Reconstruction of Rainfall Records at 24 Observation Stations in Sumatera, Colonial Indonesia, from 1879 to 1900

Ryosuke Kajita aResearch Institute for Humanity and Nature, Kyoto City, Kyoto, Japan

Search for other papers by Ryosuke Kajita in
Current site
Google Scholar
PubMed
Close
,
Manabu D. Yamanaka aResearch Institute for Humanity and Nature, Kyoto City, Kyoto, Japan
bKobe University, Kobe City, Hyogo, Japan

Search for other papers by Manabu D. Yamanaka in
Current site
Google Scholar
PubMed
Close
, and
Osamu Kozan aResearch Institute for Humanity and Nature, Kyoto City, Kyoto, Japan
cCenter for Southeast Asian Studies, Kyoto University, Kyoto City, Kyoto, Japan

Search for other papers by Osamu Kozan in
Current site
Google Scholar
PubMed
Close
Open access

Abstract

This study reconstructed historical monthly rainfall data at 24 observation stations in Sumatera, Indonesia, for 22 years (1879–1900) during the Dutch colonial era. The literature research was conducted in libraries in Indonesia and the Netherlands to collect meteorological data recorded in a series of Regenwaarnemingen in Nederlandsch-Indië, which have never been reconstructed or analyzed previously. The results showed that the six western coast stations (97°–103°E, 3°N–4°S) have abundant rainfall exceeding 3000 mm yr−1, and the 10 eastern coast stations and three stations east of the Barisan ridge obtained rainfall below 3050 mm yr−1. Abundant coastal rainfall appeared only along the western coast, but not in the eastern coast and east of the Barisan ridge. The El Niño events of 1888/89 had low impact on rainfall amounts during the dry season (May–October) at the six stations of the Southern Hemispheric climate type. A significant decrease in annual dry-season precipitation was observed in 1881, 1885, and 1896. When comparing monthly data for 1931–60 and 1971–2014 at the northernmost station Banda Aceh (5.5°N, 95.3°E) and the eastern coast station Bengkalis (1.5°N, 102.1°E), a noticeable decrease and increase occurred in June–December and January–May, respectively, throughout the last century. The recent El Niño events have resulted in a decline in rainfall and dry conditions in Sumatera, but the Indian Ocean dipole had much influence on the amount of rainfall in the late-nineteenth century.

© 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Kajita’s current affiliation: Center for Graduate School Innovation, Tohoku University Advanced Graduate School, Sendai, Japan.

Corresponding author: Ryosuke Kajita, kajita0709@gmail.com

Abstract

This study reconstructed historical monthly rainfall data at 24 observation stations in Sumatera, Indonesia, for 22 years (1879–1900) during the Dutch colonial era. The literature research was conducted in libraries in Indonesia and the Netherlands to collect meteorological data recorded in a series of Regenwaarnemingen in Nederlandsch-Indië, which have never been reconstructed or analyzed previously. The results showed that the six western coast stations (97°–103°E, 3°N–4°S) have abundant rainfall exceeding 3000 mm yr−1, and the 10 eastern coast stations and three stations east of the Barisan ridge obtained rainfall below 3050 mm yr−1. Abundant coastal rainfall appeared only along the western coast, but not in the eastern coast and east of the Barisan ridge. The El Niño events of 1888/89 had low impact on rainfall amounts during the dry season (May–October) at the six stations of the Southern Hemispheric climate type. A significant decrease in annual dry-season precipitation was observed in 1881, 1885, and 1896. When comparing monthly data for 1931–60 and 1971–2014 at the northernmost station Banda Aceh (5.5°N, 95.3°E) and the eastern coast station Bengkalis (1.5°N, 102.1°E), a noticeable decrease and increase occurred in June–December and January–May, respectively, throughout the last century. The recent El Niño events have resulted in a decline in rainfall and dry conditions in Sumatera, but the Indian Ocean dipole had much influence on the amount of rainfall in the late-nineteenth century.

© 2022 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Kajita’s current affiliation: Center for Graduate School Innovation, Tohoku University Advanced Graduate School, Sendai, Japan.

Corresponding author: Ryosuke Kajita, kajita0709@gmail.com

1. Introduction

a. Background

Rainfall in Sumatera and Kalimantan affects society at regional and global scales. Rainfall, cloud activity, and their interannual variations in this region are directly related to global energy balance, circulation, and climate, which have been studied extensively elsewhere (see, e.g., Yamanaka et al. 2018, as a review). Decreased rainfall encourages the spread of fires over large areas of degraded peatlands, which causes transboundary haze beyond the Malacca Strait and the South China Sea, as well as associated health problems for people across Southeast Asian countries (Field et al. 2016). A huge peatland fire in September 2015 resulted in substantial amounts of CO2 emissions and severe health hazards. Such interannual variations are remarkable in this region and cause an extended annual dry season. Since the 1980s, broad areas of the wet peatland in Indonesia have been converted into agricultural lands by a government-led transmigration program (Siegert et al. 2001), and the water drainage and degradation of the peatland have modified the local climate and hydrological cycle. However, land modification for cultivation, meteorological observations, and data compilation dates back to the Dutch colonial era. As global climate change has a time scale of more than one century, it is important to understand historical trends to analyze long-term environmental problems.

After Indonesia’s independence, meteorological operations were maintained by the Indonesian Agency of Meteorology, Climatology, and Geophysics (BMKG). The BMKG database “Data Iklim” (BMKG 2015) includes meteorological data, such as precipitation, temperature, humidity, and wind velocity. The database has observation records since the 1960s. Historical records before the 1960s are not included but are available to download in 1-month blocks for selected years/months. Another important meteorological database is the Southeast Asian Climate Assessment and Dataset (SACA&D; Project Team SACA&D 2011), which is available for download and includes data from Indonesia and other Southeast Asian countries. SACA&D is an international project based on the cooperation between the BMKG, the Dutch Royal Meteorological Institute (KNMI), and 23 national meteorological and hydrological services, observatories, and universities in Southeast Asia, containing meteorological data since the 1960s. Notably, the dataset also contains the historical records of Batavia (now Jakarta) since 1866. Batavia was the central city of the former Dutch colonial administration.

b. Previous studies

Historical meteorological records of local regions are not available from Data Iklim and SACA&D. Previous studies have focused on reconstructing various meteorological parameters using historical records. Können et al. (1998) reconstructed rain-day counts from 1829 to 1850 and sea level pressures from 1841 to 1997 in Batavia. They also reconstructed pressure data from Tahiti to extend the Southern Oscillation index back to the early nineteenth century. Hamada et al. (2002) collected daily precipitation records from 157 observation stations from 1961 to 1990 and analyzed the characteristics of rainy seasons at 46 stations (including 5 in Sumatera and 6 in Kalimantan) with improved continuity. Siswanto et al. (2016) presented the centennial rainfall variations in Batavia and noted that the rain-day counts have decreased in recent years due to local warming caused by climate change and the rapid urbanization of Jakarta. Wati et al. (2019) evaluated the rainy-season onset determination by comparing rainfall amounts in Jawa and Bali. Kajita (2019) collected monthly precipitation records from 1879 to 1900 from six colonial stations in eastern Sumatera and southern Kalimantan to see the rainfall variations in peatland areas.

Eguchi (1983) used monthly precipitation data from October 1978 to September 1979 to define the rainfall distribution in Indonesia. He divided Indonesia into four regions: A–I, A–II, B–I, and B–II; in regions A and B, the wettest months were October–February and March–July, respectively, region I had a single primary rainfall peak, and region II had both a primary and a secondary rainfall peak. Eguchi (1983) determined that Indonesia has four climate types based on rainfall characteristics. In A–I, January is the wettest month and during April–October precipitation is less than 100 mm month−1. The A–II region features abundant rainfall throughout the year, with more than 100 mm month−1 of precipitation. The west of the Barisan ridge (Barisan Mountains, located near the western coast; the red dotted line in Fig. 1) has more precipitation than the other regions. The B–I region east of the Barisan ridge and the coastal regions have less precipitation during October–February. The B–II region features the most rainfall concentrated in March–May, but November–January rarely exceeds 300 mm month−1. Eguchi (1983) clarified that most of Sumatera is categorized as A–II, but only the southern part of Banka Island and Lampung Province are categorized as B–II and A–I, respectively.

Fig. 1.
Fig. 1.

Map showing the geographical coordinates of analysis observation stations in Sumatera: (a) 24 observation stations during 1879–1900, (b) 28 observation stations during 1931–60, listed in Pawitan (1990), and (c) 30 observation stations operated by BMKG during 1971–2014. See section 2 for the selection methods of analysis stations in each period. Open circles, filled triangles, and crosses indicate stations in the western coast, eastern coast, and east of the Barisan ridge, respectively. The red dotted line shows the Barisan ridge from northwest to southeast Sumatera.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

Aldrian and Susanto (2003) found a distinct feature of rainfall variability in Indonesia from monthly rainfall data during 1961–93 using the Global Historical Climatology Network database (GHCN; Vose et al. 1992). They divided Indonesia into regions A′, B′, and C′. The prime notation was assigned to the definition provided by Aldrian and Susanto (2003) to avoid confusion with the notations provided by Eguchi (1983). Region A′ features a primary rainfall peak in December with approximately 320 mm month−1. The driest months are July–September, with rainfall of 100 mm month−1. Region B′ has a secondary rainfall peak during October–November and March–May, with a precipitation of 310 mm month−1. Regions A′ and B′ have similar annual rainfall cycles. Region C′ has a primary rainfall peak during June–July with a precipitation of 300 mm month−1, and the driest month is November–February. According to these three climate classifications, the rainfall distribution for the northern and southern sides of the 2°S latitudes, Sumatera are classified as Regions B′ and A′, respectively.

Yamanaka et al. (2018) summarized several studies on the Indonesian “Maritime Continent” and its precipitation (Hamada et al. 2002, 2008; Aldrian and Susanto 2003; Kubota et al. 2011; Siswanto et al. 2016; Yanto et al. 2016; As-syakur et al. 2016; Lestari et al. 2016), to demonstrate the decrease in rainfall in the tropical coastlines in the late twentieth century, followed by an increase that took place after the 1997/98 El Niño events. This study also indicated that the diurnal cycle is the mechanism by which annual precipitation is deeply related to the coastline distances. The sea surface temperatures also increased due to global warming.

