Approved OT public release; £NVPREDRSCHFAC

01st ibutiu iImited. Technical Paper No, 6-72

CHANGES IN THE CHARACTERISTICS OF TYPHOONS

CROSSING THE PHILIPPINES

SAMSON SIAND

ad

JACK W. BLELLOCH

MAY 1972

NATiON Nt TECHNICALINFORMAtON SERVICE

ENVIRONMENTAL PREDICTION RF•SEARCH FACItLITYN ¶AVAL PO4 ' AT" 7 " .

,MPONTERE BUIATfi SC99OO4MONHREYI •A !FORI•L 93940

S Approved for Public Release; ENVPRiEDRSCHFACDistribution Unlimited. Technical Paper No. 6-72

CHANGES IN THE CHARACTERISTICS OF TYPHOONS.CROSSING THE PHILIPPINES

by

I° Samson Brand

and

Jack W. Blelloch

MAY 1972

ENVIRONMENTAL PREDICTION RESEARCH FACILITYNAVAL POS-GRADUATE SCHOOL

MONTEREY, CALIFORNIA 93940

rf :ro

FABSTRACT

Thirty typhoons (1960-1970) are examined to determine

the effect of the Philippines on the intensity, speed of

movement and size characteristics of tropical cyclones

crossing the Philippines. The results show an average in-I •tensity (maximum surface wind) decrease of 33%, a northward

perturbation as the storms pass through the Islands, and a

decrease of circulation size for weak typhoons. The study

also showed an increase in speed of movement as storms

approach the Philippines.

Z

Preceding page blank

ii

It ~--

CONTENTS

P ageABSTRACT ............. ....................... .iiLIST OF ILLUSTRATIONS ........ ............... vii

1. INTRODUCTION ......... ..................... 1

2. DATA SOURCES ........... ...................... .. 5

3. DISCUSSION OF RESULTS ............. .............. 7

4. SUMMARY ......................... 21

REFERENCES .................................. 23

APPENDIX A -TRACKS OF TROPI( AL STORMS ANDTYPHOONS (1945-ý.ý9) . . . . . . . . . . . 25

Preceding page blank

LIST OF ILLUSTRATIONS

Figures Page

1. Map of the western North Pacific Ocean .... ....... 2

2. Topographical map of the Philippine Islands . . . . 3

3. Track segments for 30 typhoons crossing thePhilippines ................ .................... 9

4. Average intensity and speed profiles for crossingtyphoons and average 24-hour right-angle forecasterrors ................... ....................... 10

5. Schematic presentation of 24-hour right-anglefor'ecast error ......... ................... .i.11

6. Average intensity and speed profiles for intenseand weak typhoons ........ ................. ... 14

7. Average intensity of typhoons prior to crossingthe Philippines versus change in intensity oftyphoons crossing the Philippines . . . . . . . . . 15

8. Eye diameter and size variation of typhoonsScrossing the Philippines ..... .............. ... 17

9. Eye diameter and size variation of intense andweak typhoons crossing the Philippines ....... ... 19

Table

1. Typhoons crossing the Philippines in the period1960-1970 ............ ....................... 8

Preceding page blank vii

1. INTRODUCTION

In a previous publication the forecaster was given sta-

tistical information on the geographic and seasonal variations

of tropical cyclone intensity changes for 100 x 100 latitude/

longitude areas of WESTPAC (Brand and Gaya, 1971). Weakening

tropical cyclones at low latitudes were examined in a separate

publication (Brand, 1972). It became evident from these

general studies that a detailed examination of tropical

cyclones crossing the Philippines would provide a useful input

to the tropical cyclone forecaster, since in many situations

he has to forecast the movement and intensity of storms to the

west of the Philippines while they are still to the east of

the Islands.

The purpose of this report is to familiarize the fore-

caster with those characteristics of typhoons that are af-

fected by the Philippines and to quantitatively describe these

relationships. This information, in conjunction with conven-

tional prediction techniques, should be a useful forecast aid

to the tropical cyclone forecaster.

