Octoberof2008 Journal the Meteorological Society of Japan, Vol. HAMADA 86, No. 5, J.-I. pp.et593−611, al. 2008
593
Differences of Rainfall Characteristics between Coastal and Interior Areas of Central Western Sumatera, Indonesia HAMADA Jun-Ichi Institute of Observational Research for Global Change (IORGC), Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka, Japan
Manabu D. YAMANAKA Institute of Observational Research for Global Change (IORGC), Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka, Japan Graduate School of Science, Kobe University, Kobe, Japan
Shuichi MORI Institute of Observational Research for Global Change (IORGC), Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka, Japan
Yudi Iman TAUHID and Tien SRIBIMAWATI Agency for the Assessment and Application of Technology (BPPT), Jakarta, Indonesia (Manuscript received 17 November 2004, in final form 1 June 2008 )
Abstract Pentad rainfall data from 46 stations in West Sumatera Province, Indonesia in 1992 reveal temporal and spatial variations of rainfall in coastal and inland regions. Distinct contrasts in rainfall amounts and variations are evident between the coastal and inland mountainous regions. Locality index is defined by pentad rainfall time series in coastal and inland regions to classify characteristic rainfall distribution patterns into four types: coastal-, inland-, active-, and inactive-type. In 1992 the coastal- or active-type rainfall patterns tend to appear throughout the year, whereas the inland-type rainfall pattern tends to appear during the Southern Hemisphere summer. The inactive-type tends to appear during the Southern Hemisphere winter. Correlation and composite analyses suggest that lower tropospheric zonal wind variations over central western Sumatera tend to intensify the contrast between coastal and inland rainfall distribution. The correlation coefficient between locality index and zonal wind of 850 hPa NCEP/NCAR reanalysis is 0.77 (99% confidence level) for characteristic rainfall distribution patterns during 1992−1993. This implies that rainfall in the windward side of steep mountain is relatively greater than that of leeward side with increasing wind speed. The occurrence of coastal-type pattern is relatively well related with lower tropospheric westerly wind. Furthermore, although the number is limited, diurnal variations of convective activity are clear when active-type rainfall appears. Corresponding author: Hamada Jun-Ichi, Institute of Observational Research for Global Change (IORGC), Japan Agency for Marine-Earth Science and Techno logy (JAMSTEC), 2-15 Natsushima-Cho, YokosukaCity, Kanagawa 237-0061, Japan E-mail:
[email protected] ©2008, Meteorological Society of Japan
594
Journal of the Meteorological Society of Japan
1. Introduction Indonesia is a maritime continent; its complex topography includes many large and small islands that are surrounded by warmer seawater (Ramage 1968). Convection over Indonesia is generally more active than over other tropical regions, except for the dry season in an El Niño year (e.g., Meehl 1987; Ropelewski and Halpert 1987). Aspects of seasonal rainfall variations over Indonesia, including its climatological features and interannual variability related to ENSO (El Niño and Southern Oscillation) events, have been described since the late 19th century. (see brief review by Allan 1991; McBride 1999; Hamada et al. 2002). The temporal and spatial distributions of rainfall provide important indices to help characterize any tropical climatology. In addition, convection controls large-scale motions through the release of latent heat. Hamada et al. (2002) investigated the spatial characteristics of seasonal and interannual rainfall variations using pentad rainfall data. They noted a difference in the characteristics of the rainy season, including onset, duration and rainfall amount, between the Jawa1 Sea and Indian Ocean sides in the vicinity of Jawa Island. Chang et al. (2005) also classified the annual cycles of rainfall over land area of the southeast Asia and maritime continent based on historical station rainfall data, and identified the monsoon regimes by Tropical Rainfall Measuring Mission (TRMM) rainfall measurements. They suggested that the annual cycle is dominated largely by interactions between complex terrain and a simple annual reversal of surface monsoonal winds throughout all monsoon regions from Indian Ocean to the South China Sea and the equatorial western Pacific. Yasunari (1981) showed that biennial rainfall variations dominate interannual variability over the central part of Indonesia. In contrast, variations with different periods occurred over northern and southern Jawa. Haylock and McBride (2001) showed coherence in the interannual variability of seasonal rainfall anomalies in Indonesia. High correlations with the southern oscillation index (SOI) appeared over a wide region before the rainy season (September through November). During the rainy season, 1 In this paper the standard spellings in the Indonesian language are used for geographical names such as Jawa, Sumatera, Kalimantan, Sulawesi and Papua (corresponding to Java, Sumatra, Borneo, Celebes and New Guinea in English or Dutch spellings).
Vol. 86, No. 5
such correlations with rainfall (December through February) were only seen over a limited region. This suggests that the seasonality of the relationship between rainfall and the SOI was at a regional scale over Indonesia and adjacent localities (e.g., Nicholls 1981, 84; Ropelewski and Halpert 1987; Hamada et al. 2002). Tropical convection exists at scales from cloud clusters to the Madden−Julian oscillation (MJO or the tropical 30−60 day oscillation) (Nakazawa 1988; Madden and Julian 1971, 1972, 1994). The tropical 30−60 day oscillation is a result of eastward-propagating low-frequency circulation anomalies and associated large-scale convection. Many studies have noted that the tropical 30−60 day oscillation was suppressed over Indonesia (e.g., Murakami et al. 1986; Rui and Wang 1990; Zhu and Wang 1993). Nitta et al. (1992) showed that cloud disturbances moving eastward from the Indian Ocean to the western Pacific were weakened or modified in the western part of Indonesia, possibly because of the blocking effects of Sumateran topography. These observations suggest that surface topography plays a role in determining the distribution and variation of rainfall and convection over Indonesia. However, detailed observational studies of the location of rainfall variabilities, especially for intraseasonal variability related to tropical 30−60 day oscillations (MJO) or cloud clusters have not yet been carried out. Therefore, this study focused on the locations of rainfall variability and associated variations of wind and convection in coastal and inland mountainous regions of central western Sumatera, at the western edge of Indonesia. In this region, some meteorological studies have been done based on atmospheric radar and satellite observations. RISH (Research Institute for Sustainable Humanosphere, Kyoto University) has recorded UHF and VHF radar observations that describe the vertical structure of the tropical atmosphere at Kototabang since 1998 and 2001, respectively (cf. Widiyatmi et al. 2001; Renggono et al. 2001; Murata et al. 2002; Fukao et al. 2003; Seto et al. 2004). Renggono et al. (2001) investigated the diurnal variation of precipitating clouds classified by UHF radar, and Murata et al. (2002) and Seto et al. (2004) examined the relationship between wind and precipitation using UHF/VHF radars, rawinsonde and surface observational data. Recently, intensive rawinsonde and weather radar observation was done around Kototabang to examine the vertical coupling processes
October 2008
HAMADA J.-I. et al.
of equatorial atmosphere (cf. Fukao 2006). GPS and intensive rawinsonde observation also showed diurnal variations of atmospheric water vapour at around Kototabang (Wu et al. 2003; Sasaki et al. 2004). However, these studies used mainly single station data in an inland region. Thus, the widerscale significance of the rainfall variations at each location is limited, and their relevance to the larger picture remains, unknown. The diurnal cycle of rainfall/convection over Sumatera was investigated using satellite data (Mori et al. 2004; Sakurai et al. 2005); however, the relationship between convection and rainfall variations has not been described. Thus, this study seeks to explain observations of rainfall variations in coastal and inland mountainous regions of central western Sumatera, and to discuss the relationship among rainfall locations, wind variations and convective activities. Observational data used in this study are described in Section 2. Data analyses that focus on temporal and spatial rainfall variability are described in Sections 3 and 4. In Section 5, a relationship between rainfall locations and lower tropospheric wind variability is described. Section 6 provides some discussions on a relationship between rainfall locations and diurnal variations. Section 7 presents the conclusions of this paper.
