The Source of the Mekong River, Qinghai, China


Photo from Tokyo University of Agriculture Archives, 1994.


By international geographic convention, the geograpic source of a river is the spring or glacial discharge within the drainage basin which is furthest from the mouth of the river.

The geographic source of the Mekong River on the Tibetan Plateau in Qinghai, China was formally located at Longitude 94 41 44E, Latitude 33 42 31N, elevation 5224 m (17,135 ft) in 1999 by the Chinese Academy of Sciences. However, in October, 2002, a different location was announced by another division of the Academy, and in 2007, another location, between the 1994 and 1999 locations and very near to the 1994 location, was identified by the China Exploration and Research Society. The 1999 and 2002 locations were at the base of retreating glaciers and thus transient. The 2007 location is at the drainage divide and probably the most accurate.

The glacier pictured above is on the north side of Guosangmucha Mountain, elevation 5514 m (18,086). The glacial stream is called Lasagongma, which flows into the Gaoshanxigu, then the Guoyonggu, then the Zayaqu, then the Zaqu, which becomes the Lancang Jiang at Qamdo on the Sichuan-Tibet border. At the border between China and Myanmar (formerly Burma), the Lancang is called the Mekong River, which then flows through Laos, Thailand, Cambodia and Vietnam. It flows into the South China Sea near Ho Chi Min City (Saigon), about 4500 km from its source.

A Summary of Attempts to Locate the Source of the Mekong

The geographic sources of most major rivers on earth were identified in the late 1800's or early 1900's, including the six large rivers draining the Tibetan Plateau: Indus, Yellow, Yangtze, Mekong, Salween and Bramaputra. However, errors made in the location of the geographic source of the Mekong were not corrected until 1999, and the controversy continues.

This summary is based on reports published by:

Michel Peissel in the Geographic Journal of the Royal Geographic Society (RGS) of Britain (Geographical Journal, Vol. 161, Part 2, July 1995)

Jin Chanxing and Zhou Changjin of the Commission for Integrated Survey of Natural Resources of the Chinese Academy of Sciences (CISNR,1995)

Guan Zhihua, Zhou Changjin and Tao Baoxiang (CISNR,1999)

Zhou Changjin and Guan Zhihua of the Institute of Geographic Science and Natural Resources Research of the Chinese Academy of Sciences (IGSNRR, 2001)

Masayuki Kitamura of Toykyo University of Agriculture (TUA, 2001)

Tamotsu Nakamura of the Japanese Alpine Club (JAC, 2001)

Zhou Changjin, (Zhou, 2001)

Correspondence submitted to the Geographic Journal by Pete Winn of Earth Science Expeditions (ESE) and Tamostu Nakamura of JAC (GJ_ESE_JAC, 2002)

Liu Shaochuang, Institute of Remote Sensing Application of the Chinese Academy of Sciences (IRSA, 2002)

How Man Wang, China Exploration and Research Society (CERS, 2007)

Links to references in parentheses are included at the end of this summary.

See maps for the locations of referred to in this summary.

The Chinese reports were translated into English by Tamotsu Nakamua and provided as attachments to his JAC, 2001 report. Mr. Nakamura is a Director of the Japanese Alpine Club and editor of the Japanese Alpine News. He is the author of "Deep Gorge Country - Journeys to Blank Areas on Maps in the Eastern Himalayas," Parts I and II, and has extensive experience climbing in the Himalaya Mountains.

Tibetans living in the Mekong headwaters believe there are several spiritual sources to the Zaqu: Lungmug Spring (drains into the Zanaqu), Zanarigen Mountain (drainage into both the Zanaqu and Yangzte), and Zaxiqiwa Lake (drains into the Zayaqu). None of these are the geographic source. In 1894, Dutreuil de Rhins and Fernand Grenard of France were the first Europeans to claim they had discovered the geographic source. They entered the Mekong drainage from the Yangtze headwaters at the pass above Lungmug Spring and declared it to be the source. This spring discharges into Zanaqu, a tributary to the Zaqu which drains from the west.

In 1994, the Frenchman Michel Peissel visited the springs below Rupsa Pass in the upper reaches of Zanaqu and claimed that they were the geographic source. These springs are further upstream than Lungmug Spring. Peissel used 1:500,000 scale 1987 jet navigation maps which implied the Zanaqu was longer, but he did not attempt to measure the relative lengths of the Zanaqu and Zayaqu using Russian 1:200,000 or Chinese 1:100,000 scale maps or satellite images of the area that were available at the time. He persuaded the RGS to publish a record that he had discovered the geographic source located at 93 52E, 33 16N and published a book describing his expedition (The Discovery of the Source of the Mekong, 1995) and eventually a video with the same title.

At the same time Peissel visited Rupsa Pass, a TUA-CISNR team travelled to Guosangmucha Mountain near the northeastern reaches of the Zayaqu. They chose this region for the source based on Chinese 1:100,000 scale maps. The TUA team was led by Masayuki Kitamura of the Exploration Club, and Jin Chiangxing and Zhou Changlin represented CISNR. Kitamura recorded the source as 94 41 37E, 33 42 41N, the point where Lasagongma Creek emerges from a glacier at the base of Guosangmucha Mountain at 5160 meters (Japanese Alpine News, Vol. 1, 2001).