A series of meteorological studies were published in Verhandelingen, Koninklijk Magnetisch en Meteorologisch Observatorium te Batavia (Research, Royal Magnetic and Meteorological Observatory in Batavia) in the Dutch colonial era. Braak (1921) issued three volumes of Het Klimaat van Nederlandsch-Indië (The Climate in the Netherlands Indies) on the meteorological characteristics of local regions that described the uniqueness of the tropical climate region in Indonesia in the nineteenth and early-twentieth centuries. Berlage (1949) listed the precipitation records of 4339 observation stations from 1879 to 1941 in Verhandelingen and made a list of daily precipitation, monthly/annual average precipitation, and rain-day counts. The precipitation data were originally recorded in Regenwaarnemingen in Nederlandsch-Indië (see section 2 for details). In Sumatera, there were a total of 621 observation stations by 1941. Operational observations were difficult during the transition from the Dutch colonial era to the Japanese occupation.

This study focuses on data reconstruction and analysis of historical rainfall records in Sumatera. Section 2 introduces Dutch colonial materials published in the late-nineteenth century and compares the observations with those of current stations. Section 3 describes the rainfall variation in the longitudinal and latitudinal locations, along with the coastal regions in Sumatera. Section 4 shows the analysis results of monthly precipitation and considers a seasonal cycle in Sumatera, while also discussing climate classification. Section 5 describes the interannual variation in Sumatera and compares the analysis with historical El Niño events. Section 6 summarizes and concludes the study.

2. Data sources

Previous studies (e.g., Können et al. 1998; Hamada et al. 2002; Project Team SACA&D 2011; Siswanto et al. 2016) have used the meteorological records listed in the Verhandelingen, Koninklijk Magnetisch en Meteorologisch Observatorium te Batavia. This series of materials covered the records of central observation stations in Batavia, the most important city for colonial government administration. In the late-nineteenth century, the city of Batavia was narrower than the present capital city of Jakarta. Rapid urbanization after the independence of Indonesia expanded its capital area much more than in colonial days. Unlike in Batavia, meteorological data in Sumatera in the nineteenth century have not yet been reconstructed or effectively used in previous meteorological analyses. The administrative division in Sumatera also differs from that of the modern Indonesian administration. In particular, the eastern coast of Sumatera was characterized by areas of lowland with overgrown tropical plants and peatlands (Loeb 1935). Bengkalis (1.5°N, 102.1°E), the former capital city of Residency of Sumatera’s East Coast, was more populated than the current major city, Pekanbaru, in Riau Province (Paulus 1917; Stibbe et al. 1919).

a. Colonial materials

This study used the Dutch colonial materials Regenwaarnemingen in Nederlandsch-Indië (Rainfall in the East Indian Archipelago), volumes 1–22, which include the precipitation records observed in colonial Indonesia (Bergsma 1880, 1881, 1882; Figee 1885, 1899, 1900, 1901; Poortman 1884; van der Stok 1883, 1886, 1887, 1888, 1889, 1890, 1891, 1892, 1893, 1894, 1895, 1896, 1897, 1898). The first volume was published by Batavia Landsdrukkerij (Batavia Printing Office) in 1880. The 22 volumes included monthly precipitation records from 1879 to 1900.

The colonial government ordered the Batavia Observatory to publish annual reports, which focused only on rainfall observation records. This is intended to separate rainfall observation records from various other records. In 1850, the Dutch colonial government ordered the Koninklijk Natuurkundige Vereeniging in Nederlandsch-Indië (Royal Physics Association in the Netherlands Indies) to start the observation of geophysical phenomena such as earthquakes and volcanic eruptions throughout the Indonesian archipelagoes. The association listed the observation records, local statuses for various geophysical phenomena, and biological characteristics of unique tropical plantations in their annual reports, Natuurkundige Tijdschrift voor Nederlandsch-Indië (Journals of Physics in the Netherlands Indies). Since 1879, the rainfall records were recorded separately in Regenwaarnemingen in Nederlandsch-Indië. For this reason, we selected these materials for data reconstruction.

b. Access to colonial materials

Meteorological observations in colonial Indonesia were conducted continuously during 1879–1941. Colonial materials are usually written in Dutch and are currently archived in several libraries. These meteorological records have not yet been compiled into a database and hence are difficult to use. Some volumes have already been digitized, but low-resolution images sometimes hinder the reading of numerals. Reading the materials directly in the libraries is preferable to avoid misreading numerals or descriptions. We conducted literature research at 1) Universitaire Bibliotheken Leiden (UBL; Leiden University Libraries) and 2) Perpustakaan Nasional Republik Indonesia (Perpusnas, National Library of Republic Indonesia). UBL hosts the Reading Room Special Collections for precious and fragile historical materials, which may not be removed from the room. Perpusnas prohibits the use of cameras inside the library, and visitors must ask library staff for photocopies. This study reconstructed historical observation records in Sumatera in the late-nineteenth century from limited materials, making the data available for meteorological research on colonial Indonesia.

c. Observation methods in the nineteenth century

Bergsma (1880) introduced a detailed observation method in the preface section of the material. Observers were strictly instructed to conduct precise rainfall observations. Official and nonofficial staff worked on observation duties. Können et al. (1998) stated that the official staff included surgeons and medical officers from the army and civil services, respectively; the others were categorized as nonofficial staff. The observers used equipment distributed by the Batavia Observatory to conduct observations once per day at 0600, 0700, 0800, or 0900 local time (UTC + 7 h).

d. Original monthly precipitation data

This study reconstructs rainfall records from 24 observation stations in Sumatera. These 24 stations operated continuously during 1879–1900. The other 41 stations were also installed but were not continuously monitored during this period. Our analysis focused only on the 24 stations in Sumatera. Table 1 lists the detailed information of the 24 stations based on Dutch materials. There was a total of 264 monthly precipitation records over 22 years. We used the island, province, and region names currently used in Indonesia. Additionally, “Sumatera” represents Sumatera Island, the surrounding islands (e.g., Bengkalis), and adjacent archipelago. In the late-nineteenth century, the administrative divisions were different from those of the current provincial areas. Sumatera was divided into residentie (residencies) and afdeeling (divisions). Brill (1897) listed the names of the residencies and divisions used in 1897.

Table 1

Geographic coordinates of 24 observation stations located in Sumatera from 1879 to 1900. Listed stations are ordered from north to south. “Colonial name” refers to the name derived from the colonial materials. “Province” indicates current Indonesian administrative division. Latitude and longitude are given in degrees and minutes. “Missing months” and “Percentage” indicate total missing months and missing percentages, respectively.

Table 1

Figure 1a shows a map of the colonial stations in 1879–1900. Because Sumatera Island is elongated in the northwest–southeast direction, the stations are partitioned into the western and eastern sides of the Barisan ridge. Therefore, we divided Sumatera into three distinct regions: the western coast, eastern coast, and east of the Barisan ridge. The table in Microsoft Excel worksheet A in the online supplemental material indicates the numerical data as monthly precipitation at 24 observation stations in Sumatera for 1879–1900. Worksheet B in the online supplemental material shows the annual precipitation at 24 stations in 1879–1900. Worksheet C in the online supplemental material lists the monthly rain-day counts at 24 stations in 1879–1900. Worksheet D in the online supplemental material contains the captions for the tables in worksheets A–C.

e. Current observation data for comparison

As we reconstructed colonial data in this study, comparisons were made with more recent data using two datasets, Pawitan (1990) and BMKG (2015). Pawitan (1990) compiled rainfall data from 569 Sumatera stations during the middle of the twentieth century. They listed the average annual and monthly precipitation for each station. The list covers the periods primarily from 1931 to 1960, but some stations cover shorter (longer) periods and limited observation years. In addition, the precipitation records in Pawitan (1990) are the average rainfall amounts of all observed years. Therefore, we selected 28 of 569 stations for our analysis so that the observation years amount to at least 30. Table 2 shows the selected 28 stations with basic information cited from the appendix of Pawitan (1990). See Fig. 1b for the location of these 28 stations.

Table 2

Geographic coordinates of 28 observation stations in Sumatera from 1931 to 1960 (Pawitan 1990). A total of 569 stations are listed in Pawitan (1990); if the number of observation years is less than 30, the stations are excluded. “Obs yr” indicates the total number of observation years during the “Period” of operation.

Table 2

According to Data Iklim, the BMKG operates 44 observation stations in Sumatera. Table 3 provides detailed information on the BMKG stations. These 44 stations are located in the 10 provinces in Sumatera, with at least 2 stations in each province. Daily data from Data Iklim are not available for specific years and regions for unknown reasons. In Sumatera, 9 of the 44 stations began observations in the 1970s, 17 in the 1980s, 12 in the 1990s, and 6 in the 2000s.

Table 3

Geographic coordinates of current 44 BMKG observation stations in Sumatera from 1971 to 2014. Each station started their operation since the year shown in the right column. “WMO No.” refers to the World Meteorological Organization’s observing station number listed in Region V (“South-West Pacific”), Indonesia.

Table 3

Because Data Iklim has missing daily records, we selected a valid analysis year for calculation using the analysis method as stated in Hamada et al. (2002). We divided one analysis year into 73 pentads. Daily data on 29 February in a leap year were included in pentad 12. We allowed missing daily data for a maximum of 2 days in each pentad. If any analysis year was missing pentad data, all the data in that analysis year were omitted from the calculation. We chose stations for which we had at least 3 years of data. The missing rate was less than 10% (8 pentads) per year for missing daily data; therefore, the influence of the missing data was not large in the analysis. Consequently, we selected 30 of the 44 stations that obtained at least three analysis years for calculation (Tables 3 and 4). Figure 1c shows the locations of the 30 selected stations.

Table 4

List of BMKG rainfall year data for the 1971–2014 period analyzed in this study. Asterisks and slashes indicate analyzed 73-pentad datasets and incomplete datasets, respectively. The maximum number of days allowed for the missing daily data was two days in each pentad. If any pentad data were lacking in an analysis year at a station, all of the data in that analysis year are omitted from calculations. The omitted year is indicated with a slash. Thirty BMKG stations are listed below, with at least three available analysis years. See Table 2 for details on the location of the stations.