2. Data In order to analyze phenomena with intraseasonal time scales, which is essentially important to consider actual rainfall characteristics in the Indonesian maritime continent (as will be shown later), we need to collect (pentad or) daily rainfall data. Such daily rainfall data in West Sumatera Province were collected and compiled for several years after 1980s, which were provided in book form (BMG 1993) by Sicincin station of the Indonesian Meteorological and Geophysical Agency (BMG: Badan Meteorologi dan Geofisika) in West Sumatera. However, because the continuities of the data at several stations for many years have not been sufficient to show spatial rainfall distribution and its interannual variations, we mainly used the data in 1992 in which the continuity and spatial density of rain gauge were highly better than the others. Although the data book of 1992 included records from 112 stations, only stations with complete time series of daily rainfall were used in this study to make pentad (5-day) rainfall amount time series2. Stations with anomalously small or large rainfall compared to climatological means, as determined
595
by LMG (1973) and Pawitan (1990), or in comparison with surrounding stations in 1992, were excluded. Figure 1 shows the geographical distribution of 46 rain gauge observation stations used in this study (see Table A1 in the Appendix). About 1/3 of the stations were on the western side of steep mountains; the remainder were in the inland mountainous region (above 400 m altitude) or on the eastern side of mountains. Minimum sensitivity of manual rain gauge observation was 1 mm except for Tabing station (0°53´S, 100°21´E, 14 m altitude, and 0.1 mm sensitivity). Though number of stations was limited (total 34 stations: 20 inland, 13 coastal and 1 Siberut Island station) and the spatial distribution was localized, we also used the data in 1993 in order to investigate the relationship between rainfall locality and dominant wind direction as described in Section 5. Only 17 stations (11 inland; 6 coastal) had continuous data during 1992−1993, and stations in the central basin and the south-eastern edge (relatively flat) of inland region had lacked in 1993. Thus, we will mainly show observational results based on the data of 1992. We also collected daily rainfall data at the coastal station Tabing from January 1990 through June 1998 to determine characteristic temporal rainfall variations. The following data helped reveal the relationships among rainfall, cloud and wind variations around Sumatera Island. • Satellite cloud data This study used GMS 3-hourly IR-TBB (black body temperature) data from 1992. The horizontal resolution was 1.0° in longitude and latitude. Linear interpolation was used to compensate for irregular time-sampled data and missing data to provide complete and regular 3-hourly data. Because TBB indicates both cloud top and surface temperature, an index of convection, Ic, demarcated clouds. Ic is defined as
2 The data book reported daily rainfall. BMG and other institutes recorded rainfall amounts each morning around 0700 local standard time (corresponds to 0000 UTC), and recorded that as the rainfall data for the day. This data book included daily rainfall amounts from 31 December, 1991 to 30 December, 1992.
596
Journal of the Meteorological Society of Japan
Vol. 86, No. 5
Fig. 1. Surface topography of central western Sumatera and location of rainfall observation stations. Altitude contour interval is 400 m. Dot (•) and plus (+) indicate coastal and inland stations, respectively. See Section 4 for details. Annual rainfall distribution in 1992 is also shown as circles. Circle area shows rainfall amount, and black (gray) color indicates rainfall of equal or greater than (less than) 3,000 mm.
250 − TBB (TBB < 250 ) Ic = 0 (TBB ≥ 250 )
where 250 K is the air temperature near 8,000 m above sea level over central western Sumatera. Ic denotes deep convective clouds with tops exceeding 8,000 m. • Objective analysis data This study also used 6-hourly 850 hPa wind and mixing-ratio data from the NCEP/NCAR reanalysis (Kalnay et al. 1996) to examine relationships between wind variations and rainfall localities. The spatial resolution is 2.5° in longitude and latitude. Operational rawinsonde observation data at Tabing station were
also collected from BMG. However, because the number and continuity of the data were not sufficient, we did not use rawinsonde data. The conditions of operational rawinsonde observation in Indonesia were described in Okamoto et al. 2003. Pentad mean time series of satellite cloud data and objective analysis data were calculated. In this study, we actually used the data during December 31, 1991 and December 30, 1992 to be corresponded to the period of rainfall data. We made 73 pentads data for a year; the first pentad is form December 31, 1991 to January 4, 1992, and the last pentad is from December 26 to 30, 1992. Data on February 29, 1992 has been included into pentad 12 (February 24 to 29, 1992).
October 2008
HAMADA J.-I. et al.
3. Interannual and intraseasonal rainfall variations at Tabing Over central Sumatera Island, climatological variations in seasonal rainfall were dominated by a semi-annual cycle; the maximum rainfall occurred between September and February with abundant rainfall almost all year round (e.g., Schmidt and Ferguson 1951; LMG 1973; Hamada et al. 2002). In an individual year, however, intraseasonal rainfall variations were (of which the dominant temporal scale is close to one month) clearer than seasonal variations (e.g., Inoue and Nakamura 1990; Nakamura et al. 2001). Figure 2 shows the interannual variability in monthly rainfall at the coastal station Tabing during 1990−1997. Monthly rainfall amount generally exceeds 100 mm; however, monthly or 31-day running sum of rainfall is less than 100 mm in several periods, notably August through October of 1994 and 1997. Saji et al. (1999) described a dipole pattern in the variability, manifested as high sea-surface temperatures in the western Indian Ocean and anomalously low sea-surface temperatures off Sumatera with accompanying wind and precipitation anomalies. Drought months occurred during an Indian Ocean dipole event and also an El Niño event. Annual rainfall amounts for 1992 are relatively large compared to 1990−1997. In contrast, rainfall amounts in West Sumatera Province are smaller than the climatological average during
597
1931−1960, especially in the mountains, as will be noted in Section 4. Daily rainfall data at Tabing were statistically analyzed. Number of rainfall days (≥ 1.0 mm day−1) was 1,349 (43.5% of the total observation days). Average daily rainfall was 24.1 mm day−1 with a standard deviation of 31.7 mm day−1. Maximum daily rainfall was 238.9 mm day−1, on 4 April, 1992 (in pentad 19 of this analysis year). Rain totaling greater than 60 mm fell on about 10% of the days; these heavy rainfall days (≥ 60 mm day−1) collectively accounted for more than 40% of the total rainfall. The frequency of heavy rainfall is an important factor in seasonal and interannual rainfall variability. The frequency distributions of rainy and dry days suggest that rainy and dry days tend to be clustered in groups (e.g., McGregor and Nieuwolt 1997). Thus, an analysis of dry spells was undertaken to investigate seasonal and interannual variations in rainfall. Figure 3 shows the temporal distribution of dry spells. A dry spell was defined as more than 3 days with daily rainfall amount less than a threshold value, which was set at 1.0 mm day−1. Thus, the figure shows periods of continuous dry days. Dry spells are mostly absent from March−April and November−December. These are the months of the climatological rainfall maxima. Conversely, August through October showed interannual variations in dry spells. Dry days occur during these months in 1994 and 1997, whereas
Fig. 2. Monthly rainfall variations at Tabing (0°53´S, 100°21´E, and 14 m altitude) during 1990−97 (bar). Thick solid line shows 13-month running average of monthly rainfall. Dotted and dashed lines show 13-month running average of the southern oscillation index and the Indian Ocean dipole mode index, respectively. Plus (+) shows a month which has the day that 31-day rainfall (31-days running sum) is less than 100 mm.
