• 検索結果がありません。

DISSERTATION FOR A DEGREE OF DOCTOR OF PHILOSHOPHY (SCIENCE) TOKYO METROPOLITAN UNIVERSITY

N/A
N/A
Protected

Academic year: 2021

シェア "DISSERTATION FOR A DEGREE OF DOCTOR OF PHILOSHOPHY (SCIENCE) TOKYO METROPOLITAN UNIVERSITY"

Copied!
150
0
0

読み込み中.... (全文を見る)

全文

(1)

首都大学東京 博士(理学)学位論文(課程博士)

論文名 インドシナ 半 島

はんとう

東部

とうぶ

における 降 水

こうすい

の季節

きせつ

変化

へんか

の気候学

きこうがく

と 年 々

ねんねん

変 動

へんどう

著者 グエン レ ズン

審査担当者

主査 委員

委員

上記論文を合格と判定する

(2)

平成 年 月 日

首都大学東京大学院都市環境科学研究科教授会 研究科長

DISSERTATION FOR A DEGREE OF DOCTOR OF PHILOSHOPHY (SCIENCE)

TOKYO METROPOLITAN UNIVERSITY

TITLE: Seasonal transition of precipitation over the eastern Indochina Peninsula: Climatology and interannual variability

AUTHOR: Dung Nguyen-Le

EXAMINED BY Chief examiner:

Examiner:

Examiner:

QUALIFED BY THE GRADUATE SCHOOL OF

URBAN ENVIROMENT SCIENCES

(3)

TOKYO METROPOLITAN UNVERSITY Dean:

Date:

(4)

Seasonal transition of precipitation over the eastern Indochina Peninsula: Climatology and interannual variability

Dung Nguyen-Le

Department of Geography

Graduate School of Urban Environmental Sciences Tokyo Metropolitan University

August 2015

(5)

iv

Contents

List of figures ... vii

List of tables ... xv

Acronyms ... xvi

Acknowledgments ... xviii

Abstract ... xx

Chapter 1. Introduction ... 1

1.1. Motivational background information ... 1

1.2. Main objectives... 3

1.3. Dissertation outline ... 4

Chapter 2. Climatological onset of summer monsoon over different sub-regions of Vietnam ... 6

2.1. Introduction ... 6

2.2. Data ... 8

2.3. Onset of summer monsoon over various sub-regions of Vietnam ... 9

2.3.1. Determination of the onset dates in the northern and southern regions ... 9

2.3.2. The Foehn wind and onset in central coastal plain ... 11

2.3.3. Moisture flux convergence and Western Pacific sub-tropical high activity ... 13

2.3.4. Analysis of minimum daily relative humidity and maximum daily temperature ... 14

2.4. Large-scale convective activity and atmospheric circulation associated with the onset

period ... 15

(6)

v

2.4.1. Convective activity and low-level flow pattern ... 15

2.4.2. The Meiyu front ... 17

2.4.3. The Tropical Easterly Jet ... 18

2.5. Conclusions ... 19

Chapter 3. Onset of the rainy seasons in the eastern Indochina Peninsula: Climatology and interannual variability ... 22

3.1. Introduction ... 22

3.2. Datasets ... 24

3.3. Determination of the rainy season onset date ... 24

3.4. Climatological features of the rainy season onset over the eastern ICP ... 28

3.4.1. Onset of the SRS ... 28

3.4.2. Onset of the ARS ... 30

3.5. Precursory signals associated with interannual variability of rainy season onset over the eastern ICP ... 33

3.5.1. SRS onset ... 33

3.5.2. ARS onset ... 35

3.6. Concluding remarks ... 39

Chapter 4. Delayed withdrawal of the autumn rainy season over central Vietnam in recent decades ... 42

4.1. Introduction ... 42

4.2. Datasets ... 44

4.3. Decadal change in the withdrawal of the ARS over the CCV ... 46

(7)

vi

4.3.1. Definition of onset and withdrawal of the ARS ... 46

3.2. Decadal change in ARS withdrawal ... 47

4.4. The change of atmospheric circulation and convection associated with the ARSW ... 49

4.4.1. Climatological features ... 49

4.2. Differences between the two epochs 1979–92 and 1993–2006 ... 50

4.5. Potential factors responsible for the decadal change in the ARSW ... 51

4.5.1. Change in low-level circulation and zonal flow ... 52

4.5.2. Impact of circulation change on rain-producing efficiency ... 53

4.5.3. Change in ISO activity ... 54

4.5.4. Change in TC activity and TC rainfall ... 55

4.6. Discussions ... 55

4.7. Concluding remarks ... 56

Chapter 5. General conclusions and outlook ... 61

Bibliography ... 67

Figures ... 80

Tables ... 124

(8)

vii

List of figures

Fig. 2.1. Topography height (gray shaded; m) and climatological summer monsoon onset date (colored dot; pentad number), following the definition of summer monsoon onset date by Matsumoto (1997), at 54 stations in Vietnam during a 25-year period (1979-2003). Topography height depicted at 500 m intervals with values higher than 500 m shaded. The crosses denote stations that have non-typical monsoon rainfall pattern.

Fig. 2.2. The 105°-110°E time-latitude sections of pentad mean (a) OLR (W m

-2

) and (b) zonal winds at 850 hPa (m s

-1

), averaged over 25 years (1979-2003). OLR analyzed at 10 W m

-2

intervals with less than 240 W m

-2

shaded. Zonal wind at 850 hPa analyzed at 2 m s

-1

intervals with westerly wind shaded.

Fig. 2.3. The 13°-18°N time-longitude sections of pentad mean (a) OLR (W m

-2

), (b) the wind vector at 850 hPa (m s

-1

) and relative humidity at 2 m (%, b) and (c) zonal winds at 850 hPa (m s

-

1

), averaged over 25 years (1979-2003). OLR analyzed at 10 W m

-2

intervals with less than 240 W m

-2

shaded. Relative humidity at 2m analyzed at 4% intervals with values higher than 80%

shaded. The reference wind-vector arrow is 10 m s

-1

. Zonal wind at 850 hPa analyzed at 2 m s

-1

intervals with westerly wind shaded. The dashed line denotes P29, the SMOD in the central coastal plain of Vietnam.

Fig. 2.4. Vertically integrated moisture flux convergence (shaded; mm day

-1

) and geopotential height at 850 hPa (contour; gpm) during P22-P31, averaged over 25 years (1979-2003).

Vertically integrated moisture flux convergence analyzed at 5 mm d

-1

intervals. Geopotential height at 850 hPa analyzed at 10 gpm intervals.

Fig. 2.5. Difference between two consecutive pentads of pentad-mean minimum relative

humidity at 2 m (%) during P22-P33 at 54 stations in Vietnam, averaged over 25 years (1979-

2003). Closed and open circles denote stations having statistically significant difference at a 90%

(9)

viii

level, according to the two-tailed Student’s t test, and has difference but not significant, respectively.

Fig. 2.6. Same as Fig. 2.5 but for the maximum temperature at 2 m (°C)

Fig. 2.7. Same as Fig. 2.4 but for OLR (shaded; W m

-2

) and wind vector at 850 hPa (m s

-1

). OLR analyzed at 10 W m

-2

intervals with values less than 240 W m

-2

shaded. The reference wind- vector arrow is 10 m s

-1

.