Table 4

3. Annual precipitation

a. Results from 1879 to 1900 data

Figures 2a and 2b show the longitudinal and latitudinal distributions of annual precipitation averaged for 1879–1900 at 24 colonial stations in Sumatera.

Fig. 2.
Fig. 2.

The annual precipitation of (left) the 24 colonial stations (1879–1900) and right) the 28 stations (1931–60) and the 30 BMKG stations (1971–2014) in red and black shapes, respectively. Open circles, filled triangles, and crosses indicate stations on the western coast, eastern coast, and east of the Barisan ridge, respectively. The horizontal axis indicates (a),(d) the longitudes; (b),(e) latitudes; and (c),(f) the distance from the nearest coastline (km). In (c) and (f), crosses show the distance from the eastern coast and not from the western coast.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

The annual average precipitation in the eastern stations had less rainfall than most of the western coast stations. The nine eastern coast stations had precipitation of less than 3050 mm yr−1, but that of the six western coast stations for 97°–103°E and 3°N–4°S was more than 3000 mm yr−1, and that of the other three western coast stations for 99°–101°E and 2°N–1°S was 2000–3000 mm yr−1. The two stations east of the Barisan ridge for 103°–104°E and 3°–4°S had more than 3000 mm yr−1 of precipitation, and that of the other three stations east of the Barisan ridge was 2000–3000 mm yr−1. The six western coast stations located close to 100°–101°E and 0°–1°S had annual rainfalls of 2000–5000 mm yr−1. In general, the western and eastern coast stations have relatively larger (maximum: 4780 mm at Sibolga, 1.7°N, 98.8°E) and smaller (minimum: 1727 mm at Banda Aceh, 5.5°N, 95.3°E) annual rainfall, respectively. The south–north distribution shown in Fig. 2b is roughly reversed to the west–east in Fig. 2a. This difference is clear if we plot the data as a function of the distance from the west and east coastlines, as shown in Fig. 2c. The five western and five eastern stations are located on the coastlines of Sumatera. The five stations east of the Barisan ridge are plotted as the distance from the eastern coastline shown in Fig. 2c. At the western stations, the coastal stations had a larger amount of rainfall than the stations away from the coast. Figure 2c shows the stations east of the Barisan ridge, and the eastern coast stations experienced a similar amount of precipitation. The eastern coast stations are all devoid of coastal rainfall characteristics.

b. Comparison with other periods

In Figs. 2d–f, the annual rainfall at 28 stations (1931–60) and at 30 BMKG stations (1971–2014) are plotted for comparison. The general features are similar to those for 1879–1900 shown in Figs. 2a–c. In particular, the maximum precipitation is 4600–4800 mm for more than a century. However, the total average for all stations decreased by 227 mm from the annual average precipitation of 1879–1900 to 1971–2014. If we focused on three regions separately, the western coast stations have increased for 198 mm, but the stations in the east of the Barisan ridge and the eastern coast stations have decreased by 710 and 169 mm, respectively. A significant decrease appeared in the eastern part of Sumatera over the century.

Hamada et al. (2002) analyzed the rainfall characteristics for 46 stations over Indonesia, including five stations in Sumatera (from north to south, Sigli, Medan, Tarempa, Jambi, and Pangkal Pinang; see Table 3) for 1961–90. These five stations are all located in the eastern part of Sumatera, but we need to compare differences in rainfall variations with the western part of Sumatera. They clarified that the western part of Indonesia had abundant rainfall (>3000 mm yr−1), but smaller than that in the eastern part of Indonesia (2000 mm yr−1). Although the western part of Indonesia was mentioned, the western coast stations of Sumatera were not included in the average annual rainfall. As shown in Figs. 2a, 2b, 2d, and 2e, three periods (1879–1900, 1931–60, and 1971–2014) had abundant rainfall, specifically in most of the western stations of Sumatera, which had a value of more than 2000 mm yr−1. Moreover, the western coast stations had a recorded value of more than 3000 mm yr−1, as shown in Figs. 2c and 2f.

Ogino et al. (2016, 2017) analyzed the Tropical Rainfall Measuring Mission satellite (TRMM 3A25, December 1997–January 2011) and objective reanalysis (a part of JRA55, 1981–2010) data and showed that the rainfall was concentrated within a distance of 300 km from the coastline. Additionally, they noted that abundant precipitation of over 3500 mm yr−1 occurred in almost all coastal regions and rarely in other regions. However, they measured the distance from the nearest coastline without considering the side of the Barisan ridge. Therefore, this study considered the Barisan ridge as an important geographic feature to analyze rainfall in Sumatera. Multisatellite analysis (TRMM 3B42; 1998–2014) by As-syakur et al. (2019) and gridpoint data (GPCP; 1901–2007) by Qian (2020) showed seasonal mean precipitation distributions over Sumatera. Figure 2 suggests that at least the western coast stations have features similar to previous studies through the late-nineteenth and early-twenty-first centuries.

4. Seasonal cycle

a. Results of data analysis

Figure 3 shows the standardized monthly average precipitation. The stations operated in 1931–60 and 1971–2014 were not located in the exact same locations as the colonial stations. Therefore, we compared the standardized monthly average precipitation of the 24 colonial stations (red solid line), with the nearest Pawitan stations (thick black line) and BMKG stations (dotted line). Table 5 shows the local names of the stations for comparative analysis. The geographic distances between neighboring stations are also listed. Because recent observations have shown that rainfall variations in Sumatera, including seasonal cycles, are strongly affected by diurnal sea–land breeze circulations on both sides of the Barisan Mountains (e.g., Yamanaka et al. 2018), we divided Sumatera into three regions, as shown in Figs. 3a–c: the western coast, the east of the Barisan ridge, and the eastern coast. There were 9 colonial stations (Fig. 3a), 5 stations (Fig. 3b), and 10 stations (Fig. 3c) between 1879 and 1900.

Fig. 3.
Fig. 3.

Standardized monthly averaged precipitation in Sumatera. The red solid line, black solid line, and black dotted line indicate seasonal cycle in colonial stations (1879–1900), Pawitan’s stations (1931–60), and BMKG stations (1971–2014), respectively. See Table 5 for the nearest stations to colonial stations in 1931–60 and 1971–2014. The plots indicate (a) 9 western coast stations, (b) 5 stations in the east of the Barisan ridge, and (c) 10 eastern coast stations.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

Table 5

Changes of local names and places of stations in Sumatera: 24 colonial stations from 1879 to 1900, 28 stations from 1931 to 1960 (Pawitan 1990), and 30 stations from 1971 to 2014 (BMKG 2015). See Fig. 1 for the location of the stations and the Barisan ridge. The numbers in parentheses indicate the distances (km) between the colonial stations (boldface type) and the other neighboring stations.

Table 5

Figure 3a shows the standardized monthly average precipitation at the western coast stations. The two northwestern colonial stations, Singkil (2.3°N, 97.8°E) and Sibolga, are compared with the two Pawitan stations, Sibolga and Gunung Sitoli (1.3°N, 97.6°E), and one BMKG station, F.L. Tobing (1.6°N, 98.9°E). The other seven western coast colonial stations in Fig. 3a are also compared with the nearest stations of 1931–60 and 1971–2014 (see Table 5). At the three north colonial stations, monthly precipitation has increased from June–August over the last century. The third rainfall peak in 1971–2014 is visible in July but not visible in 1879–1900. In the last century, monthly precipitation has increased and decreased in June–December and January–February, respectively, at the three equatorial western coast colonial stations. The rainy-season rainfall decreased but was not visible in Padang (1.0°S, 100.4°E) and Lubuk Sulasih (1.0°S, 100.6°E). At the southwestern colonial station, monthly precipitation has increased and decreased in January–March and September–November, respectively, over the last century.

Figure 3b shows the standardized monthly average precipitation in the east of the Barisan ridge. The five colonial stations, mostly located in southwestern Sumatera, do not show marked variations in comparison with the western coast stations. The differences among the three periods east of the Barisan ridge are more stable than the western coast regions.

Figure 3c shows the standardized monthly average precipitation on the eastern coast of Sumatera. The northernmost colonial station was compared with the Pawitan station, Sigli (5.4°N, 96.0°E), and two BMKG stations, Sultan Iskandar Muda (5.5°N, 95.4°E) and Cut Bau Maimun Saleh (5.9°N, 95.3°E). The other nine eastern coast colonial stations in Fig. 3c were also compared with the nearest stations (see Table 5). The rainfall variations on the eastern coast, especially in Bengkalis, decreased in monthly precipitation during the dry season (July–November) and increased during the rainy season (January–May) in the last century. Similar to Bengkalis, the rainfall variations on the northernmost colonial station, Banda Aceh, decreased in June–October and increased in January–April.

Figure 4 shows the monthly precipitation anomaly for 1879–1900. In the following subsections, the seasonal cycle on the western coast, east of the Barisan ridge, and eastern coast of Sumatera during the rainy and dry seasons are analyzed.

Fig. 4.
Fig. 4.

The seasonal cycle in 24 stations indicating monthly precipitation anomalies from 1879 to 1900. The horizontal axes indicate January–December. Station numbers and geographical coordinates are the same as those provided in Table 1. The solid line shows monthly precipitation anomalies to see the seasonal cycle. The plots indicate (a) 9 western coast stations, (b) 5 stations in the east of the Barisan ridge, and (c) 10 eastern coast stations.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

1) Seasonal cycle in western coast stations

Figure 4a shows the nine western coast stations. January is the wettest month in Solok (0.8°S, 100.7°E); April is the wettest month in Bukittinggi (0.4°S, 100.5°E); October is the wettest in Singkil, Sibolga, and Padang; December is the wettest for the remaining four stations. February is the driest month in Singkil and Padang, whereas in the remaining seven stations, June or July is the driest month. In Singkil, Padang Panjang (0.5°S, 100.5°E), and Lubuk Sulasih, a secondary rainfall peak occurred in April, whereas in Sibolga and Bukittinggi, a secondary peak occurred in March and October, respectively. Figure 4a also shows the rainfall fluctuations and rainfall periodicity at the nine western coast stations. Singkil, Sibolga, and Bukittinggi feature semiannual periodicity. In Padang Panjang and Lubuk Sulasih, a secondary rainfall peak occurred four months after the wettest month. The remaining stations showed only a primary rainfall peak.