598
Journal of the Meteorological Society of Japan
Vol. 86, No. 5
Fig. 3. Dry spells in daily rainfall time series at Tabing. Black bar shows rainfall cessation period, i.e. when the daily rainfall amount was less than 1.0 mm day−1 for more than 3 days.
rainfall occurs frequently during these months in other years. Dry spells frequently appear around February and during June and July, but heavy rains (≥ 60 mm day−1) also fall during these months (not shown). Some months with relatively small rainfall totals (less than 200 mm), such as November 1996 and August 1993, naturally lack days with heavy rainfall. However, continuous dry periods are not observed during these months either. Intraseasonal variations in the recurrence interval of heavy rainfalls are important for explaining interannual variations in rainfall amounts.
4. Inhomogeneity of rainfall variations in coastal and inland regions 4.1 Spatial and temporal rainfall variations Figure 1 shows annual rainfall distribution in West Sumatera Province. In general, an east−west (inland−coast) contrast related to surface topography dominates the annual rainfall distribution. Annual rainfall is generally more than 3,000 mm on the western side of the mountains, and 1,000−3,000 mm in the inland basin or on the eastern side of the mountains. Annual rainfall exceeds 4,000 mm at some stations around Tabing. Annual rainfall total occurs also high in the inland region of the northern slopes of Mt. Kerinci (3,800 m MSL), the highest mountain in Sumatera, that may be respon-
sible for abundant rainfall on the western side of steep mountains and around high mountains. This rainfall pattern in 1992 shows the same inlandversus-coast pattern in the climatological distribution for 1931−1960 from Pawitan (1990). However, rainfall totals in the inland basin are smaller than during the period 1931−1960 (2,000−3,000 mm). Pentad rainfall distributions have locations with maxima at intraseasonal time scales. In order to focus on the contrast between the eastern (inland mountainous) and western (coastal) sides of steep mountains, stations were classified into two regions based on their locations relative to the mountain ridge: coastal and inland. Stations near the ridge were classified by annual rainfall amount. As shown in Fig. 1, annual rainfall generally exceeds 3,000 mm on the western side of the mountains. Stations with rainfall approaching or exceeding 3,000 mm year−1 in 1992 were classified as coastal. Table A1 shows station classifications. There were 15 coastal and 30 inland stations3. Figure 4a−b shows a time series of pentad rainfall amounts averaged for coastal and inland stations, respectively. 3 One station (Muara Siberut (26a, 1°35´S, 99°12´E, and 2 m altitude)) is on Siberut Island, about 150 km from Sumatera. This analysis excluded Siberut data, focusing instead on the mountains of Sumatera.
October 2008
HAMADA J.-I. et al.
Fig. 4. Time series of pentad rainfall amounts in 1992 averaged for the stations in (a) the coastal region (15 stations) and (b) the inland region (30 stations). Coefficients of variation (CV, spatial variability) are indicated with a cross (×). Horizontal solid (dashed) line shows annual mean pentad rainfall amount (one standard deviation from annual mean). (c) Time series of locality index (bar). Solid and dashed lines show normalized anomaly rainfall variations in the coastal and inland regions, respectively. Triangle (△), circle (○), square (□) and plus (＋) show pentads when coastal-, inland-, active- and inactivetype spatial rainfall distribution patterns appeared, respectively.
599
600
Journal of the Meteorological Society of Japan
Figure 4 suggests that heavy rainfall occurs at about 20- to 60-day intervals in the coastal regions. Lighter rainfall (inactive rainfall period) continues for a few pentads. By comparison, seasonal variations with relatively continuous rainfall in Southern Hemisphere summer (November through March) dominate the inland region. Variations with periods of 10−20 days appear mainly during the Southern Hemisphere summer. Intraseasonal (20−60 day) variations with maxima simultaneous with coastal stations, or with a lag of a few pentads, are found. The simultaneous correlation coefficient between inland and coastal rainfall variability was 0.46 (99% confidence level). However, intraseasonal rainfall variations also show opposite tendencies between inland and coastal regions, especially during January and February. Figure 4a−b also shows the coefficient of variation (CV) of rainfall for evaluating spatial rainfall variability. CV was defined as the ratio of the standard deviation of pentad rainfall to the average pentad rainfall (e.g., Kuraji et al. 2001 used this definition for indicating rainfall locality in a mountainous region of Thailand). A small CV value reflects a spatially uniform rainfall distribution. Seasonal variations in CV are not clear both in coastal and inland regions, but CV is relatively small in and around November when heavy rainfall (more than average) normally occurs. Figure 3 shows that clear days are rare in November and December of 1990−1997 at Tabing. These results suggest that rainfall in and around November is continuous and uniform over West Sumatera Province. Heavy rainfall also occurs with relatively small values of CV on intraseasonal time scales. 4.2 Characteristic spatial distribution patterns of rainfall We defined the locality index to measure the locality of rainfall variations based on the time series of pentad rainfall amounts in coastal and inland regions. It is calculated as follows: P Locality Index = diff , SD
where Pdiff = (normalized anomaly of average rainfall for coastal region) −, (normalized anomaly of average rainfall for inland region) SD = standard deviation of Pdiff .
Vol. 86, No. 5
The index is the difference between inland and coastal rainfall variations and was calculated by subtracting the normalized anomaly rainfall varia tions for inland regions from those of coastal regions, and dividing the result by the standard deviation for each pentad (shown in Fig. 4c). Four types of rainfall distribution pattern were defined as follows: • Coastal-type: Rainfall was significantly active in coastal regions, but not in inland regions (Locality index ≥ 1.0, and a normalized anomaly rainfall in coastal regions of ≥ 1.0). • Inland-type: Rainfall was significantly active in inland regions, but not in coastal regions (Locality index ≤ −1.0, and a normalized anomaly rainfall in inland regions of ≥ 1.0). • Active-type: Rainfall was significantly active over all of West Sumatera Province (Absolute value of locality index < 1.0, and a normalized anomaly rainfall in both coastal and inland regions of ≥ 1.0). • Inactive-type: Rainfall was small over all of West Sumatera Province (Absolute value of locality index < 1.0, and a normalized anomaly rainfall in both coastal and inland regions of ≤ −1.0). Table 1 shows the occurrence time (pentad) of the characteristic rainfall distribution patterns. The coastal- and overall active-type rainfall patterns occur 6 and 3 times, respectively, with intraseasonal time scales (about 30−60 days) throughout the year. Besides, the inland rainfall pattern occurs 6 times most frequently from November to March (in Southern Hemisphere summer). The occurrence interval is shorter (about 10−20 days) than that of coastal- or active-type rainfall. Rainfall over West Sumatera Province is relatively scant (the inactive-type occurs 4 times) from June to August (in Southern Hemisphere winter).