A map of the western North Pacific Ocean area and a

topographical map of the Philippine Islands are presented

as Figures 1 and 2,respectively.

lii

0 0 0 0 00 C2 3 CD C

0

C3-

0

0 04

C-

C1.

04

0

K ..ii Hf

7

1150 1200 1250 1300

20020-

I UZON

PHILIPPINESEA

- -~ 150 ~--

SOUTH CHINA

, ';

U Q 100-

S?7 MINDANAO

5 0-

0 1000 3000 R14XY.

I11 I11MR-- I t IMIIT480Y F4 Ipf IIFthi

Figure 2. Topographical map of the Philippine Islands

3

2. DATA SOURCES

The data used for this study were extracted from the

Annual Typhoon Reports published by FWC/JTWC, Guam and from

a history file of tropical storms and typhoons of the western

North Pacific from 1945-1969, compiled by the National Climatic

Center (NCC) for and with the Navy Weather Research Facility.

The history file for the period 1945-1969 contains 6-hourly

information on such storm characteristics as the location,

movement, size and intensity of tropical cyclonic circulations

which, during their life cycle, reached tropical storm or

typhoon intensity.

In this study the typhoons from 1960-1970 were examined.

During this 11-year period there were 34 typhoons 1which hit

the Philippines; of this number, two recurved and two dissipated

while in the Islands. The remaining 30 typhoons were studied

in the time from -48 hours prior to hitting the Philippines

to +24 hours after leaving the Philippines. The parameters

examined in this time period were position (from fix 2 and

ltshould be noted that in the context of this paper, a t~iphoonis defined as a tropical cyclonic circulation which reached

* typhoon intensity at some stage in its life cycle. This intensityis not necessarily achieved during that part of the circulationexistence considered in this paper.

2 The determination of the position of a tropical. cyclone at aprecise time, generally by reconnaissance aircraft penetrationof the center or by airborne, land or ship radar or satellitephotographs.

Preceding page blank 5

best track3 data), intensity (maximum surface wind), speed,

size of circulation (average diameter of outer closed surface

isobar) and eye diameter.

3A post analysis position incorporating all available data.

3. DISCUSSION OF RESULTS

* Table 1 provides a list of the 30 typhoons investigated

in this study, as separated by m',,nth of crossing the Islands.

Fourteen of the 30 typhoons crossed in the October-November

period. The later months of the typhoon season have a great

deal of tropical cyclone activity in the area of the Philippines

as can be seen in Appendix A, which shows the tracks for trop-

ical storms and typhoons for monthly and half-monthly periods

(1945-1969).

The track segments for the 30 typhoons investigated in

the time frame from -48 hours prior to hitting the Philippines

to +24 hours after leaving the-Islands (center position of

storms) can be seen in Figure 3. All typhoon movement _s

from east to west. If the 6-hourly values of intensity and

speed during this before and after period are averaged, an

average intensity and speed profile for the 30 typhoons

crossing the Philippines can be compiled. This is shown in

Figures 4(a) and 4(b). The shaded area in the middle represents

the Philippines; to the right is before crossing and to the

left, after crossing. 1 The intensity profile (Figure 4(a))

shows an average increase in intensity during tre period 48

to 24 hours prior to hitting the Philippines. The intensity

then levels off at about 92 knots and then decreases to about

62 knots while crossing the region of the Islands (a decrease

IThe average time the 30 typhoons existed in the Philippineswas 14½ hours.