Fig. 2.8. Same as Fig. 2.7, but for the difference between two consecutive pentads. Difference of OLR analyzed at 10 W m

-2

intervals with negative values shaded. The reference difference wind- vector arrow is 5 m s

-1

.

Fig. 2.9. Vorticity (10

-6

s

-1

) and wind vector (m s

-1

) at 850 hPa in (a) April 15-May 15 and (b) May 16-June 15, averaged over 25 years (1979-2003). Vorticity analyzed at 310

-6

s

-1

intervals with positive values shaded. Moisture transport ( qV ) (arrows; m s

-1

g kg

-1

) and moisture flux convergence (-   qV ) at 850 hPa (shaded; 10

-5

s

-1

g kg

-1

) in (c) April 16-May 15 and (d) May 16-June 15, averaged over 25 years (1979-2003). Moisture flux convergence analyzed at 210

-5

s

-1

g kg

-1

intervals with more than 210

-5

s

-1

g kg

-1

values shaded. Dashed line indicates the frontal position as defined by the axis of maximum vorticity. The missing values denote where the pressure surface is lower than topography.

Fig. 2.10. Same as Fig. 2.4, but for geopotential height (gpm) and wind vector (m s

-1

) at 200 hPa.

Geopotential height analyzed at 20 gpm intervals with values higher than 12430 gpm shaded.

The reference wind-vector arrow is 20 m s

-1

. The dashed line indicates the boundary between westerly and easterly winds.

Fig. 2.11. Geopotential height (contour; gpm) and wind vector (m s

-1

) at 200 hPa and horizontal

distribution of the 200-500 hPa layer mean temperature (shaded, °C) during (a) April 16-May 15,

(10)

ix

(b) May 16-June 15 and (c) difference between these two periods, averaged over 25 years (1979- 2003). Geopotential height analyzed at 20 gpm intervals. The 200-500 hPa layer mean temperature analyzed at 2°C intervals with values higher than -30°C shaded in (a, b) and at 1°C intervals with positive values shaded in (c). The dashed line in (a) and (b) indicates the boundary between westerly and easterly winds during April 16-May 15 and May 16-June 15, respectively.

The dashed line in (c) is the same as the one in (b). In (c), only the evolution of easterly wind is plotted. The reference wind vector arrow is 20 m s

-1

.

Fig. 3.1. (a) Topography (m), seasonal precipitation expressed as a percentage of total annual precipitation (%), and 850 hPa wind vector (m s

-1

) for: (b) December–January–February; (c) March–April–May; (d) June–July–August; and (e) September–October–November averaged during 1958–2007 in the eastern ICP (8.5°–23.5°N, 100°–110°E).

Fig. 3.2. Eigenvector patterns (left) and principal components (right) of two leading EOF modes of daily rainfall in the eastern ICP (8.5°–23.5°N, 100°–110°E) for the period 1958–2007. (a) and (c) denote the spatial distribution of EOF1 and EOF2, respectively. Gray thin lines in (b) and (d) indicate PC1 and PC2 for each individual year, respectively. Black solid curve represents the mean PC over the entire period.

Fig. 3.3. Latitude–time section of 5-day mean (b) actual precipitation amount (mm day

-1

) and (c) standardized precipitation along the eastern Vietnam–China boundary and east coast of the ICP (blue dots in the geographic map (a)). Red and purple lines indicate the mean SRS and ARS onset date over the eastern ICP, respectively.

Fig. 3.4. (a), (c), (e), (g), (i) Evolution and (b), (d), (f), (h), (j) differences between two

consecutive pentads of pentad-mean precipitation (mm day

-1

) over the eastern ICP (8.5°–23.5°N,

100°–110°E) centered on SRS onset dates. The contour interval is 1 mm/day; shaded areas in (b,

d, f, h, j) indicate a significant difference above the 95% confidence level.

(11)

x

Fig. 3.5. Same as Fig. 3.4 but for the ARS onset date. Contour interval is 2 mm/day.

Fig. 3.6. Composite evolutions of pentad mean OLR (gray shaded, W m

-2

), geopotential height at 500 hPa (red contour; gpm), and wind streamlines at 850 hPa (black contour) from three pentads before SRS onset dates to two pentads after SRS onset dates. Only OLR values lower than 230 W m

-2

and geopotential heights exceeding 5860 gpm are plotted; the interval of the purple contours is 5 gpm. The missing areas represent the below ground position of 850 hPa.

Fig. 3.7. (a), (b) Longitude–time section along 10°–20°N; and (c), (d) latitude–time section along 100°–110°E of 10–20DV and 30–60DV of OLR, respectively. Day 0 is the SRS onset.

Fig. 3.8. Same as Fig. 3.6 but for the ARS onset.

Fig. 3.9. Same as Fig. 3.7 but for the ARS onset.

Fig. 3.10. The bimonthly early-minus-late SRS onset composite of: (a)–(c) 850 hPa wind vector, and (d)–(f) SST, in the preceding December–January, February–March, and April–May, respectively. In (a)–(c), the reference difference wind-vector arrow is 4 m s

-1

, and in (d)–(f) the contour interval is 0.2 °C. Shaded areas indicate a significant difference above the 95%

confidence level.

Fig. 3.11. (a)–(c) Composite bimonthly 500 hPa geopotential height indicated by 5865 gpm (thin) and 5860 gpm (thick) contours for early (solid) and late (dashed) SRS onset years; and (d)–(f) the bimonthly early-minus-late SRS onset composite of OLR in the preceding December–

January, February–March, and April–May, respectively. In (d)–(f), the contour interval is 5 W m

-2

; shaded areas indicate a significant difference above the 95% confidence level.

Fig. 3.12. Difference in the mean standard deviation of filtered OLR (W m

-2

) on (a) 10–20- and

(b) 30–60-day time scales in March (early minus late SRS onset). The contours denote

(12)

xi

climatological values during 1979–2007; hatched areas indicate that the difference is significant above the 95% confidence level.

Fig. 3.13. The monthly early-minus-late SRS onset composite of: (a) TT, (b) vertical shear of zonal wind, and (c) 850 hPa specific humidity in March. Shaded areas indicate that the difference is significant above the 95% confidence level.

Fig. 3.14. The bimonthly early-minus-late ARS onset composite of: (a)–(c) 850 hPa wind vector;

and (d)–(f) SST in the preceding May–June, July–August, and September–October, respectively.

In (a)–(c), the reference difference wind-vector arrow is 3 m s

-1

; in (d)–(f) the contour interval is 0.2 °C. Shaded areas indicate a significant difference above the 95% confidence level.

Fig. 3.15. (a)–(c) The composite bimonthly 500 hPa geopotential height indicated by 5865 gpm (thin), and 5860 gpm (thick) contours for early (solid) and late (dashed) SRS onset years; and (d)–(f) bimonthly early-minus-late SRS onset composite of OLR in the preceding May–June, July–August, and September–October, respectively. In (d)–(f), the contour interval is 5 W m

-2

. Shaded areas indicate a significant difference above the 95% confidence level.

Fig. 3.16. Difference in the mean standard deviation of filtered OLR (W m

-2

) on: (a) 10–20-; and (b) 30–60-day time scales in August (early minus late ARS onset). Contours denote climatological values during 1979–2007; hatched areas indicate a significant difference above the 95% confidence level.

Fig. 3.17. Monthly early-minus-late ARS onset composite of: (a) TT, (b) vertical shear of zonal

wind; and (c) 850 hPa specific humidity in August. Shaded areas indicate a significant difference

above the 95% confidence level.