Table 1. Pentad for occurrence of characteristic rain fall distribution patterns (coastal-, inland-, activeand inactive-type) in 1992. See text for further information. Distribution pattern
Pentad number
Coastal-type Inland-type Active-type
8, 26, 27, 32, 48 and 60 4, 6, 14, 63, 66 and 70 19, 54 and 64
Inactive-type
35, 36, 44 and 50
October 2008
HAMADA J.-I. et al.
Figure 5 shows composite of the characteristic rainfall patterns defined above. The coastal type is recognizable as rainfall is heavy over the coastal region, especially around Tabing, though moderate rainfall occurs over the inland regions. With a different typical rainfall distribution, the inland type, heavy rainfall appears in the inland region, especially in the eastern edge of the mountains, and they might be classified into two distribution patterns. One typical rainfall distribution pattern is that rainfall is limited in the inland region and a very few rainfall appears in the coastal region in January. Another pattern is that relatively uniform distribution is found all over West Sumatera Province mainly during November and December. For the active-type, the rainfall distribution pattern is
601
similar to the annual rainfall distribution pattern (see Fig. 1). In this study we used pentad rainfall amounts data to examine intraseasonl variations (more than a few pentads) of rainfall localities. We also calculated the locality index and the occurrence of characteristic rainfall distribution pattern with the same definition for daily rainfall amounts data in order to investigate day-to-day rainfall variations within a pentad. In each pentad for the characteristic rainfall types, daily rainfall distribution pattern is generally coincident with the pentad one. As for the coastal- or the inland-type appears, each pentad does not consist of daily rainfall distribution type with the opposite tendencies. On the other hand, a few pentads include an active type daily rainfall dis-
Fig. 5. Composite distribution of pentad rainfall amount averaged for pentads in the coastal, inland, active, and inactive rainfall types, respectively. Circle area indicates pentad rainfall amount.
602
Journal of the Meteorological Society of Japan
tribution pattern in inland- (pentads 14 and 63) or coastal-type (pentads 26 and 32), thus the rainfall distribution tends to be uniform. Based on these results, characteristic pentad rainfall distribution pattern basically consists of similar type of daily rainfall distribution within the pentad. Figure 6 shows the relationship between the rainfall distribution pattern and CV. CV is generally small when average rainfall totals are large. However, CV is large (> 1.0) during coastal-type rainfall, even where high pentad rainfall amounts occur. Such a large value of CV occurs because the heaviest rain falls only in the coastal regions. In contrast, CV is small and rainfall in West Sumateran Province is large when rainfall is dominant in the inland region, except during pentads 4 and 6. Rainfall totals in inland regions are generally smaller than coastal totals, so relatively uniform rainfall distribution is characteristic for an inland-type rainfall pattern. When rains are abundant in both inland and
Fig. 6. Relationship between coefficient of variation (CV) and pentad rainfall amount averaged for 46 stations in West Sumatera Province. Triangle (△), circle (○), square (□) and plus (＋) indicate the typical rainfall distribution pattern, i.e. coastal, inland, active and inactive types, respectively. The pentad number is also indicated. Vertical solid (dashed) line shows annual mean pentad rainfall amount (one standard deviation from annual mean).
Vol. 86, No. 5
coastal regions, CV has values between those of the coastal and the inland rainfall periods. Finally, CV is large during inactive rainfall periods.
5. Comparison with rainfall variability and lower tropospheric wind variations 5.1 Overall features of circulation and convective activities Figure 7 shows longitudinal and temporal variations of convection and lower tropospheric horizontal wind near the Equator in 1992. Intraseasonal variations of convection are dominant over the eastern Indian Ocean and the western part of Sumatera (80°E−100°E), whereas the magnitude and intraseasonal variations of convection are weaker over the central maritime continent sector (100°E− 120°E) than over the Indian Ocean, except during the Southern Hemisphere summer. Intraseasonal variations of convection appear over the central western Sumatera from January through April, and also from May to August when the ITCZ (Inter Tropical Convergence Zone) is in the Northern Hemisphere (10°−15°N). Intraseasonal variations are also recognized during September−December, but convection is rather continuously active. According to the index for monitoring the MJO based on a pair of empirical orthogonal functions of the combined fields of near-equatorially averaged 850 hPa zonal wind, 200 hPa zonal wind, and OLR data (cf. Wheeler and Hendon 2004), a few MJO events are clearly recognized around Sumatera in 1992 (by calculating the pentad averaged amplitude and phase of the MJO from the pair of daily principal components provided by the Australian Bureau of Meteorology, not shown). The eastward propagating MJO convections with accompanying lower westerly wind cross over Sumatera Island in April −June at around pentads 19, 25 and 32 with sufficient amplitude of convection activity. The activeor coastal-type rainfall distribution pattern occurs when these MJO clearly propagate over Sumatera (see Table 1). In addition, the heaviest rain (238.9 mm day−1) falls at Tabing during 1990−1998 in pentad 19 (the active-type rainfall pattern appeared). In Southern Hemisphere summer, though convection activities are active over Sumatera and the central maritime continent, the MJO convection is located in western Hemisphere rather than maritime continent. Besides, the ITCZ migration over the SumateraMalay Peninsula (along 100°E) in 1992 shows similar climatological features described by Murakami
October 2008
HAMADA J.-I. et al.
603
Fig. 7. Time-longitude cross-section of Ic (shade) and pentad mean 850 hPa horizontal wind (arrow) near the Equator (3°S−3°N, average). Time series of pentad mean zonal wind (thick solid) and meridional wind (thick dashed) over central western Sumatera (100°E, the Equator), and locality index (bar) are shown in the right-hand figure. Triangle (△), circle (○), square (□) and plus (＋) show the pentad when coastal-, inland-, active- and inactive-type spatial rainfall distribution patterns appeared, respectively.