7

P44

>C >i CN 00N 00 woc ()0 F14 w k kz w - Y 0 0% L'

1-3rH i.14 r- C-1 r- H P 4 H o

o- E- 4 1 o - , ' - E1r0 H

o~~~~~r Ha 0 -'0' 0 0U0 U

w ~04

HO 0(1)

(1) CN mOH

w ~04r' H0' H0'4koH ko

0" 0)ON )-

dr41a)

004

~z3N U)

z B

~I0 -ri

1 ii W~OHC -

CD 4 L C

C-44n L 4)4JU)Q

:5 444

4J '9T H U)

1 4- -I--

CD C

... ...-..-.. -- - 100

(ktsl

(a) Z 80LLJ

z 70

S. .. ........ [~ ... . ... . 0

(b) u k

U, 11

-- - 10

PositiveferIcasts to the North

CM 114.1 nm"11081- -10

1 - ' m '°0-3.ný .O m in(c) a=.- 3,9nm -2.0nm 0

C.Z 11081 1821,,,r [ - 10

,-18.3nin Negative.Forecasts to the South

~I7)jINumber offorecasts

f24 012 0 0 -12 -24 -36 -48

AFTER CROSSING IHours) BEFORE CROSSING

Figure 4. Average intensity (a) and speed (b) profilesfor 30 tyjhoons crossing the Philippines from -48hours prior to hitting the Philippines to +24 hoursafter leaving the Islands. The shaded area rep -e-sents the Philippines. The average 24-hour right-angle forecast errors (from best track) are pre-sented in section (c) of the figure.

10

of approximately 33%) before starting to increase in intensity

in the South China and Sulu Seas.

The average speed of movement profile (Figure 4(b)) shows

a decrease in speed until about 20 hours before hitting the

Philippines, then a slight acceleration approaching the Islands

(approximately a 10% increase in speed). and finally a decreaseL__•- in bpeed again as thf- stornms crcoss i Ip.L'L" "ines an ^

into the South China and Sulu Seas.

"Another parameter that was examined was the movement

variations of storms crossing the Philippines. This was done

, indirectly by examining 24-hour right-angle forecast errors;

that is, the forecast error to the right or left of the best

track (see Figure 5). In general, a 24-hour forecast position

- - - 24-Hour Forecast Position

24-HourRight-Angle Forecast Error

__ Initial

Best Track Position

Figure 5. Schematic presentation of 24-hour right-angle forecast error

i1

indicates where persistence and/or steering is tending to move

the storm. Any deviation from this, such as an average right-

angle forecast error from the best track, would point to an

effect not considered in the normal forecast. For example,

Figure 4(c) shows that the average right-angle forecast error

for the 82 available forecasts in the 48 to 24 hours prior to

hitting the Philippines is 2 n mi to the south of the best tracJ

This is reasonably near zero, which is to be expected if no

bias is present. In the 24 hours prior to hitting the Islands

the forecasts are again slightly to the south but near zero. It

will be noted, however, that the average forecast error in

the Philippines is 18.3 n mi to the south of the best track.

This suggests storm movement to the north not considered by

the forecaster. As storms leave the Islands the average fore-

cast error is 14.1 n mi to the north of the best track, suggest-

ing that either the storm centers jumped south or that the

northerly movement in the Philippines was accepted by forecasters

and this had an affect on the persistence element of the forecast.

The question then arises as to whether the effect of the

Philippines is the same for intense and less intense typhoons.

If those typhoons having an initial average intensity of > 90

knots in the 24-hour period prior to hitting the Philippines

are separated from those typhoons having an average initial

intensity of < 90 knots in the 24-hour period prior to hitting

the Philippines, a dramatic difference is noted. For example,

12

as seen in Figure 6(a), intense typhoons are considerably affected

by the Philippines, showing an average decrease in maximum wind of

45-50% while crossing the Islands. However, the weaker typhoons

show a decrease in maximum wind of only 10-15% and return to their

initial intensity within 24 hours of leaving the Philippines.

The speed of movement profiles (Figure 6(b)) show an

acceleration towards the Islands within 18 hours of landfall¶ for both weak and intense systems. However, the weak typhoons

are generally faster moving systems and are more affected by

the Islands. Note the marked decrease in speed through the

Islands shown by the weak typhoons.

Another manner in which to examine the effect of the

Philippines on typhoon intensity is to plot the average inten-

* •sity of the storm approaching the Islands versus the change in

storm intensity which occurred while traversing the Islands.