(13)

xii

Fig. 4.1. (a) Topography (shaded, m) of the Indochina Peninsula and surrounding regions and (b) locations of stations used to construct the CCV rainfall index (blue dots). Red box in (a) indicates the domain in (b).

Fig. 4.2. Annual cycle of the CCV rainfall index (mm day

-1

) during the: (a) Gregorian year and (b) climatic year. Gray lines indicate annual cycle of rainfall for each individual year during 1979–2006. (c) Time series of annual mean of CCV rainfall index (mm day

-1

) during the: 1) climatic year (black line); and 2) Gregorian year (red line).

Fig. 4.3. Time series of the ARS: (a) onset and (b) withdrawal dates. The expression “pentad*”

represents the number of pentads in climatic year starting from April 1. The black and red line in (a) denotes the ARS onset dates given by this study and Nguyen-Le et al. (2015), respectively.

The red and green line in (b) denotes the epochal mean for 1979–1992 and 1993–2006, respectively.

Fig. 4.4. Composite evolutions of pentad mean precipitation (shaded, mm day

-1

) and 850 hPa streamline chart (contour) from three pentads before to two pentads after the ARSW. The blacked-out areas represent points at which the level of 850 hPa is below the ground. Red box indicates the domain in Figure 4.1b.

Fig. 4.5. Latitude–time section from five pentads before to five pentads after the ARSW of climatological (1979–2006) pentad mean (b) rainfall (mm day

-1

), (c) outgoing longwave radiation (OLR, shaded, W m

-2

) along the eastern Vietnam–China boundary and east coast of the ICP (blue dots in the geographic map (a)), and 850 hPa wind flow (vector) and its zonal component (contour) along 110°E. Red line indicates the climatological ARSW date.

Fig. 4.6. Latitude–time section from P47* (Nov. 17–21) to P57* (Jan. 6–10 of the following

year) of pentad mean (a) and (b) rainfall (mm day

-1

), (c) and (d) outgoing longwave radiation

(OLR, shaded, W m

-2

) along the eastern Vietnam–China boundary and east coast of the ICP

(14)

xiii

(blue dots in Figure 4.5a), and 850 hPa wind flow (vector) and its zonal component (contour) along 110°E during 1979–1992 and 1993–2006, respectively. Red lines indicate the epochal mean ARSW dates.

Fig. 4.7. The 850 hPa streamline chart superimposed with zonal wind in December during (a) 1979–92, (b) 1993–2006 and (c) the difference between these two epochs (1993–2006 minus 1979–92). Only differences that are significant above a 95% confidence level are plotted in (c).

The blacked-out areas represent points at which the level of 850 hPa is below the ground.

Fig. 4.8. (a) Time series of SST anomalies in December averaged over the Niño-3.4 area (5°S–

5°N, 170°–120°W),  SST(Niño-3.4) (bar), and over a region (128°–132°E, 8°–12°N) in the western tropical Pacific  SST(WTP) (line). The latitude–time section of December monthly mean (b) rainfall along the eastern Vietnam–China boundary and east coast of the ICP (blue dots in the Figure 4.5a), and (c) 850 hPa zonal wind along the line connecting two locations: (125°E, 35°N) and (105°E, 0°).

Fig. 4.9. (a) The correlation map between the ARSW dates and 850 hPa zonal wind in December. The time series of December monthly mean 850 hPa zonal winds at (b) (110°E, 17.5°N) and (c) (110°E, 7.5°N) during 1979–2006. The red and green lines in (b) and (c) denote the epochal mean for 1979–92 and 1993–2006, respectively.

Fig. 4.10. The potential function of water vapor flux (

Q

) or its departure superimposed with precipitation/precipitation departures in December during (a) 1979–92, (b) 1993–2006 and (c) the difference between these two epochs (1993–2006 minus 1979–92). Only differences that are significant above a 95% confidence level are plotted in (c).

Fig. 4.11. Longitude-time section of the intraseasonal oscillation of OLR on (a) and (b) 5-day,

(c) and (d) 12–24-day and (e) and (f) 30–60-day time scales along 15°N in November–following

(15)

xiv

January during 1979–92 and 1993–2006, respectively. The red line denotes the position of the CCV.

Fig. 4.12. Year-time section of the intraseasonal oscillation of OLR on (a) 5-day, (b) 12–24-day and (c) 30–60-day time scales at (109°E, 15°N) in November–following January during 1979–

2006. The black line denotes the ARSW time series.

Fig. 4.13. Tropical cyclone tracks passing through the SCS (to the west of 120°E) and/or western North Pacific in December during: (a) 1979–92; and (b) 1993–2006. Red dots indicate genesis positions. (c) Time series of the December CCV total rainfall (black) and TC-related rainfall (red).

Fig. 4.14. (a) The composite 5870 gpm contours of the 500-hPa geopotential height in December for 1979–92 (solid) and 1993–2006 (dashed). (b) The epochal difference (1993–2006 minus 1979–92; shaded) and the mean during 1979–92 (contour) of SST in December. Hatched areas denote that the significance is above a 95% confidence level. Red and green boxes indicate the SCS (110°–120°E, 10°–20°N) and WTP (128°–132°E, 8°–12°N) region, respectively.

Fig. 4.15. Time series of the December monthly: (a) Enso Modoki Index (EMI); (b) Dipole

Mode Index (DMI); (c) PDO; (d) East Asian winter monsoon (EAWM) index; and SST

anomalies over the SCS (110°–120°E, 10°–20°N)  SST(SCS). Solid lines denote the 5-year

moving average.

(16)

xv

List of tables

Table 3.1. Onset date of the SRS over the eastern ICP from 1958–2007. Bold type denotes early onset and italics denote late onset.

Table 3.2. Onset date of the ARS over the eastern ICP from 1958–2007. Bold type denotes early

onsets and italics denote late onsets; years marked with “*” are early-onset years occurring

during the El Niño developing phases; “**” are late-onset years related to La Niña; “***” are

late onset in the year preceding the development of El Niño.

(17)

xvi

Acronyms

10–20DV Intraseasonal oscillation on a 10–20-day variation 30–60DV Intraseasonal oscillation on a 30–60-day variation

AM April–May

ARS Autumn rainy season

ARSW Autumn rainy season withdrawal

ASM Asian summer monsoon

BOB Bay of Bengal

CCV Central coastal plain of Vietnam

DJ December–January

DMI Dipole Mode Index

EMI Enso Modoki Index

ENSO El Niño–Southern Oscillation

EOF Empirical Orthogonal Function

FM February–March

HRF Heavy Rainfall–Flood

ICP Indochina Peninsula

IO Indian Ocean

IPCC Intergovernmental Panel on Climate Change

ISO Intraseasonal oscillation

JA July–August

MJ May–June

MRI Meteorological Research Institute

JMA Japan Meteorological Agency

JTWC Joint Typhoon Warning Center

(18)

xvii

NCAR National Center for Atmospheric Research NCEP National Centers for Environmental Prediction

OLR Outgoing Longwave Radiation

PC Principal component

PSAC Philippine Sea anticyclone

PDO Pacific Decadal Oscillation

SCS South China Sea

SMO Summer monsoon onset

SMOD Summer monsoon onset date

SO September–October

SLP mean sea level pressure

SRS Summer rainy season

SST Sea surface temperature

TC Tropical cyclone

TMU Tokyo Metropolitan University

TT Tropospheric Temperature

VMFC vertically integrated moisture flux convergence

WNP western North Pacific

WPSH Western-Pacific subtropical high

WTP western tropical Pacific

(19)

xviii

Acknowledgments

I would like to gratefully and sincerely thank my supervisor, Prof. Jun Matsumoto, for his advice, guidance, and encouragement throughout the course of my PhD research. I am very happy that we will continue collaborating on new projects. I am also grateful to my thesis examiners, Prof. Hideo Takahashi and Prof. Hiroshima Matsuyama, who give valuable comments and suggestions for the improvement of this work. The advice and discussions that I received from other members of the Laboratory of Climatology, Tokyo Metropolitan University (TMU) in our laboratory seminars and sub-seminars encouraged me to explore further; therefore, I’m very much thankful to them.