and Matsumoto (1994) (not shown). Active convection occurs over a 15°S−20°N latitudinal range. The convection region is limited near the Equator and in the Southern Hemisphere during the Southern Hemisphere summer, whereas active convection occurs over a wider latitudinal range in the
Southern Hemisphere winter. 5.2 A relationship between locality index and zonal wind variations In order to investigate the relationship between rainfall variability and the lower tropospheric wind
Journal of the Meteorological Society of Japan
variations over central western Sumatera, we used 850 hPa wind of NCEP/NCAR reanalysis data. Murata et al. (2002) showed that lower-tropospheric zonal wind variations of UHF radar at Kototabang in the inland region of central western Sumatera and NCEP 850 hPa zonal wind variations at the nearest grid point to Kototabang (100°E, the Equator) during September−October 1998 were quantitatively agreed well with more than a few days time scales. Thus, the NCEP data would have a potential to analyze temporal wind variations in this region, but may not be sufficient to represent a magnitude of wind speed. The right panel of Fig. 7 and Fig. 8 indicates the relationship among zonal/meridional wind variations over the central western Sumatera, locality index and the occurrence of characteristic rainfall distribution types in 1992 (see Table 1). As for the wind variations, in westerly to northwesterly (easterly to southeasterly) is the most frequent (the rarest) wind direction over the central western Sumatera. Positive rainfall anomaly of coastal region is relatively larger than that of inland in case of southwesterly to northwesterly wind (not shown). Besides, rainfall amount of inland region is greater than the average in northerly to easterly wind, whereas (although variation is large) rainfall amount of coastal region is almost the same with or less than the average. Thus, inland rainfall anomaly is relatively positive in comparison with the coastal rainfall anomaly. In northwesterly to northerly wind, both of coastal and inland rainfall amount are greater than average. Thus, locality index tends to be negative in easterly to northerly wind range, whereas it tends to be positive in westerly wind range. The correlation coefficient between locality index and zonal wind was 0.56 (99% confidence level). This implies that rainfall is distributed on the windward side of steep mountains over West Sumatera Province, and the contrast between coastal and inland conditions tends to be large with increasing zonal wind speed. It is also suggested that the rainfall locations and zonal wind variations over central western Sumatera, especially for the occurrence of coastal-type rainfall distribution pattern and westerly wind, are related. The coastal-type (totally 6 pentads) occurs at northwesterly wind maximum such as the pentads 48 and 60, and appears at pentad with relatively strong westerly wind range. Wind speed is weak at pentad 8, but it is westerly anomaly. On the other hand, the inland-type (totally 6 pentads) oc-
Vol. 86, No. 5
Fig. 8. Relationship between 850 hPa horizontal wind over the central western Sumatera (100°E, the Equator) and locality index. Triangle (△), circle (○), square (□) and inverse triangle ( ) indicate the typical rainfall distribution pattern, i.e. coastal, inland, active and inactive types, respectively. The pentad number is also indicated. Vertical (horizontal) solid line shows annual mean zonal (meridional) wind speed. Vertical and horizontal dashed line show one standard deviation from annual mean zonal and meridional wind speed, respectively. The shading of the characters shows locality index.
△
604
curs at northeasterly wind maximum clearly such as pentads 4 and 6. However, the inland-type also occurs at strong northerly wind such as pentads 14 and 70, and with weak wind speed at pentads 63 and 66. The inland-type would be caused by not only zonal wind variations, but also others of factors as discussed in Section 6. We also used the data of 1993 in order to examine the wind-rainfall locality relationship more statistically. We examined the locality index in 1993 and defined characteristic rainfall distribution patterns as same as those in 1992. The correlation coefficient between locality index and zonal wind variations in 1993 was 0.34 (99% confidence level),
October 2008
605
HAMADA J.-I. et al.
thus, the wind-rainfall locality relationship shows the similar tendency with that of 1992. Occurrence time of the rainfall patterns is shown in Table 2. The occurrence frequency and the circulation and convection patterns at coastal- and active-type rainfall patterns in 1993 are almost similar to the features in 1992 (not shown). However, inland-type occurred less frequently and they are not so concentrated in Southern Hemisphere summer. They may be caused by that zonal wind tends to be westerly in Southern Hemisphere summer over central western Sumatera in 1993 and/or the spatial distribution of inland stations was localized. Though the observational number and location are not the same in 1992 and 1993, the correlation coefficient between locality index and zonal wind variations was 0.77 (99% confidence level) in case that characteristic rainfall distribution pattern appeared in 1992/1993 (see Fig. 9). Thus, these observational results inferred that the inhomogeneity of rainfall distribution patterns, especially in the coastal− inland contrast, may be related with the phase of zonal wind variations.
Table 2.
Same with Table 1, but for the data in 1993.
Distribution pattern
Pentad number
Coastal-type Inland-type Active-type
5, 15, 23, 54, 66 and 69 13, 26 and 61 4, 25, 29 and 62
Inactive-type
6, 16, 19, 32, 33, 43, 44 and 45
6. Discussions on a relationship between rainfall localities and diurnal variations
Fig. 9. Relationship between 850 hPa zonal wind over the central western Sumatera (100°E, the Equator) and locality index for 40 pentads when characteristic rainfall distribution patterns appeared during 1992−1993. Triangle (△), circle (○), square (□) and inverse triangle ( ) indicate the typical rainfall distribution pattern, i.e. coastal, inland, active and inactive types, respectively. The pentad number for coastal-, inland- and activetype is also indicated. Solid (dashed) mark indicates the pentad of 1992 (1993).
△
As described above, lower tropospheric zonal wind variations tend to intensify the difference of rainfall distributions between coastal and inland regions. It may be suggested that localized water vapor convergence in the windward side of the mountains induces coastal- or inland-type rainfall distributions. However, there are some inconsistencies for the relationship between the occurrence of rainfall distribution pattern, especially for the inland-type, and dominant wind direction. Some of the inland-type occur with weak westerly wind conditions during November−December. Besides, the ITCZ passes over the equatorial Sumatera from mid-October to early-November and it is continuously located in the Southern Hemisphere near the Equator during November−December. Rainfall distribution of the inland-type is relatively uniform as indicated by the coefficient of variation (see Figs. 4 and 6). This relatively uniform rainfall distribution in both of coastal and inland regions may be induced by large-scale convergence of the ITCZ. A few previous studies showed that rainfall was active at Kototabang in inland region of central western Sumatera under easterly or weak westerly, and it was related with intraseasnal variations of the passage of cloud system (e.g., Murata et al.
2002; Seto et al. 2004). Thus, characteristic rainfall distribution patterns during the ITCZ that are located near central Sumatera may be caused by lower tropospheric zonal wind change as intraseasonal variation of cloud system movement. On the other hand, some of the inactive-type rainfall pattern occur in strong westerly phase during Southern Hemisphere winter. Murata
606
Journal of the Meteorological Society of Japan
et al. (2006) showed that severe rain occurred as a single organized synoptic-scale cloud system passed over Sumatera, and dry air intrusion with westerly wind suppressed convections in the rear part. Thus, as a further study subject, we need to consider not only for dominant wind direction, but also atmospheric water vapor conditions in zonal wind phase for rainfall locality in this region.
Vol. 86, No. 5
Besides, different characteristics in rainfall varia tions between coastal and island regions may also result from different characteristics in diurnal variations of convection, because the diurnal variations of convection are strongly dependent on local topo graphic conditions. Figure 10 shows the diurnal and seasonal variations of convection over central western Sumatera (100.5°E, 0.5°S) and the equato-
Fig. 10. Time series of pentad mean Ic (pentad−1), and averaged Ic every 3 hours in each pentad (hour−1 pentad−1) over (a) central western Sumatera (1°×1° region centered at 100.5°E, 0.5°S) and (b) equatorial eastern Indian Ocean (centered at 95.5°E, 0.5°S). Triangle (△), circle (○), square (□) and plus (＋) show the pentad when the coastal-, inland-, active- and inactive-type spatial rainfall distribution patterns appeared, respectively. Diurnal variations averaged for typical rainfall distribution patterns are also indicated for the active (solid), coastal (dot-dashed), inland (dashed) types, and overall average (thick-solid), respectively.