This can be seen in Figure 7 which is a plot of the average

24-hour intensity prior to hitting the Philippines (I ), versus

the change in intensity (I, -I 2 ),where 12 is the average 24-hour2

intensity after leaving the Philippines The distribution

has a correlation coefficient of 0.83 and the computed re-

gression line based on the distribution could be of use to the

forecaster. For example, Figure 7 shows that if the maximum

2It should again be pointed out that we are examining tropicalcyclones which at some stage reached typhoon intensity (Ž 64knots) but that this level of development did not necessarilyfall within the period 48 hours before reaching the Philippinesto 24 hours after leaving the Philippines.

13

-- 130

-120

INTENSE TYPHOON- -110

100I-INTENSE TYPHOONS -These typhoons having IktsJ

Luprier to hitting the Philippines ofatil knots. 9I- W1 oEAK

WEA TYPHOONS -These typheens having

70

-60

/ ¾. ~WEAK TYPHOONNS ___ 5

-13

b -12

00,-11

10

+24 +12 0 0 -12 -24 *N- -48AFTER CROSSING (Hours] BEFORE CROSSING

Picgurc 6. Averacgo intensity(a) and spood (b) profilesfor intense and wo,-k typhoons crossinct thue Philip--":i !es.

14

h = AVERAGE 24-HOUR INTENSITY PRIOR TO HITTING THE PHILIPPINES

12 =AVERAGEU24-HOUR INTENSITY AFTER LEAVING THE PHILIPPINES

S- 11-12 CHANGE IN AVERAGE INTENSITY

r - 2

<'z

JC

100-______

~- 80.V .-

a-

"60Cm,

4 0- -

,r* OA-00203

0*

-20- -

0 20 40 60 80 100 120 140 160

INTENSITY (11 in knots)

Figure 7. Average intensity of typhoons 24-hours priorto hitting the Philippines (I,) versus the change inintensity (I 1 -I 2 ),where I is the average 24-hourintensity after leaving tde Philippines.

15

wind in a typhoon averages 120 knots in the 24-hour period

prior to hitting the Islands, then the average intensity in

the 24-hour period after leaving the Islands will be 50 knots

less. It will be noted that, based on the regression line,

tropical cyclones with average intensities of less than 60

knots in the 24-hour period prior to hitting the Philippines

exhibit an increase in average intensity in the 24-hour period

after leaving, the Islands. Thus, there is an indication that

the intensity of tropical cyclonic circulations of less than

typhoon intensity are affected very little by the Philippines.

During the course of analysis it was observed that, in

general, the tracks of the more intense typhoons were further

to the north than the tracks of less intense storms.

Two other parameters were examined for typhoons crossing

the Philippines. These were the eye diameter of the storms

and circulation size of the storms. The eye diameter data

were extracted from the Annual Typhoon Reports and the size

of 'circulation data were extracted from the history file of

tropical storms and typhoons discussed earlier. The size

used here is the diameter of the circulation given by the

average diameter of the outer closed surface isobar.

Figure 8 shows the variations of these parameters for the

30 typhoons considered. Eye diameter data were available for

21 of the typhoons and size data for 24. From Figure 8(a) it

can be seen that eye diameters increase throughout the period

16I;I - _ _______

21 IYPIIOOK CASES

S•-- - -- 35

Lij

mF5

124 TPHOO CASE; - .12S-"-11051

- 0

i 15

'24 TON 2ON 0.A2ES 22

S-- -10

AFTER CROSSING IHoursi BEFORE CROSSING

Pibure 8. 'he eye diameter (a) and size variation (b)8)' tyrphoorts crossing the Philippines. The "size"icrameter is a ieasuro of circulation size and re-lates to the averageo C&anieter of the outer closedsurface isobar. TAhe number of observations availableis prcsentet. in parentheses.