Despite not being my official supervisor, I don’t think I would have been able to complete this PhD thesis without the input from Assoc. Prof. Thanh Ngo-Duc, Department of Meteorology and Climate Change, Hanoi College of Science, Vietnam National University . His positive feedback and superb methodological suggestions really brought the work together and helped me culminate the thesis. The constructive criticisms and the great ideals persuasively provided by the anonymous reviewers during the peer-review process of my papers, which are the main components of this thesis, are sincerely acknowledged as well. Also the fruitful discussions that I had with several experts at the conferences I attended helped me to improve my research knowledge and analyses approach.

I am highly indebted for the support provided by the Tokyo Metropolitan Government

through the “Asian Human Resources Fund”; my study would not have conducted without its

financial support given to me. I also want to thank all of my university colleagues and friends for

their support and friendship throughout my PhD. There are many of them, but particular thanks

go to Mrs. Mio Tanahashi, Mr. Hayato Suzuki, Ms. Moeka Yamaji, Ms. Nozomi Kamizawa, and

(20)

xix

Mr. Kohei Okuhara. The Vietnamese students in TMU and entire Vietnamese community in Japan in general helped me also in several ways.

Some of the biggest thanks go to my family who have encouraged me every step of the way.

I am lucky to have my parents and my younger sister, who always support me. The very biggest

thanks go to my wife and my son, the most important people in my life. They give me praise on

the good days and encourage me on the bad days. I couldn’t have done it without their support. I

would like to dedicate this PhD to one special person, my daughter who, as I am writing this

thesis, should seriously be planning her entrance into this world!

(21)

xx

Abstract

The summer monsoon onset (SMO) is considered to be the key characterizing indicator of the transition between dry and rainy season in the Indochina Peninsula (ICP). However, the SMO have not been sufficiently examined over Vietnam and the eastern ICP in general. Thus, first, the detailed climatological SMO over various sub-regions of Vietnam is examined. In the central coastal plain of Vietnam (CCV), rainfall reaches its maximum intensity in autumn;

whereas, a strong mountain shadow effect causes relatively dry summer. Thus, a new specific onset criterion for this region has been proposed: The summer dry season onset date, which is indicated in late May and marks the appearance of dry conditions. From a large-scale perspective, the SMO in Vietnam and surrounding areas coincides with the northward retreat of the mid-latitude westerlies, and the eastward retreat of the easterly trade winds associated with the Western Pacific sub-tropical high (WPSH). A relationship between the occurrence of the Baiu front in East Asia and the SMO over the South China Sea (SCS) has also been identified.

Next, the climatology and precursory signals associated with the interannual variability of both the summer and autumn rainy season (SRS and ARS) onsets over the eastern ICP are investigated. Results indicated that during 1958–2007, the average onset of the SRS and ARS in the CCV occurs respectively on May 6 and September 16 with a standard deviation of 13 days and 12 days. The SRS onset is characterized by the northward propagation of strong convection over Sumatra and the evolution of summer monsoon westerlies. Conversely, the withdrawal of the summer monsoon over northern and central Indochina in autumn favors the onset of ARS.

Both onsets are strongly related with an intraseasonal oscillation (ISO) on the scale of 30–60 and 10–20 day. The results also suggest that ENSO has a considerable influence on interannual variations of the onset dates. In La Niña years, the following SRS tends to have early onsets.

Simultaneously, the WPSH weakens and retreats eastward earlier. In addition, advanced ARS

onset generally occurs during an El Niño developing autumn with weakened equatorial easterlies

(22)

xxi

and suppressed convection over the central Indian Ocean from the preceding summer, as evident of a weakening Walker circulation. However, robust precursory signals in SST are observed only from mid–summer (July-August). Also an earlier ARS onset is associated with a development of an anomalous Philippine Sea anticyclone and a westward-extended WPSH from mid- to late summer. However, no coherent correlation is found between the late onset and La Niña.

Finally, a significant and abrupt delay in the withdrawal of the ARS over the CCV since

around 1992/1993 is detected and associated dynamical mechanisms are discussed. During 1979-

92, the mean withdrawal is in early December, which is three pentads earlier than that during

1993-2006. Since rainfall over the CCV is primarily produced by cold surge vortices formed by

the interaction of easterly waves with the cold surge flow, the ARS withdrawal is characterized

by a gradual equatorward propagation (retreat) of dry continental air by the northeasterly winter

monsoon (tropical easterlies). Thus, the relatively late withdrawal in the recent epoch is

determined by a westward extension of the tropical easterlies during early-to-mid December that

causes an increase in the easterly wave activity over the Philippine Sea and the South China Sea

(SCS), in response to the cold sea surface temperature (SST) anomalies in the central-eastern

Pacific and warm SST anomalies in the western Pacific. As a result, since 1993, remarkable

increases of convection and moisture convergence around the CCV and SCS have occurred

during December. Large-scale circulation changes also favor an enhancement of the

intraseasonal oscillation (ISO) activities on 30–60, 10–20 and 5-day scales that formed and

maintained rainfall in the CCV. Another potential factor is a distinct increase in the number of

tropical cyclones (TC) passing through the southern SCS in 1993-2006 compared with those

occurring during 1979–92, which is related to a strengthening and more westward-extended

WPSH and a significant sea surface warming over the western Pacific. This enhanced ISO and

TC activity, and the markedly delayed ARS withdrawal, which is associated with interdecadal

circulation changes, may be attributed to a mean state changes in the Pacific basin since the mid-

(23)

xxii

to-late 1990s characterized by a grand La Niña-like pattern, which also results in the

simultaneous advance of the SCS summer monsoon onset.

(24)

1

Chapter 1. Introduction

1.1. Motivational background information

The Asian monsoon precipitation represents the life-blood of two-third of the world’s population. As its variability can have devastating effects on agriculture, ecosystem, economics, and society, it has been the subject of scientific research for several decades. Recently, several studies showed that global monsoon precipitation and area possess profound increasing trends in recent three decades (e.g., Hsu et al. 2011, Wang et al., 2012). Furthermore, the future climate projections based on Intergovernmental Panel on Climate Change (IPCC) models predict intensified global monsoon precipitation in the 21

st

century under anthropogenic global warming (Hsu et al. 2012; Lee and Wang, 2014). However, Wang et al. (2012) argued that this recent global monsoon variability is highly attributed to the mean state change in the Pacific basin which is primarily ascribable to natural decadal variability.