October 2008
HAMADA J.-I. et al.
rial eastern Indian Ocean (95.5°E, 0.5°S) where convection was well correlated with coastal rainfall variations (not shown). As examined by recent studies based on TRMM rainfall and GMS cloud data (e.g., Mori et al. 2004; Sakurai et al. 2005; Nitta and Sekine 1994), diurnal variation is notable over land and surrounding Ocean. The amplitude and the maximum phase of diurnal variations are different according to intraseasonal variations of convection, but a diurnal cycle with an afternoon/ night maximum is clear throughout the year. On the other hand, over the Indian Ocean, the variance of intraseasonal variations is larger than the variance of diurnal variations, although the annual mean Ic is almost the same as the annual mean Ic over Sumatera. Hence, intraseasonal variations are clearer over the Ocean than over Sumatera. Figure 10 also shows the difference in diurnal variations of convection among characteristic rainfall distribution types. Although the number of events is limited, diurnal variation over central western Sumatera is the clearest (amplitude is large) during the period of active-type rainfall, and the maximum phase (1200UTC; LST = UTC + 0700) appears earlier than those of coastal and inland types (1500UTC). For the coastal and inland types, the amplitudes of diurnal convection are almost the same as the annual mean, while midnight (1500UTC−1800UTC) convection is significantly active. On the other hand, diurnal convection variations over the Indian Ocean do not differ among the characteristic rainfall types except for the mean value of convection. Convection is significantly active in coastal and active types. Sakurai et al. (2005) described a systematic migration of clouds with the diurnal cycle in and around Sumatera. In one-year analysis (from May 2001 to April 2002) of them, the westward migration from western steep mountains appeared almost every month, while the eastward migration tended to occur mainly in/near the ITCZ. Eastward migration of convection with the diurnal cycle was found during lower tropospheric westerly wind in/ near the ITCZ. Besides, Shibagaki et al. (2006) described westward moving cloud system with diurnal cycle from Kalimantan to Sumatera in the active (easterly) phase of intraseasonal variation during November 2002. The coastal-type appears throughout the year, and coastal rainfall variations are wellrelated with convection activity off the west coast of Sumatera as investigated by correlation analysis (not shown). The late diurnal convection peak over
607
central western Sumatera in coastal- and inlandtype rainfall distribution and the seasonality of their occurrence may be related to the diurnal migration of cloud system. A future study should investigate the characteristics of cloud migration systems and related rainfall distribution in different phases of intraseasonal variations based on high temporaland spatial-resolution data.
7. Conclusions In this study, dense rain gauge observation data in 1992 (46 stations) were used to investigate spatial rainfall variability in coastal and inland regions of West Sumatera Province, Indonesia. The most remarkable characteristic of the rainfall variation is the contrast between coastal and inland mountainous regions. Annual rainfall is generally abundant on the western side of the mountains (coastal region) and sparser on the eastern side (inland region). The locality index was defined to measure the locality of pentad rainfall distributions based on the time series of pentad rainfall amounts in coastal and inland regions. Rainfall distribution patterns were classified into coastal, inland, active and inactive types. As for the occurrence of each distribution pattern in 1992, coastal- and activetype rainfall patterns tend to appear throughout the year, whereas inland-type rainfall pattern tends to appear during a Southern Hemisphere summer. Inactive-type tends to appear during a Southern Hemisphere winter. The relationship between rainfall and lower tropospheric wind over central western Sumatera was also investigated. Correlation and composite analyses suggest that lower tropospheric zonal wind variations tend to intensify the contrast between coastal and inland rainfall distribution. The windward side of rainfall tends to be greater than the leeward side. The correlation coefficient between locality index and zonal wind over the central western Sumatera was 0.77 (99% confidence level) for the characteristic patterns during 1992−1993. The occurrence of coastal-type rainfall distribution pattern and westerly wind is well-related. In addition, a clear diurnal cycle of convection is found in the active-type rainfall distribution pattern. However, analysis data has been limited in this study, so it is necessary to analyze longer time period data and/ or high-time resolution data for investigating the relationship statistically as the further study subject. Besides, the importance of atmospheric stability and distribution of water vapour for the rainfall
608
Journal of the Meteorological Society of Japan
Vol. 86, No. 5
Table A1. Rainfall observation stations in West Sumatera Province used in this study. Station number, station name, location (degree ∙ minute), altitude and annual rainfall amount in 1992 are given. Type I (II) means that the station is located in the coastal (inland) region. Further details in Section 4. Type
No.
Station name
Latitude
Longitude
Altitude (m)
Rainfall (mm)
— II II II II II II II II II II II I I I I I I II II II II II II I II II I I I II I II II II II II II II II I I I I II II
26a 27a 28 28b 30a 34a 36 36a 36b 37 38 42a 42g 43 43a 43b 43f 44a 46 47 47i 49a 49c 50 50a 50b 50d 51 51a 52 52b 52e 53 53d 53f 55b 56e 57a 57b 57c 58c 59 59a 59b 62 63b
Muara Siberut Sukarami Huberta Peconian Lubuk Gadang Tanpadi Kunangan Sijunjung Sitiung Silago Alhan Panjang Ulubangko Pasar Baru Indarung Tabing Duku Bandar Buat Dam Air Baling Lubuk Sulasih Solok Muara Panas Kumanis Tabat Patah Batusangkar Lubuk Alung Galo Gandang Rambatan Kayu Taman Sicincin Pariaman Maninjau Pilubang Padang Panjang Merapi Atas Gumarang Situjuhbatur Guguk Muara Paiti Tanjung Pati Gunung Malintan Bonjol Lubuk Sikaping Sasak Sukamenanti Rao Parit
−1 ∙ 35 −0 ∙ 56 −1 ∙ 38 −1 ∙ 36 −1 ∙ 32 −0 ∙ 58 −0 ∙ 45 −0 ∙ 42 −1 ∙ 03 −1 ∙ 01 −1 ∙ 05 −1 ∙ 35 −1 ∙ 17 −0 ∙ 57 −0 ∙ 53 −0 ∙ 48 −0 ∙ 57 −0 ∙ 56 −1 ∙ 00 −0 ∙ 48 −0 ∙ 47 −0 ∙ 28 −0 ∙ 19 −0 ∙ 28 −0 ∙ 42 −0 ∙ 29 −0 ∙ 32 −0 ∙ 32 −0 ∙ 32 −0 ∙ 38 −0 ∙ 17 −0 ∙ 27 −0 ∙ 27 −0 ∙ 23 −0 ∙ 13 −0 ∙ 18 −0 ∙ 10 0 ∙ 08 −0 ∙ 09 0 ∙ 06 0 ∙ 01 0 ∙ 07 0 ∙ 00 0 ∙ 07 0 ∙ 34 0 ∙ 30
99 ∙ 12 100 ∙ 37 101 ∙ 10 101 ∙ 09 101 ∙ 15 101 ∙ 33 101 ∙ 06 100 ∙ 59 101 ∙ 38 101 ∙ 15 100 ∙ 47 101 ∙ 09 100 ∙ 29 100 ∙ 29 100 ∙ 21 100 ∙ 19 100 ∙ 25 100 ∙ 29 100 ∙ 36 100 ∙ 40 100 ∙ 42 100 ∙ 54 100 ∙ 32 100 ∙ 35 100 ∙ 16 100 ∙ 33 100 ∙ 34 100 ∙ 19 100 ∙ 17 100 ∙ 07 100 ∙ 13 100 ∙ 00 100 ∙ 24 100 ∙ 26 100 ∙ 28 100 ∙ 36 100 ∙ 35 100 ∙ 30 100 ∙ 39 100 ∙ 40 100 ∙ 13 100 ∙ 10 99 ∙ 42 99 ∙ 53 100 ∙ 01 100 ∙ 10
2 991 — 926 — — — 160 88 238 1458 1000 — 250 14 7 76 — 1000 266 — — 1330 460 2 570 — 144 — 2 479 — 780 1550 718 608 — 140 80 164 244 450 2 108 298 240
2386 2510 4818 4339 3559 3134 3813 1826 2165 3340 1630 1372 2861 5996 4355 4961 4317 6321 2673 1771 1894 2020 2188 1384 2621 1763 2009 4001 4447 3346 2812 3266 2100 2800 2702 1961 2020 2901 2268 2987 3622 2975 3879 3749 3104 5058
October 2008
609
HAMADA J.-I. et al.