17

considered. 3 However, note the large values of average eye

diameter for the 24 hours after leaving the Philippines (33.2

n mi) compared with the average 24-hour eye diameter prior to

hitting the Philippines (22.4 n mi)ý Thus, there is an in-

dication that the Islands did have an effect on the eye dia-

meter. The circulation size parameter shows that storms

maintain their size throughout the 48-hour period prior to

hitting the Islands and then decrease in size (approximately

17% in areal extent) after leaving the Islands.4

These parameters were further examined for intense and

weak typhoons (same criteria as before) and the results cana be seen in Figure 9. The eye diameter variations appear to

be similar for each of the two intensity categories. Notice

that the eye diameter of the intense typhoons is consistently'

smaller (approximately 13%) than the weaker typhoons.

The effect on circulation size varies with typhoon in-

tensity. Intense typhoons decrease in size throughout the

period considered. Weaker typhoons appear to increase in

circulation size prior to hitting the Islands and then to

decrease during the remainder of the period. Notice that the

average circulation diameter of the intense typhoons is approx-

imately 1-2½O latitude greater then the weaker systems.

3 It should be noted that the average time the 21 typhoonsexisted in the Philippines was 13 hours.4 The average time the 24 typhoons existed in the Philippineswas 13½ hours.

18

¶• INTENSE TYPHOONS-Those typhoons having an average intensity inthe'24-hour period prior to hitting the Philippinesof >-90 knots.

{1 WEAK TYPHOONS-Those typhoons having an average intensity inthe 24-hour period prioi to hitting the Philippinesof <90 knots.

INTENSE TYPHOONS 110 cases)- "WIAK TYP'0tONS (11 cases4

-30

L• • 1371(a) •mInm "

a. -- - -- - - 20

LA 1461

-~ ----- 0L--- i20

M -- 5 1 4)15"- INTENSE TYPHOONS (10 cases)

WEAK TYPHOONS 114 casesl

* m .12

- 11

L1400

(b ) I, 1 Latl

.. . . - -. . -- -- - -

1701)

,24 ,12 0 0 .12 -24 -36 -48AFTER CROSSING 1H4u5sl BEFORE CROSSING

Figure 9. The eye diame::er (a) and size variation (b)of intense and weak typhoons crossing the Philippines.The number of obser-zitions available is presented inparentheses.

19

7'. W" M,

4. SUMMARY

An attempt has been made to document the effects of the

Philippines on tropical-cyclones. The results presented here

show -that' the intensityi speed, movement and size parameters

of typhoons are influenced by the Philippines. Primary causes

include the frictional effects of the land mass of the Philip-

pines and the reduction in heat and moisture supplied by the

ocean. However, especially in the later months of the typhoon

season, the ocean environment and mean synoptic conditions of

the South China Sea are different from those of the Philip-

pine Sea, and these characteristics may well affect the behav-1

ior of storms. The degree of land influence is a function

of the area and topography of the terrain over which the storm

is passing. The terrain of the Philippine Islands varies a

great deal (see Figure 2), ranging from extensive mountainous

regions on Luzon and Mindanao to a sea-island mix in the central

Philippines. The forecaster should be aware of these topographi-

cal variations and keep in mind that the central region should

have less effect on the storms.

The synoptics of the situation should also be taken into

consideration. For example, if a tropical cyclone is crossing

1 Some of the effects of the summer and fall season on tropicalcyclones in the South China Sea have been discussed in previouspublications (Adler et al.; 1970; and Brody and Jarrell, 1969).

Preceding page blank 21

the Philippines in the late typhoon season and this crossing

coincides with an early surge of the northeast monsoon, cool

air may enter the circulation of the storm with a resulting

decrease in intensity. However, a surge may only increase the

low-level cyclonic shear in the region of the South China Sea

trough, if present, and this wil] present an excellent path

for the storm.

In the later months of the typhoon season 'the ocean be-

comes a factor since the sea6.surface temperature in the South

China Sea decreases more rapidly than the Philippine Sea. The

thermal structure parameters, other than sea-surface temperature,

are also important in that they determine how much warm water is

available for the energy of the storm (Brand, 1971). For ex-

ample, in the tater months of the typhoon season (October

through Decezý.ber) there is less available warm water in thL

South China Sea than in the Philippine Sea.