In addition to the seasonal mean rainfall variability, understanding the interannual and long-

term changes in the rainfall seasonal cycle is a critical issue. In particular, changes in the onset

and withdrawal dates of the rainy seasons have important implications for agriculture planning

and water management, such as crop selection, planting, and irrigation. In this respect, a

significant decadal change in the onset of the Asian summer rainy season (SRS) has been

observed around 1993/1994 (Kajikawa et al., 2012). In detail, the mean onset during 1994–2008

occurred earlier over the Bay of Bengal (BOB) and the western Pacific region by approximately

10–15 days compared with that during 1979–1993. In the paper of Kajikawa and Wang (2012),

the South China Sea (SCS) summer monsoon exhibits a similar feature with a simultaneous

advanced onset after 1993/94. Wang et al. (2009) also showed that since 1993 the summer and

autumn rainfall has increased over southern China and northern SCS, while it has decreased over

the central SCS.

(25)

2

The study on the linkage between the Asian summer monsoon/SRS onset and El Niño–

Southern Oscillation (ENSO) has been advanced rapidly in recent years despite most papers address the Indian and SCS monsoon, in which La Niña (El Niño) tends to advance (delay) the onset date (e.g. Lau and Yang 1997; Xavier et al. 2007). Lau and Yang (1997) showed that the delayed (advanced) SCS monsoon onset might be related to basin-wide warm (cold) events of the Pacific and IO. Zhou and Chan (2007) also found that in years associated with a warm (cold) ENSO event or the year after such years, the SCS monsoon tends to have late (early) onsets.

Recently, Xiang and Wang (2013) revealed that the advance in the SCS summer monsoon onset since 1994 (Kajikawa and Wang, 2012) is primarily attributed to a mean state change in the Pacific basin characterized by a grand La Niña–like pattern. Meanwhile, less attention has been paid to the withdrawal phase of summer monsoon/SRS. According to Xavier et al. (2007), the La Niña (El Niño) condition favors a delayed (an early) withdrawal of the Indian summer monsoon.

In the ICP, Zhang et al. (2002) demonstrated that warm sea surface temperature (SST) anomalies in the western Pacific and cold SST anomalies in the central–eastern Pacific in the preceding winter–spring correspond to an early SRS onset. Nevertheless, previous researches have not included reference to the eastern part of the ICP.

The summer monsoon onset (SMO) is generally considered to be the key characterizing

indicator of the seasonal transition between dry and intense rainy season in the Indochina

Peninsula (ICP), and the Asian monsoon region in general. However, the SMO have not been

fully investigated over the eastern ICP in previous literatures. For example, Matsumoto (1997)

and Wang and LinHo (2002) demonstrated that the central coastal plain of Vietnam (CCV)

exhibits a non-typical pattern of summer monsoon rainfall, in which its maximum rainfall

belongs to the autumn regime. In the CCV along which the mountain ranges are mostly oriented

northwest–southeast with elevations over 1000 m (see Fig. 2.1), a strong mountain shadow effect

causes relatively dry summer, this unique characteristic causes substantial regional differences

(26)

3

compared to adjacent regions during the rainy season (Wang and LinHo, 2002). The westerly summer monsoon begins to retreat in September and the East Asian winter monsoon (EAWM) subsequently appears around October (Matsumoto, 1997), bringing north-easterly surface winds that cause orographic rainfall on the windward coastal plain in the CCV (Chen et al., 2012a;

Nguyen-Le et al., 2015). A question as to when is the SMO in this region is naturally raised. In addition, to date, because of its localized nature, few studies have characterized the particular autumn rainy season (ARS) along the CCV, although it is of utmost importance for the patterns of agricultural, forestry, and fishery activities throughout the region. During the ARS, heavy rainfall events can cause severe flooding that directly affects the local agricultural and energy sectors as well as the economy and human society (e.g. Yokoi and Matsumoto, 2008; Chen et al., 2012a). Therefore, it is also essential and meaningful to depict the variations in the timing of seasonal transition of autumn rainfall in the CCV.

1.2. Main objectives

The primary objective of this dissertation is to gain a better understanding of the climatological aspects and further explore the variations at interannual and decadal time scales in the timing of seasonal transition of precipitation over the eastern ICP, focusing on their relationships with ENSO. In order to achieve the main objective of this study, extensive observational analyses were conducted addressing the following specific aims:

1. To examine the climatological summer monsoon and summer rainy season onset over different sub-regions of Vietnam, particularly over the CCV;

2. To determine the climatology and precursory signals associated with the interannual variability of both the SRS and ARS onsets over the eastern ICP;

3. To detect the decadal changes in the onset and withdrawal dates of the ARS along the

CCV, and to suggest the potential factors responsible for this change, if any.

(27)

4 1.3. Dissertation outline

This dissertation is a compilation of two published articles and a manuscript that is currently under minor revision; therefore, each of these independently written papers with specific objectives, which are subsets of the main aim, appears as a separate chapter of this dissertation.

Because of these, although the readers may find it inconvenient by repeatedly seeing each part of a scientific paper, one can immediately achieve a conclusion by just reading any of the succeeding chapters.

Chapter 2 focuses on investigating the climatological SMO date in Vietnam using the five- day (pentad) main observations at 54 meteorological stations across the country and the JRA25 reanalysis dataset for the 25-year period from 1979 to 2003. In addition to the outbreak of summer monsoon rainfall in other sub-regions, the commencement date of a relative summer dry season in the CCV is defined for the first time. Furthermore, large-scale convective activity and atmospheric circulation associated with the onset period are discussed. The results have already been published in the International Journal of Climatology (Nguyen-Le et al., 2014).

Chapter 3 deals with climatology and precursory signals associated with the interannual variability of the onset of the rainy seasons over an extended study region, the eastern ICP (8.5°–

23.5°N, 100°–110°E). The SRS and ARS onset dates are objectively determined for individual

years from 1958–2007. The climatological features of atmospheric circulation, convection and

the intraseasonal oscillation (ISO) activities associated to the onsets are presented. Then,

composite analyses are conducted between the early/late SRS (ARS) onset categories to

investigate the precursory signals in the preceding winter and spring (summer), and their

underlying processes are discussed. It is based on the article that has been published online in the

Journal of Climate (Nguyen-Le et al., 2015).

(28)

5

Chapter 4 presents the first attempt in depicting the changes in the onset and withdrawal of the ARS along the CCV on a decadal scale. It utilizes the CCV rainfall index represented by pentad area-averaged rainfall from 14 meteorological stations among the region for the period 1979–2007. The non-parametric Lepage test (Lepage, 1971) is used to detect the significant decadal change in the time series of the onset and withdrawal dates; then, the epochal differences in atmospheric conditions and ISO activities before and after the decadal shift are intensively analyzed to examine the potential factors that explain this change, if any. It is based on a manuscript that has been submitted for publication and currently conditionally accepted (under minor revision) in the International Journal of Climatology (Nguyen-Le and Matsumoto, 2015).

Chapter 5 provides the concluding remarks and potential future researches that were not

covered in this study; then, it is followed by the bibliographical references. The figures and

tables are separated from the text following the University’s guidelines; they can be found

eventually after the Bibliography.

(29)

6

Chapter 2. Climatological onset of summer monsoon over different sub- regions of Vietnam

2.1. Introduction

Over recent decades, the Asian monsoon has become a critical issue in many studies.