variability should be investigated with reliable data such as intensive rawinsonde soundings/weather radar and continuous GPS observation. Some of forgoing studies showed the importance of an interaction between monsoonal wind and complex topography on seasonal rainfall variations over the maritime continent using monthly data (e.g., Chang et al. 2005). In this study we investigated that the characteristics of spatial and temporal rainfall variations over West Sumatera Province and their relation to intraseasonal variations of lower tropospheric wind using pentads data. The frequency of MJOs (large-scale convections over the Indian Ocean) and the accompanying lower tropospheric wind variations may be an important factor to determine seasonal and interannual rainfall variations in this region. This is inferred from the heaviest rainfall at Tabing, coastal region, occurred when an eastward moving MJO convection and accompanying westerly wind have just passed over Sumatera from Indian Ocean. Recently Manton et al. (2001) examined long-term climate variability over Southeast Asian countries including Indonesia based on daily rainfall and maximum/minimum temperature data during 1961−1998. Meteorological observations in Indonesia began to be recorded in the middle of the 19th century, so these data have the potential to cast light on climate change over the maritime continent. It is necessary to classify the stations into distinct orographic locations such as coastal and inland regions (windward/leeward sides of dominant wind direction) for distinguishing the factors that affect their local climate variability. Thus, further studies are necessary to elucidate the general and regional structures of rainfall variation not only in Sumatera, but also in and around the Indonesian archipelago, and their association with wind. To investigate the spatial variation in precipitation over complex topography by comparing local measurement with rainfall estimation by satellite observation is also an important subject for further study. For this purpose, surface rainfall data should be compared with satellite observations to understand the climatology over Indonesia.
Appendix As mentioned in Section 2, we used daily rainfall data at 46 stations over West Sumatera Province in 1992 compiled by the Indonesian Meteorological and Geophysical Agency (BMG). Table A1 lists the geographical location of rain gauge station and
annual rainfall amount in 1992. Some stations in the data book lacked location information (latitude, longitude and altitude); the latitude and longitude of those stations in Table A1 were obtained from BMG operational and commercial maps.
Acknowledgments We thank all the operators of BMG who engage in constructing Indonesian meteorological data base. We also thank the Editor (Dr. Hiroaki Ueda) and two anonymous reviewers for their comments to improve this paper. The present study was financially supported by IORGC Project of Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Hydrometeorological ARay for ISVMonsoon AUtomonitoring (HARIMAU) project of Japan Earth Observation System Promotion Program (JEPP) and also Grants-in-Aid (13440141 and 14595007) of the Japanese Ministry of Education, Science and Culture. Generic Mapping Tools (GMT) were used to create some figures in the present paper.
References Allan, R.J., 1991: Australasia. Teleconnections Linking Worldwide Climate Anomalies: Scientific Basis and Societal Impact. Cambridge University Press, 73−120. BMG, 1993: Pemeriksaan curah hujan propinsi Sumatera Barat: Tahun 1992. Sicincin Climatological Station of Indonesian Agency for the Meteorology and Geophysics, 129 pp. (in Indonesian). Chang, C.-P., Z. Wang, J. McBride, and C.-H. Liu, 2005: Annual cycle of Southeast Asia−maritime continent rainfall and the asymmetric monsoon transition. J. Climate, 18, 287−301. Fukao, S., 2006: Coupling processes in the equatorial atmosphere (CPEA): A project overview. J. Meteor. Soc. Japan, 84A, 1−18. Fukao, S., H. Hashiguchi, M. Yamamoto, T. Tsuda, T. Nakamura, M.K. Yamamoto, T. Sato, M. Hagio, and Y. Yabugaki, 2003: Equatorial Atmosphere Radar (EAR): System description and first results. Radio Sci., 38, 1053, doi: 10.1029/2002RS 002767. Hamada, J.-I., M.D. Yamanaka, J. Matsumoto, S. Fukao, P.A. Winarso, and T. Sribimawati, 2002: Spatial and temporal variations of the rainy season over Indonesia and their link to ENSO. J. Meteor. Soc. Japan, 80, 285−310. Haylock, M. and J.L. McBride, 2001: Spatial coherence and predictability of Indonesian wet season rainfall. J. Climate, 14, 3882−3887. Inoue, T. and K. Nakamura, 1990: Physical and biologi-
610
Journal of the Meteorological Society of Japan
cal background for insect studies in Sumatra. in Natural History of Social Wasps and Bees in Equatorial Sumatra. Hokkaido University Press, 1−11. Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, S. Saha, G. White, J. Woollen, Y. Zhu, M. Chelliah, W. Ebisuzaki, W. Higgins, J. Janowiak, K.C. Mo, C. Ropelewski, J. Wang, A. Leetmaa, R. Reynolds, R. Jenne, and D. Joseph, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc., 77, 437−471. Kuraji, K., K. Punyatrong, and M. Suzuki, 2001: Altitudinal increase in rainfall in Mae Chaem watershed, Thailand. J. Meteor. Soc. Japan, 79, 353− 363. LMG, 1973: Rainfall atlas of Indonesia. Vol. II: Sumatra, Kalimantan, Sulawesi, Nusatenggara, Maluku, Irian Jaya (1911−1940). in Meteorological Note No. 9. Department of Communications Meteorological and Geophysical Institute Republic of Indonesia, 47 pp. Madden, R.A. and P.R. Julian, 1971: Detection of a 40− 50 day oscillation in the zonal wind in the tropical Pacific. J. Atmos. Sci., 28, 702−708. Madden, R.A. and P.R. Julian, 1972: Description of global-scale circulation cells in the tropics with a 40−50 day period. J. Atmos. Sci., 29, 1109−1123. Madden, R.A. and P.R. Julian, 1994: Observations of the 40−50-day tropical oscillation — a review. Mon. Wea. Rev., 122, 814−837. Manton, M.J., P.M. Della-Marta, M.R. Haylock, K.J. Hennessy, N. Nicholls, L.E. Chambers, D.A. Collins, G. Daw, A. Finet, D. Gunawan, K. Inape, H. Isobe, T.S. Kestin, P. Lefale, C.H. Leyu, T. Lwin, L. Maitrepierre, N. Ouprasitwong, C.M. Page, J. Pahalad, N. Plummer, M.J. Salinger, R. Suppiah, V.L. Tran, B. Trewin, I. Tibig, and D. Yee, 2001: Trends in extreme daily rainfall and temperature in southeast Asia and the south Pacific: 1961− 1998. Int. J. Climatol., 21, 269−284. McBride, J., 1999: Indonesia, Papua New Guinea, and tropical Australia: The southern hemisphere monsoon. Meteor. Monogr., 49, 89−99. McGregor, R. and S. Nieuwolt, 1997: Tropical precipitation. in Tropical Climatology. John Wiley and Sons, 187−216, (Second Edition). Meehl, G.A., 1987: The annual cycle and interannual variability in the tropical Pacific and Indian ocean regions. Mon. Wea. Rev., 115, 27−50. Mori, S., J.