2) Seasonal cycle in eastern coast and east of the Barisan ridge

Figure 4b shows the five stations east of the Barisan ridge. January is the wettest month in the three southern colonial stations, and December is the wettest month for Pagar Alam (4.1°S, 103.4°E) and Pajakumbuh (0.3°S, 100.8°E). In Pagar Alam, a secondary rainfall peak occurred in April. All five stations had the driest month in July, yet east of the Barisan ridge, the month with the highest rainfall appeared a month later than that of the western coast stations.

Figure 4c shows the 10 eastern coast stations. The two northeastern colonial stations had the highest rainfall in October, Bengkalis had the highest rainfall in November, and the other seven stations had the highest rainfall in December. The four stations at a latitude of 2°–6°N had the lowest rainfall in February and March. The six stations at a latitude of 2°N–6°S had the lowest rainfall in February, June, or July.

Figure 4c also shows a secondary peak of rainfall in the annual cycle. In the northern part of Sumatera, the secondary peak occurred in May, whereas in the adjacent eastern islands, such as Bengkalis, Tanjung Pinang (0.9°N, 104.4°E), and Tanjung Pandan (2.8°S, 107.6°E), a secondary peak occurred in April. Inland stations such as Jambi (1.6°S, 103.6°E) and Palembang (3.0°S, 104.8°E) have a secondary rainfall peak in March, while the eastern coast stations, Idirayeuk (4.9°N, 97.8°E) and Muntok (2.1°S, 105.2°E), only have a primary rainfall peak.

Figure 4c shows the rainfall fluctuations and the onset of the rainy season on the eastern coast. Four northeastern coast stations feature semiannual periodicity. Two stations located at a far distance from Sumatera mainland, a secondary rainfall peak occurred 4 months after the wettest month, whereas that of two southeastern inland stations occurred 3 months after the wettest month. Such characteristics indicate that the eastern coast of Sumatera has a rainy season from October to December, and a secondary rainfall peak from March to May. These characteristics did not appear in Idirayeuk and Muntok. The seasonal cycle of Sumatera by area distribution in the late-nineteenth century is difficult to accurately determine because of the effects of the uneven distribution of land masses, which leads to increased continental climatic ambiguity. Inoue and Nakamura (1990) noted that the tropical climate of Sumatera could not be explained solely by the latitudinal location because of its unique landforms.

3) Remarkableness of rainy and dry seasons

We compared the gap between the maximum and minimum monthly precipitation at each station (Figs. 4a–c). The two eastern coast stations, Idirayeuk and Muntok, have the largest gaps between maximum and minimum monthly precipitation of 406 and 441 mm, respectively; no secondary rainfall peak occurred at these two stations. The gap between the maximum and minimum monthly precipitation at two southeastern stations, Tanjung Pandan and Palembang, is 277 and 249 mm, respectively. These four eastern stations exhibit a drastic increase in precipitation during the rainy season, making the onset of this season unambiguous. The other six eastern coast stations have gradual and less drastic rainfall variations.

From the southwest to the equator, Padang Panjang, Tebing Tinggi (3.6°S, 103.1°E), and Lahat, have the largest gaps between maximum and minimum monthly precipitation of 308, 355, and 398 mm, respectively. These three southwestern stations feature a drastic increase in precipitation during the rainy season. From the equator to the northeast, Singkil, Sibolga, Lubuk Sulasih, and Padang have smaller gaps between maximum and minimum monthly precipitation of 263, 250, 224, and 283 mm, respectively. These four stations show gradual variations in precipitation that signal the onset of the rainy season. The onset of the rainy season in the remaining seven western coast stations is not easily defined due to less rainfall variation.

The three highest records of monthly rainfall among 24 colonial stations in 1879–1900 were 1878 mm (Padang Panjang, December 1879), 1443 mm (Padang Panjang, September 1879), and 1112 mm (Muntok, December 1898). The three lowest records of monthly rainfall were 0 mm (Tanjung Pinang, February 1881, and Bandar Lampung, September 1881) and 1 mm (Muntok, September 1881).

b. Climate classifications

Previous studies (Eguchi 1983; Aldrian and Susanto 2003; Aldrian et al. 2003) have used precipitation data after the 1960s, but in the present study, we used colonial data for 1879–1900. Two methods were followed to clarify the rainfall distribution in the colonial era. First, we compared the data with the definition of rainfall distribution proposed by Eguchi (1983) and only Bukittinggi, in the western Sumatera highlands, was classified as B–II. In Bukittinggi, the wettest month and secondary rainfall peak occurred in April and December, respectively. The remaining 23 colonial stations were classified as Region A. The 11 stations with a single primary rainfall peak (A–I) are Idirayeuk, Pajakumbuh, Padang Sidempuan (1.4°N, 99.3°E), Padang, Solok, Muntok, Bengkulu, Tebing Tinggi, Lahat, Pagar Alam, and Bandar Lampung (5.5°S, 105.3°E). The 12 stations with a secondary rainfall peak (A–II) are Banda Aceh, Medan (3.6°N, 98.7°E), Medan Putri (3.6°N, 98.7°E), Sibolga, Singkil, Lubuk Sulasih, Padang Panjang, Bengkalis, Jambi, Tanjung Pinang, Palembang, and Tanjung Pandan.

Second, we compared the data with the rainfall distribution definitions proposed by Aldrian and Susanto (2003). The borderline for Regions A′ and B′ is drawn near the 2°S latitude to divide Sumatera. Region A′, with a primary rainfall peak in December, includes nine colonial stations: Padang Sidempuan, Pajakumbuh, Solok, Bengkulu, Lahat, Tebing Tinggi, Pagar Alam, Muntok, and Bandar Lampung. We included Solok, Tebing Tinggi, Lahat, and Bandar Lampung into Region A′ because the wettest month at these four stations is also January. The 14 stations in Region B′ are Banda Aceh, Medan, Medan Putri, Singkil, Sibolga, Bengkalis, Jambi, Padang Panjang, Lubuk Sulasih, Palembang, Tanjung Pinang, and Tanjung Pandan. The rainfall distribution in Idirayeuk, Padang, and Bukittinggi could not be classified. In Idirayeuk, one of the driest regions in Sumatera throughout the year, the precipitation was below 100 mm month−1 during February–July. Additionally, the lowest monthly average precipitation occurred in March, at 55 mm month−1. The Idirayeuk rainfall characteristics did not correspond to the definition of Region A′, as the driest month should appear in July–September. Padang was not suitably categorized in Region A′, as rainfall was abundant throughout the year and lowest in February, with 262 mm month−1. Bukittinggi could not be categorized in Region B′ (although its rainfall characteristics are similar) because rainfall peaked in April and December at 269 and 246 mm month−1, respectively. Therefore, Idirayeuk, Padang, and Bukittinggi did not correspond to any of the rainfall distributions defined by Aldrian and Susanto (2003).

A–I stations were observed in northern Sumatera in the colonial era, but A–I migrated southward until the modern era. B–II station was observed in Bukittinggi but migrated southeast, Banka Island. Region A′ and B′ stations were found near the northwestern and southeastern areas, respectively, in the colonial era. Region A′ on the northwestern coast migrated southward, and Region B′ in the southeastern coast migrated northward over the century. The results showed that the 24 colonial stations did not clearly follow the rainfall distribution defined by the two previous studies, which used rainfall data from 1978 to 1979 and from 1961 to 1993. Notably, six colonial stations surrounding the Padang region, Sumatera Barat Province were relatively difficult to classify. Although the locations of these six stations were not distant from each other, the rainfall distribution could not be integrated as a single definition.

5. Interannual variations

a. Centennial and decadal variations

To clarify the centennial variations of rainfall in Sumatera, we compared the total average annual precipitation in the three regions of Sumatera, the eastern coast, the western coast, and the east of the Barisan ridge, during each period of 1879–1900, 1931–60, and 1971–2014. For the 1879–1900 period, the average annual precipitation at the colonial stations on the eastern coast, western coast, and east of the Barisan ridge was 2441, 3466, and 2844 mm yr−1, respectively. For the 1931–60 period, the average annual precipitation at the Pawitan stations on the eastern coast, western coast, and east of the Barisan ridge was 2603, 3357, and 2566 mm yr−1, respectively. For the 1971–2014 period, the average annual precipitation at the BMKG stations on the eastern coast, western coast, and east of the Barisan ridge was 2273, 3664, and 2134 mm yr−1, respectively. The annual average precipitation decreased during the 1971–2014 period in stations located on the eastern coast and east of the Barisan ridge over the last century. Annual precipitation at the western coast stations was found to increase gradually toward current day values.

For the entire Sumatera Island, the average annual precipitation for the 1879–1900, 1931–60, and 1971–2014 periods were 2917, 2842, and 2690 mm yr−1, respectively. For the colonial rainfall data (1879–1900) and Pawitan data (1931–60), the total annual precipitation in Sumatera slightly decreased toward the BMKG data (1971–2014). Notably, although observation methods and meteorological equipment differ between the modern and colonial eras, Siswanto et al. (2016) noted that the measurement techniques, calibration, and maintenance procedure at the Batavia station in colonial times remained similar in the current observation period. All of the local stations, including Sumatera, adopted the strict rule provided by Batavia, which is why our study also used colonial rainfall data as a valid dataset, as Siswanto et al. (2016) did. Sumatera appears to have been affected by global warming, leading to a decline in annual precipitation since 1879.

Figures 5a and 5b show the annual precipitation anomalies for the 1879–1900 and 1971–2014 periods, respectively. The precipitation anomalies during 1879–1900 and 1971–2014 are shown in three regions: the western coast, eastern coast, and the east of the Barisan ridge. The rainfall data for Pawitan from 1931 to 1960 were excluded from Fig. 5 because annual precipitation data had already been averaged, and therefore the precipitation observed for each year is not shown. Figure 6 indicates the standardized annual average precipitation in 24 colonial stations during 1879–1900. Figure 7 shows the standardized annual average precipitation during the dry season (May–October) for the 1879–1900 period.

Fig. 5.
Fig. 5.