Research into the annual cycle of the Asian monsoon system suggests that the seasonal changes in atmospheric circulation are accompanied by corresponding changes in rainfall (Ramage, 1971). The Asian summer monsoon (ASM) is often referred to as the rainy phase of a seasonally changing pattern. Therefore, summer monsoon onset is a key indicator that characterizes the transition from dry to rainy seasons. In this respect, the summer monsoon onset date (SMOD) is considered to be same as the summer rainy season onset date. For example, Wang and LinHo (2002) used relative climatological pentad mean rainfall to describe the spatial–temporal structure of the Asian-–Pacific summer monsoon rainy season. The results showed that monsoon rainfall first increases in the South China Sea (SCS) in mid-May before extending to the northwest Pacific in early to mid-June. Zhang et al. (2002) studied the climatology and interannual variations of the summer rainy season onset over the Indochina Peninsula (ICP) by using the observed daily rainfall at 30 stations and the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis data recorded from 1951 to 1996. The climatological onset date was determined as May 9, with a standard deviation of 12 days. However, Matsumoto (1997) observed winter maximum rainfall along the central coastal plain of Vietnam, where the distinction between dry and rainy seasons is difficult to determine through rainfall data. Wang and LinHo (2002) also suggested that this area exhibits a non-typical pattern of summer monsoon rainfall.

However, appropriate determination of the monsoon onset must also depend on wind fields.

Orgill (1967) used wind charts recorded in 1936–64 to define the onset of the ASM over

(30)

7

Southeast Asia as the time at which lower tropospheric equatorial westerlies move northward into southern China during the months of May and June. The results showed that the mean onset date in Indochina was May 17 with a range of 33 days. Cheang and Tan (1988) defined the onset of the southwest monsoon over this region as the date on which when both 850 hPa and 700 hPa zonal wind components become positive and remain positive for at least 20 days. Wang et al.

(2004) used 850 hPa zonal winds averaged over the central SCS (5°–15°N; 110°–120°E) to determine the monsoon onset dates for the period 1948–2001. They determined that the earliest and latest onsets occur in the 25th pentad (P25: May 1–5) and the 34th pentad (P34: June 14–

19), respectively. However, Murakami and Matsumoto (1994) suggested that changes in prevailing winds are not always a proper indicator for onset. For example, Matsumoto (1997) showed that the 850 hPa winds over the northern ICP (around 20°–25°N) are already westerly during the pre-monsoon period. Prior to the summer monsoon, they are of mid-latitude origin, flowing around the southern periphery of the Tibetan Plateau (Murakami and Matsumoto, 1994).

Located in the eastern part of the ICP, Vietnam is a country with a tropical climate and

complex topography (Phan et al., 2009; Chen et al., 2012a). Given the strong influence of the

ASM system, many socio-economic sectors of Vietnam such as forestry, fisheries, and

agriculture are strongly dependent on monsoon activity. Therefore, it is of great importance to

understand the monsoon’s characteristics and its impacts on the country. However, substantial

knowledge gaps in monsoon activity over Vietnam remain. Recently, Pham et al. (2009) used

daily rainfall at six stations in southern Vietnam and 1000 hPa reanalysis zonal wind data to

determine monsoon onset dates for 1979–2004. It was shown that the mean onset date is May 12,

with the earliest onset occurring in 1979 (April 19) and the latest in 1993 (June 9). Nevertheless,

to the best of our knowledge, previous studies have not sufficiently investigated the detailed

climatological SMOD over various sub-regions of Vietnam. For example, only 12-year period

rain gauge data recorded in 1975–77 and 1979–87 were available in Matsumoto (1997). Zhang et

(31)

8

al. (2002) investigated monsoon onset variability over the ICP; however, their study did not include Vietnam. In addition, the analysis gridded data used in many studies (e.g. Murakami and Matsumoto, 1994; Qian and Lee, 2000; Wang and LinHo, 2002; Wang et al., 2004), are limited in capturing the complex monsoon and rainfall characteristics of Vietnam due to their coarse resolutions. Thus, the major objective of the present work is to depict such detailed features with a combination of in-situ and reanalysis gridded datasets.

The remainder of this chapter is organized as follows. The data used in this study is described in the next section. In Section 2.3, we present an investigation of the SMOD in Vietnam. Large-scale convective activity and atmospheric circulation associated with the onset period are discussed in Section 2.4. Finally, concluding remarks are given in Section 2.5.

2.2. Data

To accomplish the objective of this study, a 25-year dataset recorded in 1979–2003) was used. This dataset contains the following information:

1) Pentad (five-day) mean rainfall computed from daily rainfall (R), maximum temperature (Tmax), and minimum relative humidity (RHmin) at 54 meteorological stations in Vietnam provided by the Vietnamese National Hydro-Meteorological Service (VNHMS). The homogeneity of R was verified by Endo et al. (2009). When missing data in a particular pentad exceeded two days, the data from this pentad was removed from the calculation of the mean data.

2) Pentad mean zonal and meridional wind components (U, V), temperature (T), specific

humidity (Q), relative humidity (RH), and geopotential height (H) derived from the Japanese 25-

year reanalysis data (JRA25) (Onogi et al., 2007) at eight pressure levels (1000, 925, 850, 500,

400, 300, 250, and 200 hPa) and near-surface (2 m) on a 1.25° grid.

(32)

9

3) Pentad mean outgoing longwave radiation (OLR) satellite observations on a 2.5° latitude–

longitude mesh from the National Oceanic and Atmospheric Administration (NOAA) (Liebmann and Smith, 1996).

2.3. Onset of summer monsoon and summer rainy season over various sub-regions of Vietnam

2.3.1. Determination of the onset dates in the northern and southern regions

In previous studies, the SMOD in the ICP was determined by using various methods. In particular, some authors determined the onset by focusing on only rainfall (e.g. Ramage, 1971;

Tao and Chen, 1987; Matsumoto, 1997; Wang and LinHo, 2002; Zhang et al., 2002) or convective activity indicated by satellite observations (e.g. Tanaka, 1992; Murakami and Matsumoto, 1994). SMOD was also indicated by changes in the prevailing winds (e.g. Orgill, 1967; Wang et al., 2004; Li and Zhang 2009). Recently, a combination of rainfall (or convective activity) and wind field has been used (e.g. Qian and Lee, 2000; Ding and Yanju, 2001; Pham et al., 2009, Htway and Matsumoto, 2011).

However, the coarse resolutions of the current reanalysis and convective activity data present challenges in determining regional information. One should note that although Vietnam is a somewhat small country with only 329,560 km

2

of total area, its spatial classification of climate is complicated as a result of monsoon influence, heterogeneous topography, and latitudinal extent (Pham and Phan, 1993; Pham et al., 2009). For this reason, Vietnam generally could be divided into seven climatological sub-regions (Phan et al., 2009; Ho et al., 2011).

Hence, to better capture monsoonal characteristics at the sub-regional scale in Vietnam, rain-

gauge data is preferred. Additionally, rainfall information is typically more reliable than

observed surface wind, which is strongly affected by topography. Thus, the SMOD in Vietnam

was examined on the basis of the onset criterion proposed by Matsumoto (1997), which is the

(33)

10

first pentad when the mean rainfall exceeds the annual mean pentad rainfall in at least three consecutive pentads after lowering it in more than three consecutive pentads. To reduce noise amplitude, the rainfall data was smoothed with a (1–2–1) weighted mean (forward and backward) of three consecutive pentads. As mentioned previously, since this criterion is only based on rainfall, the SMOD here is also the onset date of the summer rainy season.