-I. Hamada, Y.I. Tauhid, M.D. Yamanaka, N. Okamoto, F. Murata, N. Sakurai, H. Hashiguchi, and T. Sribimawati, 2004: Diurnal land-sea rainfall peak migration over Sumatera Island, Indonesian maritime continent observed by TRMM satellite and intensive rawinsonde soundings. Mon. Wea. Rev., 132, 2021−2039. Murakami, T. and J. Matsumoto, 1994: Summer mon-
Vol. 86, No. 5
soon over the Asian continent and western north Pacific. J. Meteor. Soc. Japan, 72, 719−745. Murakami, T., L.-X. Chen, A. Xie, and M.L. Shrestha, 1986: Eastward propagation of 30−60 day perturbations as revealed from Outgoing Longwave Radiation data. J. Atmos. Sci., 43, 961−971. Murata, F., M.D. Yamanaka, M. Fujiwara, S.-Y. Ogino, H. Hashiguchi, S. Fukao, M. Kudsy, T. Sribimawati, S.W.B. Harijono, and E. Kelana, 2002: Relationship between wind and precipitation observed with a UHF radar, GPS rawinsondes and surface meteorological instruments at Kototabang, West Sumatera during September− October 1998. J. Meteor. Soc. Japan, 80, 347−360. Murata, F., M.D. Yamanaka, H. Hashiguchi, S. Mori, M. Kudsy, T. Sribimawati, B. Suhardi, and Emrizal, 2006: Dry intrusions following eastward-propagating synoptic-scale cloud systems over Sumatera Island. J. Meteor. Soc. Japan, 84, 277−294. Nakamura, K., S. Kahono, N. Hasan, I. Abbas, S. Nakano, S. Koji, A. Hasyim, D. Astuti, and Syafril, 2001: Long-term dynamics of phytophagous lady beetle populations under different climate conditions in Sumatra and Jawa, Indonesia. Tropics, 10, 427−448. Nakazawa, T., 1988: Tropical super clusters within intraseasonal variations over the western Pacific. J. Meteor. Soc. Japan, 66, 823−839. Nicholls, N., 1981: Air-sea interaction and the possibility of long-range weather prediction in the Indonesian archipelago. Mon. Wea. Rev., 109, 2435− 2443. Nicholls, N., 1984: The southern oscillation and Indonesian sea surface temperature. Mon. Wea. Rev., 112, 424−432. Nitta, T., T. Mizuno, and K. Takahashi, 1992: Multiscale convective systems during the initial phase of the 1986/87 El Niño. J. Meteor. Soc. Japan, 70, 447−465. Nitta, T. and S. Sekine, 1994: Diurnal variation of convective activity over the tropical western Pacific. J. Meteor. Soc. Japan, 72, 627−641. Okamoto, N., M.D. Yamanaka, S.-Y. Ogino, H. Hashiguchi, N. Nishi, T. Sribimawati, and A. Numaguti, 2003: Seasonal variations of tropospheric wind over Indonesia: Comparison between collected operational rawinsonde data and NCEP reanalysis for 1992−99. J. Meteor. Soc. Japan, 81, 829− 850. Pawitan, H., 1990: Climate data compilation for Sumatra. in Technical Report No. 23. Center for Soil and Agroclimate Research, Bogor, Indonesia. Ramage, C.S., 1968: Role of a tropical “maritime continent” in the atmospheric circulation. Mon. Wea. Rev., 96, 365−370. Renggono, F., H. Hashiguchi, S. Fukao, M.D. Yama naka, S.-Y. Ogino, N. Okamoto, F. Murata, B.P.
October 2008
HAMADA J.-I. et al.
Sitorus, M. Kudsy, M. Kartasasmita, and G. Ibrahim, 2001: Precipitating clouds observed by 1.3-GHz boundary layer radars in equatorial Indonesia. Ann. Geophys., 19, 889−897. Ropelewski, C.F. and M.S. Halpert, 1987: Global and regional scale precipitation patterns associated with the El Niño/southern oscillation. Mon. Wea. Rev., 115, 1606−1626. Rui, H. and B. Wang, 1990: Development characteristics and dynamic structure of tropical intraseasonal convection anomalies. J. Atmos. Sci., 47, 357−379. Saji, N.H., B.N. Goswami, P.N. Vinayachandran and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360−363. Sakurai, N., F. Murata, M.D. Yamanaka, S. Mori, J.-I. Hamada, H. Hashiguchi, Y.I. Tauhid, T. Sribimawati, and B. Suhardi, 2005: Diurnal cycle of cloud system migration over Sumatera Island. J. Meteor. Soc. Japan, 83, 835−850. Sasaki, T., P. Wu, S. Mori, J.-I. Hamada, Y.I. Tauhid, M.D. Yamanaka, T. Sribimawati, T. Yoshikane, and F. Kimura, 2004: Vertical moisture transport above the mixed layer around the mountains in western Sumatra. Geophys. Res. Lett., 31, doi: 10.1029/2004GL019730. Schmidt, F.H. and J.H.A. Ferguson, 1951: Rainfall types based on wet and dry period ratios for Indonesia with western New Guinee. Verhandelingen, 42, 77 pp. Seto, T.H., M.K. Yamamoto, H. Hashiguchi, and S.
View publication stats
611
Fukao, 2004: Convective activities associated with intraseasonal variation over Sumatera, Indonesia, observed with the equatorial atmosphere radar. Ann. Geophys., 22, 3899−3916. Shibagaki, Y., T. Shimomai, T. Kozu, S. Mori, Y. Fujiyoshi, H. Hashiguchi, M.K. Yamamoto, and S. Fukao, 2006: Multiscale aspcects of convective systems associated with an intraseasonal oscillation over the Indonesian maritime continent. Mon. Wea. Rev., 134, 1682−1696. Wheeler, M.C. and H.H. Hendon, 2004: An all-season real-time multivariate MJO index: Development of an index for monitoring and prediction. Mon. Wea. Rev., 132, 1917−1932. Widiyatmi, I., H. Hashiguchi, S. Fukao, M.D. Yama naka, S.-Y. Ogino, K.S. Gage, S.W.B. Harijono, S. Diharto, and H. Djojodihardjo, 2001: Examination of 3−6 day disturbances over equatorial Indonesia based on boundary layer radar observations during 1996−1999 at Bukittinggi, Serpong and Biak. J. Meteor. Soc. Japan, 79, 317−331. Wu, P., J.-I. Hamada, S. Mori, Y.I. Tauhid, M.D. Yama naka, and F. Kimura, 2003: Diurnal variations of precipitable water over a mountainous area of Sumatra Island. J. Appl. Meteor., 42, 1107−1115. Yasunari, T., 1981: Temporal and spatial variations of monthly rainfall in Java, Indonesia. Southeast Asian Studies, 19, 170−186. Zhu, B. and B. Wang, 1993: The 30−60-day convection seesaw between the tropical Indian and western Pacific Oceans. J. Atmos. Sci., 50, 184−199.