Anomaly in the annual average precipitation of three regions in Sumatera for (a) 1879–1900 and (b) 1971–2014. The black solid line, black dotted line, and black doubled line indicate the western coast, east of the Barisan ridge, and eastern coast, respectively. The solid red line and blue lines indicate the dipole mode index (DMI) and multivariate ENSO index, respectively. Gray shades show El Niño event years.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

Fig. 6.
Fig. 6.

Standardized annual precipitation in 1879–1900. Station numbers and geographical coordinates are the same as given in Table 1. The plots indicate (a) 9 western coast stations, (b) 5 stations in the east of the Barisan ridge, and (c) 10 eastern coast stations. Vertical gray shades indicate the 1888/89 El Niño event year.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

Fig. 7.
Fig. 7.

Standardized annual dry-season precipitation (May–October) at six Southern Hemispheric climate-type stations from 1879 to 1900. The red solid line and blue dotted line indicate DMI and extended multivariate ENSO index (MEI.ext), respectively. The black solid bar, gray solid bar, and open bar indicate a western coast station [Bengkulu (3.8°S, 102.3°E)], an eastern coast station [Muntok (2.1°S, 105.2°E)], and four stations east of the Barisan ridge [Pagar Alam (4.1°S, 103.4°E), Lahat (3.8°S, 103.5°E), Tebing Tinggi (3.6°S, 103.1°E), and Bandar Lampung (5.5°S, 105.3°E)], respectively.

Citation: Journal of Hydrometeorology 23, 10; 10.1175/JHM-D-20-0245.1

The decadal rainfall variations from 1879 to 1900 and from 1971 to 2014 are shown in Figs. 57. In the late-nineteenth century, the annual rainfall appeared to have gradually increased until 1900, but the increase was not as drastic as in 1971–2014. After the strong 1972/73 El Niño events, the annual rainfall increased until 1997. Very strong El Niño events occurred again in 1997/98, leading to unstable decadal variations until 2014. The rainfall variations in the two decades of the colonial era appeared to be more stable than in 1971–2014.

b. Comparison with ENSO

Wolter and Timlin (1993, 1998) determined the definition of the multivariate ENSO index (MEI) using the principal component of six observed fields: sea level pressure, zonal and meridional surface wind components, sea surface temperature, near-surface air temperatures, and total cloudiness. The original MEI base period was 1950–93. Wolter and Timlin (2011) extended the time series of the multivariate ENSO index (MEI.ext) from 1871 to 2005, with three overlapping periods (1870–1920, 1915–65, and 1960–2005). They analyzed the cold and warm ENSO phases with El Niño and La Niña events. In the late-nineteenth century, El Niño occurred twice, during 1877/78 and 1888/89. In these El Niño event periods, MEI.ext indicates standardized departures equal to or above +2. The El Niño event of 1877/78 was similar to the extreme El Niño events in 1982/83 and 1997/98. However, the 1888/89 El Niño event was not as severe as the 1877/78 event, with a standard deviation below +2 for MEI.ext.

We compared the annual rainfall variation with El Niño events in Sumatera (Figs. 6a–c). The actual observed rainfall amount during 1877/78 was imprecise because the colonial observations began on 1 January 1879. Therefore, we focused on the El Niño event of 1888/89. Consequently, there was a decline in rainfall at the three western coast (Bukittinggi, Padang Panjang, Solok), three eastern coast (Banda Aceh, Idirayeuk, Medan), and two east of the ridge (Pajakumbuh, Lahat) stations because of the El Niño event. The remaining 16 stations showed a modest increase in precipitation during El Niño periods, whereas the other years showed a larger decline than those during El Niño events. El Niño primarily affected the regions south of the equator, while still affecting the northernmost region. Most stations north of the equator were unaffected by El Niño events.

Hamada et al. (2012) summarized that the rainfall in northwestern Jawa Island is positively correlated with local and remote sea surface temperatures during the dry season (May–October), but uncorrelated with equatorial sea surface temperature during the rainy season (November–April). Western Jawa, Banten Province, is the closest region to southern Sumatera, especially in Lampung Province. Therefore, we compared the annual precipitation in the southern Sumatera stations with the Southern Hemispheric climate type to determine whether there are any relationships with ENSO events. We selected six southern stations determined as “A–I” using the definitions by Eguchi (1983), which correspond to the Southern Hemispheric climate type (see Fig. 7). Figure 7 only includes the dry-season precipitation during May–October, showing the six southern stations in three regions: the western coast, eastern coast, and east of the Barisan ridge. The dipole mode index (DMI) and MEI.ext values are available on the National Oceanic and Atmospheric Administration’s Physical Sciences Laboratory website (NOAA/PSL 2020, 2011). DMI values were reconstructed from 1870 to the present. Original bimonthly MEI.ext data values from December 1870 to December 2005 are available. Negative and positive values in the indices represent La Niña and El Niño events, respectively (Wolter and Timlin 2011).

The 1877/78 El Niño events were strong, but the following year, 1879, was a moderate La Niña year (see Fig. 7). Both MEI.ext and DMI indicate negative values of −0.74 and −0.22 in 1879, respectively. The standardized annual dry-season precipitation in the six stations of the Southern Hemispheric climate type indicate positive values due to moderate La Niña events. In 1888, the positive values of MEI.ext reached +1.36, which is the largest value during 1879–1900, and the annual dry-season precipitation was negative. Although 1888 had the largest positive values of MEI.ext, DMI indicated negative values of −0.15. In 1896 and 1885, MEI.ext showed positive values of +1.11 and +0.38, respectively, and the annual dry-season precipitation showed a significant decrease and higher than that of 1888. DMI also indicated positive values of +0.24 and +0.07 for 1896 and 1885, respectively. In 1881, the annual dry-season precipitation was remarkably lower relative to 1885 and 1896. MEI.ext had negative values of −0.36, but DMI indicated positive values of +0.14. Significant decreasing trends in 1896 occurred in all three regions of the western coast, eastern coast, and east of the Barisan ridge as the standardized annual dry-season precipitation showed negative values of −2.10, −2.12, and −1.79, respectively. Decreasing trends in 1885 primarily occurred only in the western and eastern coasts as −2.22 and −1.79, respectively, and decreasing trends in 1881 were primarily occurred in the eastern coast with a value of −1.55. We found a significant decrease in the annual dry-season precipitation in 1881, 1885, and 1896 when DMI was positive. Consequently, the dry-season precipitation in the nineteenth century was not affected by ENSO and El Niño events but was definitely influenced by the Indian Ocean dipole (IOD).

6. Conclusions

This study reconstructed the historical precipitation records at 24 observation stations in Sumatera between 1879 and 1900. The series of colonial materials Regenwaarnemingen in Nederlandsch-Indië is the most important data source for historical rainfall data in colonial Indonesia. Although there is missing data for some the monthly records, the amount is not large. Regional data for Sumatera are lacking in the current operational database. A comparison analysis was conducted to clarify the rainfall variation in the colonial era and the latter periods from 1931 to 1960 and from 1971 to 2014. Further study and analysis are still required to understand the relationships between peatland fires and rainfall in the colonial Sumatera.

We compared the precipitation records from the 24 colonial stations during 1879–1900 with the 28 stations in 1931–60 and with the 30 stations in 1971–2014. The longitudinal and latitudinal locations show regional differences in rainfall variations over the colonial Sumatera. Abundant rainfall can be observed on the western coast of Sumatera at 97°–103°E and 3°N–4°S, with an annual precipitation above 3000 mm yr−1. The ten eastern coast stations have an annual precipitation below 3000 mm yr−1, indicating a drier environment than the western coast (Figs. 2a,b,d,e). The coastal rainfall characteristics, specifically on the western coastlines of Sumatera, show a similar variation to the analysis conducted in Indonesia from 1997 to 2011 and from 1961 to 1990 by Ogino et al. (2016, 2017) and Hamada et al. (2002), respectively. The eastern coast regions do not represent such coastal rainfall characteristics; only limited annual rainfall is observed from north to south. The stations east of the Barisan ridge show a similar variation to the eastern coast. The Barisan ridge is elongated north–south in the western part of Sumatera, but the rainfall characteristics do not match those of the western coast regions (Figs. 2c,f).

We found that the monthly precipitation declined in a specific area of Sumatera in three different periods, 1879–1900, 1931–60, and 1971–2014 (Figs. 3a–c). In the western coast stations located north of 0.5°S, the monthly precipitation decreased in January–May, but rainfall increased during June–December in the last century. In the two eastern coast stations, Banda Aceh and Bengkalis, the rainfall variation shows opposite trends to those of the western coast stations. Monthly precipitation increased in January–May and decreased in June–December. This implies that the rainfall decline caused the recent severe dry season and frequent forest and peatland fires in the Riau and Aceh Provinces.

We attempted to clarify the differences in the rainfall distribution at each station and compared the results with the climate classifications defined by Eguchi (1983), Aldrian and Susanto (2003), and Wolter and Timlin (2011). As the rainfall characteristics in colonial Sumatera displayed only slightly similar trends to previous studies, a clear and definitive rainfall distribution in colonial Sumatera is difficult to conclude. The western coast stations on the equator, in particular, cannot be classified according to rainfall distribution.

The interannual variation analysis and a comparison with MEI.ext and DMI revealed that the El Niño events of 1888/89 had low impact on the precipitation in colonial Sumatera. The positive values of MEI.ext indicate that rainfall in the three regions decreased. Even though a strong El Niño event occurred in 1888/89, the precipitation in the three regions did not decrease as much as in 1881, 1885, and 1896. The positive values of DMI indicate a significant decrease in rainfall in 1881, 1885, and 1896; consequently, the IOD had much influence on the rainfall decline than the influence of the El Niño event. El Niño events caused severe drought conditions in modern Sumatera and massive forest and peatland fires in 1997/98 and 2015/16. This study implies that colonial Sumatera experienced fewer severe drought conditions even in the El Niño events of the late-nineteenth century.

The reconstruction of historical precipitation data in other regions of Indonesia should be the subject of future study. Moreover, peatland fires and haze in Southeast Asia are critical environmental problems that should be investigated. We clarified the rainfall variation in the late-nineteenth century, but we also need to connect climate and development in Sumatera using other social records, such as colonial newspapers and regional administrative reports. Such an interdisciplinary study is also important for the future of Earth and our society.