Figure 2.1 shows the onset dates at 54 meteorological stations across Vietnam. The earliest onset of the summer rainy season, in late April to early May (P24–P25), is located in the northwestern mountainous region. The onset occurred in early to mid-May (P26–P27) in the Red River and Mekong River deltas in northern southern Vietnam, respectively. Several highland stations in the northern and south–central regions have onset dates in early May (P25), which is approximately one pentad earlier than those of other stations located in the plain areas. However, whereas the northern and the southern regions have a rainy season under the influence of the summer monsoon, the seasonal march of rainfall in the central coastal plain is delayed until late autumn to early winter (September–November). This difference is attributed to orographic rainfall caused by highlands facing northeast, cold surges, and tropical disturbances over the SCS in early winter (Yokoi and Matsumoto, 2008; Chen et al., 2012a). Therefore, a different criterion is required to determine the onset date for this region.

Aside from these differences, the results for northern and southern Vietnam are in agreement

with previous studies. According to Matsumoto (1997), the inland region of Indochina

(Thailand) has an onset date in late April (P23–P24), which is earlier than that in the coastal

region (P26–P27). This earlier onset, known as pre-monsoon rain, occurs under a mid-latitude

wind system (Matsumoto, 1997; Kiguchi and Matsumoto, 2004). Ding (2004) summarized the

climatological onset dates of the ASM by dividing the onset process into four stages, beginning

in the central ICP in late April to early May and continuing eastward to the coastal area and the

SCS from mid- to late May. Moreover, Wang and LinHo (2002) showed that the earliest rainy

(34)

11

season in the Asian monsoon region begins in March over southeastern China. This onset is much earlier than the establishment of summer monsoon over the ICP, which may be induced by the southward intrusion of cold fronts (Wang et al., 2004).

2.3.2. The Foehn wind and onset in central coastal plain

Murakami and Matsumoto (1994) used the strong correlation between monsoon circulation and convective activity to introduce an objective monsoon criterion based on OLR data in which a threshold value of 240 W m

-2

was chosen to define the ASM onset date. Figure 2.2 presents the OLR and 850 hPa zonal wind latitude section, averaged between 105°E and 110°E. From mid- April to early May, the OLR begins to decrease and remains lower than 240 W m

-2

for at least 15 days (Fig. 2.2a). In particular, the SMOD in the northern and north–central plains of Vietnam (18°–22°N) was determined to occur near P26–P27, whereas the onset occurs much earlier, in late April (P23–P24), over the northernmost mountainous region (22°–24°N). In the southern area (8°–12°N), the SMOD is sharply defined at P26–P27. However, in the narrow coastal plain in central Vietnam (12°–18°N), where the distance between the mountains and the sea is narrow (50–150 km), it is apparent that the coarse resolution of the JRA25 and OLR data limits our ability to obtain an accurate onset date.

Over 8°–12°N, the 850 hPa zonal wind components become positive from P27, which is in good agreement with the onset date determined by OLR and rainfall (Fig. 2.1). Pham et al.

(2009) clarified that the climatic regime in this region is the best representative of monsoonal climates. On the contrary, the annual variations of winds show that westerlies dominate the 20°–

24°N zone from the beginning of the year and continue to expand southward to 15°N in late

winter before being replaced by the summer monsoon. Therefore, it appears that onset criteria

based on changes in low-level prevailing winds are of limited use north of 15°N, where the

summer monsoon is difficult to distinguish from earlier dominating mid-latitude westerlies.

(35)

12

As an effect of the ASM, the Foehn wind brings a dry rather than rainy season to the central coastal plain of Vietnam during the summer. The wet southwesterly ASM is blocked and lifted by the northeast-facing Truong Son mountain range (11°–20°N, 104°–109°E), which extends southeastward from north central Laos along Vietnam’s central coastal plain and its boundary with Laos and Cambodia to Mekong River delta (Fig. 2.1). As a result, the wet southwesterly ASM drops most of its moisture on the windward slopes located in Laos and Cambodia (Pham and Phan, 1993). Therefore, to investigate the onset date for the central coastal area, temporal changes in relative humidity should be considered. The averaged 13°–18°N time–longitude cross-section of the OLR, 2-m relative humidity (RH2m), and 850 hPa zonal wind (U850) fields are represented in Fig. 2.3, which shows a significant difference in convective activity between the windward and leeward sides of the Truong Son range at 106°E to 108°E (Fig. 2.3a). The OLR on the windward side begins at less than 240 W m

-2

from mid-April, whereas that on the lee side is under suppressed convective activity until P28.

Conversely, the distinction between dry and wet seasons over the lee side of Truong Son range is clearly displayed by using RH2m (Fig. 2.3b). The wetter period, with a relative humidity exceeding 80%, occurs from September to mid-May. The drier period, indicated by an RH2m of less than 80%, lasts from P29 to P53. Furthermore, wind fields at 850 hPa over central Vietnam exhibit considerable annual variability. The wetter period corresponds with winter monsoon northeasterlies and late-winter easterlies, whereas the drier period is associated with summer monsoon westerlies. Moisture is brought to the central coastal plain of Vietnam from the SCS by the winter monsoon easterlies. Conversely, drier conditions occur after the summer monsoon crosses the highland regions.

A peculiar feature of the lee side also appears in the 850 hPa zonal wind pattern (Fig. 2.3c).

The westerlies begin to cross over the Truong Son range in March, which agrees with results

reported by Li and Zhang (2009). However, the rapid evolution of westerlies, likely indicating

(36)

13

the onset of the summer monsoon, occurs only in late May. From P28, the westerlies break out to cover not only the SCS but also the Philippine region, coinciding with the development of convective activity. As previously noted, the relative humidity in the central coastal plain of Vietnam falls below 80% from P29, denoting the commencement of a drier period. In this study, the time of the drier period commencement, as an effect of the Foehn wind, was used as an onset indicator for the central coastal plain. According to this subjective approach, the onset of summer dry season caused by the ASM in this region, which differs substantially from the rainy season in the adjacent regions (Wang and LinHo, 2002), was determined to be P29. However, the lack of investigation on detailed mechanisms of the downslope Foehn wind in central Vietnam has limited our ability for explaining the specific process of this summer dry season onset. Thus, this topic deserves further investigation.

2.3.3. Moisture flux convergence and Western Pacific sub-tropical high activity

It is well known that the ASM plays the most important role in moisture transport and rainfall supply from the Indian Ocean to the ICP (Wang, 2006). The secondary moisture source originates from the SCS via the southern and western periphery of the Western Pacific Sub- tropical High (WPSH). On the contrary, subsiding stable air from the WPSH produces high pressure, low humidity, and cloudless conditions. The ICP region typically experiences dry, hot, and sunny weather, whereas the sub-tropical ridge fully dominates. Therefore, it is indispensable to study the moisture transport and activity of the WPSH prior to and following the summer monsoon onset dates. Figure 2.4 shows vertically integrated moisture flux convergence (VMFC) and geopotential height at 850 hPa from P22 to P31. The VMFC is given by

1000

0

p

VMFC qV

g dp

    , where , V , q, and g represent 1000 hPa, wind vector, specific humidity, and gravity acceleration, respectively.

p

1000

(37)

14

As indicated in Fig. 2.4, prior to the summer dry season onset in the central coastal plain (P29), the moisture flux divergence over Vietnam, the SCS, and the western Pacific area is highly associated with the WPSH, whereas moisture is likely transported eastward by mid- latitude westerlies from the Bay of Bengal (BOB) to inland Indochina, including Myanmar, Thailand, and Laos, and to southern China. It is important to note that the influence of the sub- tropical ridge, which is indicated by the position of the 1500 gpm contour, weakens in P24, leading to increased moisture flux convergence in the ICP. However, in the next pentad (P25), the dominance of the WPSH is established once more, represented by the westward extension of the 1500 gpm contour. Therefore, moisture flux divergence is again observed around northern and central Vietnam, which is similar to conditions that occur during P22.

