elo city and

d irectio n as w ell as w ind period A t N ha B e and

Phu An statio n s has sim ilar ch aracteristics,

hovvever, a t statio n s lying d eep in to the land,

th e eíTect o f rainfall on s is m ore clearly At

th e s e statio n s, the v alu e o f s is p re tty high in

th e en d o f d ry seaso ns (M arch and A pril) and esp e c ia lly in the m id o f rainy seaso n s (from

J u ly to S eptem ber)

T ab le 3 Standard deviation o f non -tid al w a tc r level

M onth V u n g T au N ha Be Phu An B ien H oa T h u Dau Ben Luc T an An

T able 4 T rcn d o f w ater levcl ai o bservation statio n s d u rin g p eríod 1981-2007

S tation V u n g T au N ha Be P hu A n B icn Hoa T hu Dau B en Luc T a n An

T re n d o f w a te r lcvel

ta k in g into acco un t n on-tidal c o m p o n en ts The

im pact o f w ind on w ater levcl v arialio n is

T h e w ater level trenđ at ob scrv atio n stations

is d e íin e d from a 3 m coefTicient, the calculation

resu lt is show n in T ab le 4 D ue to calculation

m ethod, th is resu lt is m ean vvater level increase

T h is resu lt sh o w th a t in the p erio d 1978 - 2007,

th e averag e w ater level at ob serv atio n station s

had increased sig n iíica n tly T he m axim um

increase is 13 cm at o b serv atio n statio ns on the

sea (V u n g T au statio n ), w h ich is app ro xim ately

equal to the risin g o f w ater level at inshore

statio ns o f V ietnam [3-5] A t statio n s on the

river, th e risin g level is low er and has values

rang in g from 8 cm to 12 cm D ue to the fact

th at th e accu racy o f the result o f w ater level

sim u latin g is n o t really high for th e statio ns on

the rivers, th is is o n ly co nsidered the

referencin g data

C o n c lu sio n s

c o n sid era b le, esp ecially in dry seaso n T he rainfall also play s an im portant role to the vvater level v ariatio n on thc riv ers w h ich is revealed

m ost c le a rly in th e m id d le o f m o n th s o f rainy season T h e su p p lem en tin g o f w ater lev el trend

c o m p o n en t into an aly sis im p ro ves th e q u a lity o f the sim u latio n and sh o w s o f f a rise a t a lert lcvel

o f w a te r level d u e to global c lim a te c h a n g e in Sai G o n - D o n g N ai riv er system

In o rd e r to im p ro ve th e q u a lity o f sim u latio n , it is n ece ssary to im p ro v e the affectin g facto rs esp ecially th e d a ta o f flo w in the u p p er stream T h e ap p licatio n o f this

m eth o d in íò rec astin g w ater level in th e stu d ied area is feasible w h en it is c o o p e ra te d w ith

P ro du cts from w cath cr m odels

T he assessm en t sh o w that the q u ality o f

w ater level sim u latio n on Sai G on - D ong N ai

riv er system is im p ro v ed co n sid erab ly and

m ax im u m error d ecre ase s sig n iĩica n tly w hen

R eíeren ces

[1] Luong Van Viet, Research to create h y d ro -

m eteorological database 10 serve inundation

L.V Vict et ứ/ / V N U Ịournaỉ o f Science Earth Sciences 25 (2009) 223-232 231

prevcntion in Ho Chi Minh city, Scientifìc

research - Center o f Managing ỉnundation

Preveniion Program Ho Chi Minh City, 2008

[2] Luong Van Viet, Building H ydrology

M cteorology database in conịunction vvith land

use planning till 2020 at Nhon Trach D istrict,

SEM LA prọịect, Dong Nai Ministry' o f Resources

and Environment, 2008

[3] M inistry o f Resources and Environmení,

D evelopm ent o f National target program on clim ate change adaptation, 2008

[4] M inistry o f Resources and Environmenty C lim ate

C hange and Sea Lcvel Rise Scenarios for Vietnam , 2009

[5] N guycn Duc N gu, C lim ate C hange and C oping Sưategy, T he H ydro m eteorological jo u m al' 567(2007), p 9 - !3th