Acknowledgments.

This research was supported by the Research Institute for Humanity and Nature (RIHN: a constituent member of NIHU) Project 14200117. In addition, this work was supported by JSPS KAKENHI Grant JP18K18269. We acknowledge Dr. Jun-Ichi Hamada for his careful reading of the original paper. We deeply appreciate the great support from the Special Collection Room of Leiden University Libraries and the National Library of Republic Indonesia for our survey material. We acknowledge the data provided by the BMKG, Data IKLIM, for meteorological data (http://dataonline.bmkg.go.id/home). We also acknowledge the data providers in the SACA&D project (http://sacad.database.bmkg.go.id; van den Besselaar et al. 2017).

REFERENCES

  • Aldrian, E., and R. D. Susanto, 2003: Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. Int. J. Climatol., 23, 14351452, https://doi.org/10.1002/joc.950.

    • Search Google Scholar
    • Export Citation
  • Aldrian, E., L. D. Gates, and F. H. Widodo, 2003: Variability of Indonesian Rainfall and the Influence of ENSO and Resolution in ECHAM4 simulations and in the Reanalyses, MPI Rep., 346, 30 pp., https://mpimet.mpg.de/en/science/publications/archive/mpi-reports-1987-2004.

  • As-syakur, A. R., and Coauthors, 2016: Maritime Continent rainfall variability during the TRMM era: The role of monsoon, topography and El Niño Modoki. Dyn. Atmos. Oceans, 75, 5877, https://doi.org/10.1016/j.dynatmoce.2016.05.004.

    • Search Google Scholar
    • Export Citation
  • As-syakur, A. R., K. Imaoka, K. Ogawara, M. D. Yamanaka, T. Tanaka, Y. Kashino, I. W. Nuarsa, and T. Osawa, 2019: Analysis of spatial and seasonal differences in the diurnal rainfall cycle over Sumatera revealed by 17-year TRMM 3B42 dataset. SOLA, 15, 216221, https://doi.org/10.2151/sola.2019-039.

    • Search Google Scholar
    • Export Citation
  • Bergsma, P. A., 1880: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 1, Batavia Landsdrukkerij, 257 pp.

  • Bergsma, P. A., 1881: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 2, Batavia Landsdrukkerij, 291 pp.

  • Bergsma, P. A., 1882: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 3, Batavia Landsdrukkerij, 305 pp.

  • Berlage, H. P., Jr., 1949: Regenval in Indonesië: Gemiddelden van de regenval voor 4339 waarnemingsplaatsen in Indonesië, berekend uit waarnemingen verricht in het tijdvak 1879–1941 (in Dutch). Verh. K. Magn. Meteor. Obs. Batavia, 37, 212 pp.

  • BMKG, 2015: Data Iklim (in Indonesian). Data Online Pusat Database, accessed 5 September 2019, http://dataonline.bmkg.go.id/home.

  • Braak, C., 1921: Het Klimaat van Nederlandsch-Indie (in Dutch). Verh. K. Magn. Meteor. Obs. Batavia , 8, 272 pp.

  • Brill, E. J., 1897: Woordenlijst van Eenige Aardrijkskundige Namen in Nederlandsch Oost-Indië (in Dutch). Het Koninklijk Nederlandsch Aardrijkundig Genootschap Boekhandel en Drukkerij, 122 pp.

  • Eguchi, T., 1983: Rainfall distribution and air streams over Indonesia. Geogr. Rev. Japan, 56, 151170, https://doi.org/10.4157/grj.56.151.

    • Search Google Scholar
    • Export Citation
  • Field, R., and Coauthors, 2016: Indonesian fire activity and smoke pollution in 2015 show persistent nonlinear sensitivity to El Niño-induced drought. Proc. Natl. Acad. Sci. USA, 113, 92049209, https://doi.org/10.1073/pnas.1524888113.

    • Search Google Scholar
    • Export Citation
  • Figee, S., 1885: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 6, Batavia Landsdrukkerij, 408 pp.

  • Figee, S., 1899: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 20, Batavia Landsdrukkerij, 483 pp.

  • Figee, S., 1900: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 21, Batavia Landsdrukkerij, 476 pp.

  • Figee, S., 1901: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 22, Batavia Landsdrukkerij, 493 pp.

  • Hamada, J., 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, 285310, https://doi.org/10.2151/jmsj.80.285.

    • Search Google Scholar
    • Export Citation
  • Hamada, J., M. D. Yamanaka, S. Mori, Y. I. Tauhid, and T. Sribimawati, 2008: Differences of rainfall characteristics between coastal and interior areas of central western Sumatera, Indonesia. J. Meteor. Soc. Japan, 86, 593611, https://doi.org/10.2151/jmsj.86.593.

    • Search Google Scholar
    • Export Citation
  • Hamada, J., S. Mori, M. D. Yamanaka, U. Haryoko, S. Lestari, R. Sulistyowati, and F. Syamsudin, 2012: Interannual rainfall variability over northwestern Jawa and its relation to the Indian Ocean dipole and El Niño southern-oscillation events. SOLA, 8, 6972, https://doi.org/10.2151/sola.2012-018.

    • Search Google Scholar
    • Export Citation
  • Inoue, T., and K. Nakamura, 1990: Physical and biological background for insect studies in Sumatra. Natural History of Social Wasps and Bees in Equatorial Sumatra, Hokkaido University Press, 1–11.

  • Kajita, R., 2019: Historical precipitation data in Sumatra and Kalimantan from 1879 to 1900, by using Dutch colonial materials. IOP Conf. Ser. Earth Environ. Sci., 361, 012003, https://doi.org/10.1088/1755-1315/361/1/012003.

    • Search Google Scholar
    • Export Citation
  • Können, G. P., P. D. Jones, M. H. Kaltofen, and R. J. Allan, 1998: Pre-1866 extensions of the Southern Oscillation index using early Indonesian and Tahitian meteorological readings. J. Climate, 11, 23252339, https://doi.org/10.1175/1520-0442(1998)011<2325:PEOTSO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kubota, H., R. Shirooka, J. Hamada, and F. Syamsudin, 2011: Interannual rainfall variability over the eastern Maritime Continent. J. Meteor. Soc. Japan, 89A, 111122, https://doi.org/10.2151/jmsj.2011-A07.

    • Search Google Scholar
    • Export Citation
  • Lestari, S., J.-I. Hamada, F. Syamsudin, Sunaryo, J. Matsumoto, and M. D. Yamanaka, 2016: ENSO influences on rainfall extremes around Sulawesi and Maluku Islands in the eastern Indonesian Maritime Continent. SOLA, 12, 3741, https://doi.org/10.2151/sola.2016-008.

    • Search Google Scholar
    • Export Citation
  • Loeb, E. M., 1935: Sumatra Its History and People. Oxford University Press, 350 pp.

  • NOAA/PSL, 2011: Extended multivariate ENSO index (MEI.ext): Outline for MEI.ext webpage (first edition on May 10th, 2011). NOAA/PSL, accessed 14 December 2020, https://psl.noaa.gov/enso/mei.ext/.

  • NOAA/PSL, 2020: DMI: Standard PSL Format. NOAA/PSL, accessed 14 December 2020, https://psl.noaa.gov/gcos_wgsp/Timeseries/DMI/

  • Ogino, S., M. D. Yamanaka, S. Mori, and J. Matsumoto, 2016: How much is the precipitation amount over the tropical coastal region? J. Climate, 29, 12311236, https://doi.org/10.1175/JCLI-D-15-0484.1.

    • Search Google Scholar
    • Export Citation
  • Ogino, S., M. D. Yamanaka, S. Mori, and J. Matsumoto, 2017: Tropical coastal dehydrator in global atmospheric water circulation. Geophys. Res. Lett., 44, 11 63611 643, https://doi.org/10.1002/2017GL075760.

    • Search Google Scholar
    • Export Citation
  • Paulus, J., 1917: Bengkalis. Encyclopaedie van Nederlandsch-Indië (in Dutch) , 2nd ed. Vol. 1, Nijhoff-Brill, 267–268.

  • Pawitan, H., 1990: Climate data compilation for Sumatra. Center for Soil and Agroclimate Research Tech. Rep. 23, 13 pp.

  • Poortman, J. J., 1884: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 5, Batavia Landsdrukkerij, 360 pp.

  • Project Team SACA&D, 2011: Southeast Asian Climate Assessment & Dataset (SACA&D) project. Accessed 5 September 2019, http://sacad.database.bmkg.go.id/.

  • Qian, J.-H., 2020: Multi‐scale climate processes and rainfall variability in Sumatra and Malay Peninsula associated with NSO in boreal fall and winter. Int. J. Climatol., 40, 41714188, https://doi.org/10.1002/joc.6450.

    • Search Google Scholar
    • Export Citation
  • Siegert, F., G. Ruecker, A. Hinrichs, and A. A. Hoffmann, 2001: Increased damage from fires in logged forests during droughts caused by El Niño. Nature, 414, 437440, https://doi.org/10.1038/35106547.

    • Search Google Scholar
    • Export Citation
  • Siswanto, S., G. van Oldenborgh, G. van der Schrier, R. Jilderda, and B. van den Hurk, 2016: Temperature, extreme precipitation, and diurnal rainfall changes in the urbanized Jakarta city during the past 130 years. Int. J. Climatol., 36, 32073225, https://doi.org/10.1002/joc.4548.

    • Search Google Scholar
    • Export Citation
  • Stibbe, D. G., W. C. B. Wintgens, and E. M. Uhlenbeck, 1919: Oostkust van Sumatra. Encyclopaedie van Nederlandsch-Indië (in Dutch) . 2nd ed. Vol. 3, Nijhoff-Brill, 140–154.

    • Search Google Scholar
    • Export Citation
  • van den Besselaar, E. J. M., G. van der Schrier, R. C. Cornes, A. S. Iqbal, and A. M. G. Klein Tank, 2017: SA-OBS: A daily gridded surface temperature and precipitation dataset for Southeast Asia. J. Climate, 30, 51515165, https://doi.org/10.1175/JCLI-D-16-0575.1.