A remarkable feature is the eastward retreat of the WPSH beginning in P26, which occurs simultaneously with the strengthening of moisture flux convergence in the ICP and BOB.

However, after the onset in the northern and southern plains in P27, moisture flux divergence occurs again around central Vietnam from P29 after the WPSH has already retreated to the western Pacific region. Meanwhile, just east of this divergence, an area of significant convergence is located in the SCS. Therefore, it is evident that the summer monsoon, rather than the sub-tropical ridge, causes dry conditions in the central coastal region of Vietnam.

2.3.4. Analysis of minimum daily relative humidity and maximum daily temperature

To better investigate the period and location of the significant changes in observations, the differences between two consecutive pentads from mid-April (P22) to early June (P33) in RHmin and Tmax at each station were computed and are displayed in Figs. 2.5 and 2.6, respectively. The Student’s t-test was applied to determine the statistical significance of these differences.

From P22 (mid-April) to P27 (mid-May), Tmax in the northwestern mountainous area does

not change significantly. According to the two-tailed Student’s t-test, increased rainfall in P24

(38)

15

results in a slight decrease in Tmax (not statistically significant) and a significant increase in RHmin at the 90% level, particularly in the northern region. However, Tmax (RHmin) in that region increases (decreases) significantly only in the one (two) succeeding pentad(s) before decreasing (increasing) in P28 (P27), indicating the onset of the summer rainy season in this area. In the case of the Mekong River delta, the most significant increase in RHmin is observed in P27 and P28, whereas Tmax begins a constant decrease from P26 as a result of the eastward retreat of the WPSH. This result is a clear signal of transition from the dry season to the rainy season brought by the summer monsoon.

Over the central coastal region, Tmax increases significantly in P23, P25, and P29, with the maximum increase occurring in P29. Conversely, RHmin statistically decreases considerably in P22 and P25 and gradually begins to decrease in P29. Although such increasing (decreasing) characteristics are quite similar, their causes differ. As mentioned in Section 2.3, the increase of Tmax and decrease of RHmin before P26 is clearly under the influence of the WPSH, whereas the most significant changes in both Tmax and RHmin, in P29, occur only in the central coastal plain and after the monsoon onset in all other sub-regions.

2.4. Large-scale convective activity and atmospheric circulation associated with the onset period

2.4.1. Convective activity and low-level flow pattern

To investigate the large-scale characteristics of convective activity and low-level atmospheric circulation, the OLR distribution and wind field at 850 hPa during P22–P31 over the Indian, Indochina, and SCS areas were examined (Fig. 2.7). The threshold of 240 W m

-2

was chosen to indicate strong convective activity.

Prior to the onset period, strong convective activity in P22 appears only over the southern

BOB, the Indonesian Maritime Continent, and the equatorial region of the western Pacific. From

(39)

16

P23, a lower OLR region extends from low latitudes to the ICP. Meanwhile, two suppressed convective-activity areas are located over the BOB and the SCS, corresponding to the presence of mid-latitude westerlies and the WPSH. During the period around the onset (P26 to P29), an extension of the lower OLR region is observed first in the ICP, then over the BOB and the SCS.

From P28, the SCS region begins to exhibit an OLR lower than 240 W m

-2

. In contrast, convective activity over the western BOB remains weak until early June. In P31, strong convection covers nearly the entire region, from the BOB to the SCS and the Philippines.

However, the northern Indian subcontinent remains a high-OLR area. In summary, it is shown that the remarkable seasonal changes in convective activity occur first in the ICP, extend eastward to the SCS, then westward to the BOB, and finally over inland India. This result is consistent with those reported by Matsumoto (1992, 1997), Zhang et al. (2002), Ding (2004), and Kiguchi and Matsumoto (2004).

The prominent wind system at 850 hPa over northern India, the BOB, and the ICP prior to the onset is classified as mid-latitude westerlies. On the contrary, the easterly trade wind related to the WPSH dominates the pre-onset climatic regime over the southern ICP and the SCS. The eastern region of the ICP, Vietnam, is located along the boundary between these wind systems.

In mid-April, the cross-equatorial flow becomes westerly under the effect of Coriolis force, and it is confined to the equatorial Indian Ocean. Over time, the strong westerlies expand from the equator to cover the southern BOB and to reach the ICP and the SCS in P27 and P28, respectively. The mid-latitude westerlies retreat northward, and the southeasterly trade wind retreats eastward to the western Pacific Ocean. From P29, the summer monsoon circulation fully controls the ICP and the SCS area.

To further elaborate on the evolution of this seasonal change in the atmosphere, the

difference between two consecutive pentads in OLR and wind field from P22 to P31 is shown in

Fig. 2.8. The most salient feature is the nearly continuous stronger evolution of westerly winds

Fig.  2.1. Topography  height  (gray  shaded;  m)  and  climatological  summer  monsoon  onset  date  (colored  dot;  pentad  number),  following  the  definition  of  summer  monsoon  onset  date  by  Matsumoto (1997), at 54 stations in Vietnam during a 2
Fig. 2.3. The 13°-18°N time-longitude sections of pentad mean (a) OLR (W m -2 ), (b) the wind  vector at 850 hPa (m s -1 ) and relative humidity at 2 m (%, b) and (c) zonal winds at 850 hPa (m s
Fig. 2.6. Same as Fig. 2.5 but for the maximum temperature at 2 m (°C)
Fig. 2.9. Vorticity (10 -6  s -1 ) and wind vector (m s -1 ) at 850 hPa in (a) April 15-May 15 and (b)  May 16-June 15, averaged over 25 years (1979-2003)
+7

参照

関連したドキュメント

The only thing left to observe that (−) ∨ is a functor from the ordinary category of cartesian (respectively, cocartesian) fibrations to the ordinary category of cocartesian

Eskandani, “Stability of a mixed additive and cubic functional equation in quasi- Banach spaces,” Journal of Mathematical Analysis and Applications, vol.. Eshaghi Gordji, “Stability

Finally, we give an example to show how the generalized zeta function can be applied to graphs to distinguish non-isomorphic graphs with the same Ihara-Selberg zeta

She reviews the status of a number of interrelated problems on diameters of graphs, including: (i) degree/diameter problem, (ii) order/degree problem, (iii) given n, D, D 0 ,

It is suggested by our method that most of the quadratic algebras for all St¨ ackel equivalence classes of 3D second order quantum superintegrable systems on conformally flat

This paper develops a recursion formula for the conditional moments of the area under the absolute value of Brownian bridge given the local time at 0.. The method of power series

It turns out that the symbol which is defined in a probabilistic way coincides with the analytic (in the sense of pseudo-differential operators) symbol for the class of Feller

Then it follows immediately from a suitable version of “Hensel’s Lemma” [cf., e.g., the argument of [4], Lemma 2.1] that S may be obtained, as the notation suggests, as the m A