It i2 suggested that the forecaster keep the above com-

ments in mind bnd subjectively modify the forecast aids pre-

sented in this report based on the above considerations.

22

REFERENCES

Adler, R. F., L. R. Brody and W. L. Somervell, Jr., 1970:A preliminary survey of SASI'A fall transformationseason weather. NAVWE2'1jSCHFAC Tech. Paper No. 10-70.

Brand, S., 1971: The effects on a tropical cyclone of coolersurface waters due to upwel.ling and mixing produced bya prior tropical cyclone. Jc.rn.al of Applied Meteoro7logy, 10,No. 5, 865-874.

Brand, S., 1972: Geographic and monthly, variation of rapidintensification and low-latitude Weakening of tropýicalcyclones of the western North Pacific Ocean.ENVPREDRSCHFAC Tech. Paper No. 5-72.

Brand, S. and R. F. Gaya, 1971: Intensity changes of tropi-cal storms and typhoons of the western North .'acificOcean. NAVWEARSCHFAC Tech. Paper No. 5-71.

Brody, L. R., and J. D. Jarrell, 1969: A technique for pre-dicting South China Sea tropical cyclones.' NAVWEARSCHFAC

TVch. Paper No. 8-69.

U.S. FLEET WEATHER CENTRAL/JOINT TYPHOON WARNING CENTEP,GUAM, Annual Typhoon Reports, 1960-1970.

23

APPENDIX A

TRACKS OF TROPICAL STORMS

AND TYPHiOONS (1945-1969)

BY MONTHLY AND HALF-MONTHLY PERIODS

FOR THE WESTERN NORTH PACIFIC OCEAN

Preceding page blank 25

A _____ ____________25

l~~~O4 so- 11' *. .' * IO' W Ii'JAN. fZf--SO- -

'/to-

FEB.

j V

MAR,

30'

-.-.-- ~-WOO

2 6'5 0 S ' SW W 7'

___ - - " . .. -. . .. ..."

APR. Ol)1IP IV

IO- 15 . . 7- - I

I ' ', . .. .. . ..._ _ _ _ -M AY "'NI,... I I -+

JUN. o0

-to

a...

27 ,.S,.

JUN . .. •... • ,, .- ,. a.. o',

\ , , . . + . : . o r.. _ . +

'., , 4.4•.1 , .

ION ..... : ...

I ) ' ; * .,

;' -,A -- -+ -*..,

2.-.,7. • ,•. / .

J L.A1*15) 2 Sr 40 Sr .0 I*t W

:: O- I V 3 4. IW-V

d 11.r si• io. 11V- 140. 150•. lo . ,0y. 1 eO O-o4

AUG.1 ] 1"-

*.J t . s,.+, *, - *.." ~I '...-,,-- .. to

I *]

JU . 16 1) '''"'"" '"'' S• * Sr "S ... Sr•... '0.' nrc so ....

) L S+)'- - ,i• " .,.

+ + . ... . .. ... ... .. ../ -S..,So.'

I \0.

282 ... .. ..

I -.

28

"10. *to- W . .0. 1*. 1..o 70og o 'Ps..11G 116-3 "Lli& . to

SEP.I1*1-51 c7,4A.;%TmV%

SEP.(l6.3 0)-.0. 'o.' o. !

xr}.30>do

29...

OCT. (1 -15) or- o SILA'. S. 7*,, -

5U I - .... . .

tt

vc W so W I~vISO M- f - I

717

NO V .,15,.\ _ L ._.'§ ,•.1 .S+4- , ( , wi , j . ' ., 5

INV " , "-<+,+' -~.5•,, -.\.I..,' '1

SI*, p. 7' iol l, ______ '_________ I

* r2 4 I ..

, "I' • 30

Wt W aetc l1w *OW Hg o. G ?o- 110-9nNOV.116-30) .

i IL/,...> 1I"/I.-

"I .< •.~... . .. .- " ~ t_ ' ! II

" Flil~leo *

31. . * i

3]..