    • Search Google Scholar
    • Export Citation
  • van der Stok, J. P., 1883: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 4, Batavia Landsdrukkerij, 341 pp.

  • van der Stok, J. P., 1886: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 7, Batavia Landsdrukkerij, 408 pp.

  • van der Stok, J. P., 1887: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 8, Batavia Landsdrukkerij, 412 pp.

  • van der Stok, J. P., 1888: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 9, Batavia Landsdrukkerij, 413 pp.

  • van der Stok, J. P., 1889: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 10, Batavia Landsdrukkerij, 420 pp.

  • van der Stok, J. P., 1890: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 11, Batavia Landsdrukkerij, 417 pp.

  • van der Stok, J. P., 1891: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 12, Batavia Landsdrukkerij, 418 pp.

  • van der Stok, J. P., 1892: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 13, Batavia Landsdrukkerij, 415 pp.

  • van der Stok, J. P., 1893: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 14, Batavia Landsdrukkerij, 413 pp.

  • van der Stok, J. P., 1894: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 15, Batavia Landsdrukkerij, 415 pp.

  • van der Stok, J. P., 1895: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 16, Batavia Landsdrukkerij, 421 pp.

  • van der Stok, J. P., 1896: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 17, Batavia Landsdrukkerij, 423 pp.

  • van der Stok, J. P., 1897: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 18, Batavia Landsdrukkerij, 468 pp.

  • van der Stok, J. P., 1898: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 19, Batavia Landsdrukkerij, 552 pp.

  • Vose, R. S., R. L. Schmoyer, P. M. Steurer, T. C. Peterson, R. Heim, T. R. Karl, and J. K. Eischeid, 1992: The Global Historical Climatology Network: Long-term monthly temperature, precipitation, sea level pressure, and station pressure data. Oak Ridge National Laboratory Environmental Sciences Division Publ. 3912, ORNL/CDIAC-53 NDP-041, 324 pp., https://doi.org/10.3334/CDIAC/cli.ndp041.

  • Wati, T., S. D. A. Kusumaningtyas, and E. Aldrian, 2019: Study of season onset based on water requirement assessment. IOP Conf. Ser. Earth Environ. Sci., 299, 012042, https://doi.org/10.1088/1755-1315/299/1/012042.

    • Search Google Scholar
    • Export Citation
  • Wolter, K., and M. S. Timlin, 1993: Monitoring ENSO in COADS with a seasonally adjusted principal component index. Proc. 17th Climate Diagnostics Workshop, Norman, OK, University of Oklahoma, 5257.

  • Wolter, K., and M. S. Timlin, 1998: Measuring the strength of ENSO events: How does 1997/98 rank? Weather, 53, 315324, https://doi.org/10.1002/j.1477-8696.1998.tb06408.x.

    • Search Google Scholar
    • Export Citation
  • Wolter, K., and M. S. Timlin, 2011: El Niño/Southern Oscillation behaviour since 1871 as diagnosed in an extended multivariate ENSO index (MEI.ext). Int. J. Climatol., 31, 10741087, https://doi.org/10.1002/joc.2336.

    • Search Google Scholar
    • Export Citation
  • Yamanaka, M. D., S.-Y. Ogino, P.-M. Wu, J.-I. Hamada, S. Mori, J. Matsumoto, and F. Syamsudin, 2018: Maritime Continent coastlines controlling Earth’s climate. Prog. Earth Planet. Sci., 5, 21, https://doi.org/10.1186/s40645-018-0174-9.

    • Search Google Scholar
    • Export Citation
  • Yanto, B. Rajagopalan, and E. Zagona, 2016: Space–time variability of Indonesian rainfall at inter-annual and multi-decadal time scales. Climate Dyn., 47, 29752989, https://doi.org/10.1007/s00382-016-3008-8.

    • Search Google Scholar
    • Export Citation

Supplementary Materials

Save
  • Aldrian, E., and R. D. Susanto, 2003: Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. Int. J. Climatol., 23, 14351452, https://doi.org/10.1002/joc.950.

    • Search Google Scholar
    • Export Citation
  • Aldrian, E., L. D. Gates, and F. H. Widodo, 2003: Variability of Indonesian Rainfall and the Influence of ENSO and Resolution in ECHAM4 simulations and in the Reanalyses, MPI Rep., 346, 30 pp., https://mpimet.mpg.de/en/science/publications/archive/mpi-reports-1987-2004.

  • As-syakur, A. R., and Coauthors, 2016: Maritime Continent rainfall variability during the TRMM era: The role of monsoon, topography and El Niño Modoki. Dyn. Atmos. Oceans, 75, 5877, https://doi.org/10.1016/j.dynatmoce.2016.05.004.

    • Search Google Scholar
    • Export Citation
  • As-syakur, A. R., K. Imaoka, K. Ogawara, M. D. Yamanaka, T. Tanaka, Y. Kashino, I. W. Nuarsa, and T. Osawa, 2019: Analysis of spatial and seasonal differences in the diurnal rainfall cycle over Sumatera revealed by 17-year TRMM 3B42 dataset. SOLA, 15, 216221, https://doi.org/10.2151/sola.2019-039.

    • Search Google Scholar
    • Export Citation
  • Bergsma, P. A., 1880: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 1, Batavia Landsdrukkerij, 257 pp.

  • Bergsma, P. A., 1881: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 2, Batavia Landsdrukkerij, 291 pp.

  • Bergsma, P. A., 1882: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 3, Batavia Landsdrukkerij, 305 pp.

  • Berlage, H. P., Jr., 1949: Regenval in Indonesië: Gemiddelden van de regenval voor 4339 waarnemingsplaatsen in Indonesië, berekend uit waarnemingen verricht in het tijdvak 1879–1941 (in Dutch). Verh. K. Magn. Meteor. Obs. Batavia, 37, 212 pp.

  • BMKG, 2015: Data Iklim (in Indonesian). Data Online Pusat Database, accessed 5 September 2019, http://dataonline.bmkg.go.id/home.

  • Braak, C., 1921: Het Klimaat van Nederlandsch-Indie (in Dutch). Verh. K. Magn. Meteor. Obs. Batavia , 8, 272 pp.

  • Brill, E. J., 1897: Woordenlijst van Eenige Aardrijkskundige Namen in Nederlandsch Oost-Indië (in Dutch). Het Koninklijk Nederlandsch Aardrijkundig Genootschap Boekhandel en Drukkerij, 122 pp.

  • Eguchi, T., 1983: Rainfall distribution and air streams over Indonesia. Geogr. Rev. Japan, 56, 151170, https://doi.org/10.4157/grj.56.151.

    • Search Google Scholar
    • Export Citation
  • Field, R., and Coauthors, 2016: Indonesian fire activity and smoke pollution in 2015 show persistent nonlinear sensitivity to El Niño-induced drought. Proc. Natl. Acad. Sci. USA, 113, 92049209, https://doi.org/10.1073/pnas.1524888113.

    • Search Google Scholar
    • Export Citation
  • Figee, S., 1885: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 6, Batavia Landsdrukkerij, 408 pp.

  • Figee, S., 1899: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 20, Batavia Landsdrukkerij, 483 pp.

  • Figee, S., 1900: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 21, Batavia Landsdrukkerij, 476 pp.

  • Figee, S., 1901: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 22, Batavia Landsdrukkerij, 493 pp.

  • Hamada, J., 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, 285310, https://doi.org/10.2151/jmsj.80.285.

    • Search Google Scholar
    • Export Citation
  • Hamada, J., M. D. Yamanaka, S. Mori, Y. I. Tauhid, and T. Sribimawati, 2008: Differences of rainfall characteristics between coastal and interior areas of central western Sumatera, Indonesia. J. Meteor. Soc. Japan, 86, 593611, https://doi.org/10.2151/jmsj.86.593.

    • Search Google Scholar
    • Export Citation
  • Hamada, J., S. Mori, M. D. Yamanaka, U. Haryoko, S. Lestari, R. Sulistyowati, and F. Syamsudin, 2012: Interannual rainfall variability over northwestern Jawa and its relation to the Indian Ocean dipole and El Niño southern-oscillation events. SOLA, 8, 6972, https://doi.org/10.2151/sola.2012-018.

    • Search Google Scholar
    • Export Citation
  • Inoue, T., and K. Nakamura, 1990: Physical and biological background for insect studies in Sumatra. Natural History of Social Wasps and Bees in Equatorial Sumatra, Hokkaido University Press, 1–11.

  • Kajita, R., 2019: Historical precipitation data in Sumatra and Kalimantan from 1879 to 1900, by using Dutch colonial materials. IOP Conf. Ser. Earth Environ. Sci., 361, 012003, https://doi.org/10.1088/1755-1315/361/1/012003.

    • Search Google Scholar
    • Export Citation
  • Können, G. P., P. D. Jones, M. H. Kaltofen, and R. J. Allan, 1998: Pre-1866 extensions of the Southern Oscillation index using early Indonesian and Tahitian meteorological readings. J. Climate, 11, 23252339, https://doi.org/10.1175/1520-0442(1998)011<2325:PEOTSO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kubota, H., R. Shirooka, J. Hamada, and F. Syamsudin, 2011: Interannual rainfall variability over the eastern Maritime Continent. J. Meteor. Soc. Japan, 89A, 111122, https://doi.org/10.2151/jmsj.2011-A07.

    • Search Google Scholar
    • Export Citation
  • Lestari, S., J.-I. Hamada, F. Syamsudin, Sunaryo, J. Matsumoto, and M. D. Yamanaka, 2016: ENSO influences on rainfall extremes around Sulawesi and Maluku Islands in the eastern Indonesian Maritime Continent. SOLA, 12, 3741, https://doi.org/10.2151/sola.2016-008.

    • Search Google Scholar
    • Export Citation
  • Loeb, E. M., 1935: Sumatra Its History and People. Oxford University Press, 350 pp.

  • NOAA/PSL, 2011: Extended multivariate ENSO index (MEI.ext): Outline for MEI.ext webpage (first edition on May 10th, 2011). NOAA/PSL, accessed 14 December 2020, https://psl.noaa.gov/enso/mei.ext/.

  • NOAA/PSL, 2020: DMI: Standard PSL Format. NOAA/PSL, accessed 14 December 2020, https://psl.noaa.gov/gcos_wgsp/Timeseries/DMI/

  • Ogino, S., M. D. Yamanaka, S. Mori, and J. Matsumoto, 2016: How much is the precipitation amount over the tropical coastal region? J. Climate, 29, 12311236, https://doi.org/10.1175/JCLI-D-15-0484.1.

    • Search Google Scholar
    • Export Citation
  • Ogino, S., M. D. Yamanaka, S. Mori, and J. Matsumoto, 2017: Tropical coastal dehydrator in global atmospheric water circulation. Geophys. Res. Lett., 44, 11 63611 643, https://doi.org/10.1002/2017GL075760.

    • Search Google Scholar
    • Export Citation
  • Paulus, J., 1917: Bengkalis. Encyclopaedie van Nederlandsch-Indië (in Dutch) , 2nd ed. Vol. 1, Nijhoff-Brill, 267–268.

  • Pawitan, H., 1990: Climate data compilation for Sumatra. Center for Soil and Agroclimate Research Tech. Rep. 23, 13 pp.

  • Poortman, J. J., 1884: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 5, Batavia Landsdrukkerij, 360 pp.

  • Project Team SACA&D, 2011: Southeast Asian Climate Assessment & Dataset (SACA&D) project. Accessed 5 September 2019, http://sacad.database.bmkg.go.id/.

  • Qian, J.-H., 2020: Multi‐scale climate processes and rainfall variability in Sumatra and Malay Peninsula associated with NSO in boreal fall and winter. Int. J. Climatol., 40, 41714188, https://doi.org/10.1002/joc.6450.

    • Search Google Scholar
    • Export Citation
  • Siegert, F., G. Ruecker, A. Hinrichs, and A. A. Hoffmann, 2001: Increased damage from fires in logged forests during droughts caused by El Niño. Nature, 414, 437440, https://doi.org/10.1038/35106547.

    • Search Google Scholar
    • Export Citation
  • Siswanto, S., G. van Oldenborgh, G. van der Schrier, R. Jilderda, and B. van den Hurk, 2016: Temperature, extreme precipitation, and diurnal rainfall changes in the urbanized Jakarta city during the past 130 years. Int. J. Climatol., 36, 32073225, https://doi.org/10.1002/joc.4548.

    • Search Google Scholar
    • Export Citation
  • Stibbe, D. G., W. C. B. Wintgens, and E. M. Uhlenbeck, 1919: Oostkust van Sumatra. Encyclopaedie van Nederlandsch-Indië (in Dutch) . 2nd ed. Vol. 3, Nijhoff-Brill, 140–154.

    • Search Google Scholar
    • Export Citation
  • van den Besselaar, E. J. M., G. van der Schrier, R. C. Cornes, A. S. Iqbal, and A. M. G. Klein Tank, 2017: SA-OBS: A daily gridded surface temperature and precipitation dataset for Southeast Asia. J. Climate, 30, 51515165, https://doi.org/10.1175/JCLI-D-16-0575.1.

    • Search Google Scholar
    • Export Citation
  • van der Stok, J. P., 1883: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 4, Batavia Landsdrukkerij, 341 pp.

  • van der Stok, J. P., 1886: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 7, Batavia Landsdrukkerij, 408 pp.

  • van der Stok, J. P., 1887: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 8, Batavia Landsdrukkerij, 412 pp.

  • van der Stok, J. P., 1888: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 9, Batavia Landsdrukkerij, 413 pp.

  • van der Stok, J. P., 1889: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 10, Batavia Landsdrukkerij, 420 pp.

  • van der Stok, J. P., 1890: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 11, Batavia Landsdrukkerij, 417 pp.

  • van der Stok, J. P., 1891: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 12, Batavia Landsdrukkerij, 418 pp.

  • van der Stok, J. P., 1892: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 13, Batavia Landsdrukkerij, 415 pp.

  • van der Stok, J. P., 1893: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 14, Batavia Landsdrukkerij, 413 pp.

  • van der Stok, J. P., 1894: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 15, Batavia Landsdrukkerij, 415 pp.

  • van der Stok, J. P., 1895: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 16, Batavia Landsdrukkerij, 421 pp.

  • van der Stok, J. P., 1896: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 17, Batavia Landsdrukkerij, 423 pp.

  • van der Stok, J. P., 1897: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 18, Batavia Landsdrukkerij, 468 pp.

  • van der Stok, J. P., 1898: Regenwaarnemingen in Nederlandsch-Indië (in Dutch). Vol. 19, Batavia Landsdrukkerij, 552 pp.

  • Vose, R. S., R. L. Schmoyer, P. M. Steurer, T. C. Peterson, R. Heim, T. R. Karl, and J. K. Eischeid, 1992: The Global Historical Climatology Network: Long-term monthly temperature, precipitation, sea level pressure, and station pressure data. Oak Ridge National Laboratory Environmental Sciences Division Publ. 3912, ORNL/CDIAC-53 NDP-041, 324 pp., https://doi.org/10.3334/CDIAC/cli.ndp041.

  • Wati, T., S. D. A. Kusumaningtyas, and E. Aldrian, 2019: Study of season onset based on water requirement assessment. IOP Conf. Ser. Earth Environ. Sci., 299, 012042, https://doi.org/10.1088/1755-1315/299/1/012042.

    • Search Google Scholar
    • Export Citation
  • Wolter, K., and M. S. Timlin, 1993: Monitoring ENSO in COADS with a seasonally adjusted principal component index. Proc. 17th Climate Diagnostics Workshop, Norman, OK, University of Oklahoma, 5257.

  • Wolter, K., and M. S. Timlin, 1998: Measuring the strength of ENSO events: How does 1997/98 rank? Weather, 53, 315324, https://doi.org/10.1002/j.1477-8696.1998.tb06408.x.

    • Search Google Scholar
    • Export Citation
  • Wolter, K., and M. S. Timlin, 2011: El Niño/Southern Oscillation behaviour since 1871 as diagnosed in an extended multivariate ENSO index (MEI.ext). Int. J. Climatol., 31, 10741087, https://doi.org/10.1002/joc.2336.

    • Search Google Scholar
    • Export Citation
  • Yamanaka, M. D., S.-Y. Ogino, P.-M. Wu, J.-I. Hamada, S. Mori, J. Matsumoto, and F. Syamsudin, 2018: Maritime Continent coastlines controlling Earth’s climate. Prog. Earth Planet. Sci., 5, 21, https://doi.org/10.1186/s40645-018-0174-9.

    • Search Google Scholar
    • Export Citation
  • Yanto, B. Rajagopalan, and E. Zagona, 2016: Space–time variability of Indonesian rainfall at inter-annual and multi-decadal time scales. Climate Dyn., 47, 29752989, https://doi.org/10.1007/s00382-016-3008-8.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Map showing the geographical coordinates of analysis observation stations in Sumatera: (a) 24 observation stations during 1879–1900, (b) 28 observation stations during 1931–60, listed in Pawitan (1990), and (c) 30 observation stations operated by BMKG during 1971–2014. See section 2 for the selection methods of analysis stations in each period. Open circles, filled triangles, and crosses indicate stations in the western coast, eastern coast, and east of the Barisan ridge, respectively. The red dotted line shows the Barisan ridge from northwest to southeast Sumatera.

  • Fig. 2.

    The annual precipitation of (left) the 24 colonial stations (1879–1900) and right) the 28 stations (1931–60) and the 30 BMKG stations (1971–2014) in red and black shapes, respectively. Open circles, filled triangles, and crosses indicate stations on the western coast, eastern coast, and east of the Barisan ridge, respectively. The horizontal axis indicates (a),(d) the longitudes; (b),(e) latitudes; and (c),(f) the distance from the nearest coastline (km). In (c) and (f), crosses show the distance from the eastern coast and not from the western coast.

  • Fig. 3.

    Standardized monthly averaged precipitation in Sumatera. The red solid line, black solid line, and black dotted line indicate seasonal cycle in colonial stations (1879–1900), Pawitan’s stations (1931–60), and BMKG stations (1971–2014), respectively. See Table 5 for the nearest stations to colonial stations in 1931–60 and 1971–2014. The plots indicate (a) 9 western coast stations, (b) 5 stations in the east of the Barisan ridge, and (c) 10 eastern coast stations.

  • Fig. 4.

    The seasonal cycle in 24 stations indicating monthly precipitation anomalies from 1879 to 1900. The horizontal axes indicate January–December. Station numbers and geographical coordinates are the same as those provided in Table 1. The solid line shows monthly precipitation anomalies to see the seasonal cycle. The plots indicate (a) 9 western coast stations, (b) 5 stations in the east of the Barisan ridge, and (c) 10 eastern coast stations.

  • Fig. 5.

    Anomaly in the annual average precipitation of three regions in Sumatera for (a) 1879–1900 and (b) 1971–2014. The black solid line, black dotted line, and black doubled line indicate the western coast, east of the Barisan ridge, and eastern coast, respectively. The solid red line and blue lines indicate the dipole mode index (DMI) and multivariate ENSO index, respectively. Gray shades show El Niño event years.

  • Fig. 6.

    Standardized annual precipitation in 1879–1900. Station numbers and geographical coordinates are the same as given in Table 1. The plots indicate (a) 9 western coast stations, (b) 5 stations in the east of the Barisan ridge, and (c) 10 eastern coast stations. Vertical gray shades indicate the 1888/89 El Niño event year.

  • Fig. 7.

    Standardized annual dry-season precipitation (May–October) at six Southern Hemispheric climate-type stations from 1879 to 1900. The red solid line and blue dotted line indicate DMI and extended multivariate ENSO index (MEI.ext), respectively. The black solid bar, gray solid bar, and open bar indicate a western coast station [Bengkulu (3.8°S, 102.3°E)], an eastern coast station [Muntok (2.1°S, 105.2°E)], and four stations east of the Barisan ridge [Pagar Alam (4.1°S, 103.4°E), Lahat (3.8°S, 103.5°E), Tebing Tinggi (3.6°S, 103.1°E), and Bandar Lampung (5.5°S, 105.3°E)], respectively.

All Time Past Year Past 30 Days
Abstract Views 3 0 0
Full Text Views 1113 319 19
PDF Downloads 941 230 16