In March, 1995, Jin Chiangxing and Zhou Changlin of CISNR published the source as 94 41 35 E, 33 44 13 N at 5167 meters (Geographical Research, Vol. 14, No. 1). This is about 3 km downstream from the TAU location. At the time, Kitamura could not find out why the Chinese published this location rather than the location identified on their joint expedition. In a 2001 email message Zhou Changjin explained that CISNR was not sure the TUA location was correct, and instead published the location of the confluence of Lasagongma Creek and another creek. Based on further research in 1999, CISNR published a revised source location that is nearly identical with the TUA location (Geography, October 1999).

In 1999, four teams visited the Mekong source area: two Chinese teams, one Japanese-Chinese team and one American-Chinese team. The two Chinese teams arrived in July, the other two teams arrived in August.

The Dexiang team, representing CISNR, went to Lasagongma in July and reported that the geographic source was located at 94 41 44E, 33 42 31N, elevation 5224 meters. This is only 360 meters upstream and 57 meters higher than the source identified by TUA in 1994. Between improvements in GPS technology and the likelihood of glacial retreat, the 1994 TUA-CISNR team was probably the first non-Tibetan group to visit the source.

The Liu Shaochuang team, representing the Remote Sensing Office of the Chinese Academy of Sciences (RSO), went to the base of a glacier on Jifu Mountain in July. They claimed the geographic source was the point at which Gaodepu Creek discharged from the glacier - 94 41 12E, 33 45 35N, elevation 5552 meters - about 6 km north of the source of Lasagongma Creek. Lasagongma Creek drains into Gaoshanxigu Creek, which becomes Guoyonggu Creek at the confluence (Yeyonsongdou) with Gaodepu Creek .

Based on GPS readings and detailed measurements of stream lengths from maps and satellite photos, each team decided that the length of the tribuary they had chosen was the longest. The CISNR believed Gaoshanxigu Creek was about one km longer than Gaodepu Creek, and RSO believed Gaodepu Creek was about two km longer than Gaoshanxigu Creek. Although CISNR is the official survey agency for the Chinese government and could have insisted on their location, they agreed to settle the difference by determining which creek had the greatest flow.

In August, Zhou Changjin of CISNR, who had been on the TUA-CISNR expedition in 1994 and the CISNR expedition in July, traveled to the upper Zayaqu with ESE, an American whitewater rafting group. They arrived in the area one day later than a second TUA expedition, again led by Masayuki Kitamura. The TUA team split into two groups, one of which returned to Lasagongma to revisit the location of the source they had identified in 1994, while the other group began a first descent of the Zayaqu.

While ESE readied its boats for their "almost" first descent, Zhou returned to Yeyonsongduo with a flow meter and measured the discharge of Gaodepu and Gaoshanxigu creeks. The discharge of Gaoshanxiqu Creek was about 20 percent higher than the flow in Gaodepu Creek, so in their report, Langcang Jiang: Identification of the True Source and Headwaters, published in Geography, October 1999, Zhou and Zhihua of CISNR recognized 94 41 44E, 33 42 31N as the official location of the geographic source of the Mekong River. In 2001, representing IGSNRR, these same authors published another report reconfirming the identification of the true source (Geographical Research, Vol. 20, No. 2).

However, in October, 2002, Liu Shaochuang, representing the Institute of Remote Sensing Application of CAS, announced a new location at 94 40 52E and 33 45 48N, very near his 1999 location. The discrepancy between the locations of the two sources was not resolved until 2007.

The TUA rafting team floated 560 km down the river to Qamdo, on the border between Tibet and Sichuan. The ESE team floated about 160 km to a large tributary 30 km below Zadoi (the first town on the Mekong). See the link below for a description of the ESE expedition (ESE, 1999). Masayuki Kitamura published three articles on the TUA expedition: Gakujin, No. 632, 2000, Mountain and Valley, No. 775, 2000, and Japanese Alpine News, Vol. 1, 2001 (TUA, 2001).

Tamotsu Nakamura, a Director of the Japanese Alpine Club and editor of the Japanese Alpine News, translated the CISRN and ISGNRR reports into English, prepared a history of the search for the source of the Mekong in English (JAC, 2001) and prepared the maps for the TUA, 2001 article. Nakamura and Pete Winn, Science Director of Earth Science Expeditions, prepared a Letter to the Editor of the Geographic Journal of RGS and a report submitted to the Himalayan Journal summarizing the CISNR data and documenting the fact that the glacial spring at the head of Lasgongma Creek is the true and official source of the Mekong.

In 2007, the Chinese Exploration and Research Society, using high resolution satellite imagery, identified the source as being the spring on Mt. Jifu discharging into Gaodepu Creek due to the large number of meanders compared to Gaoshanxigu Creek, even though it has a much smaller discharge. However, CERS did not publish a specific location. Again, the true source may vary due to differences in the rate of glacial retreat.

Pieter Neele published an article in the April, 2014 issue of Japanese Alpine News (edited by Tomatsu Nakamura) describing the efforts of he and his co-explorer, Luciano Lepre, to confirm the source of the Mekong in 2013. In the process, they discovered a new source, located between the CISRN and ISGNRR sources but at a higher elevation, also discharging from a glacier: 33 45.677 E, 94 40.562 N, altitude 5,374 m. His article also summarizes the information listed above, and provides maps and photos and a strong rationale for making a claim to have discovered a source that is further from the mouth than the other sources to date.

Links to references:

CISNR, 1995
CISNR, 1999
ESE, 1999
IGSNRR, 2001
TAU, 2001
JAC, 2001
Zhou, 2001
GJ_ESE_JAC, 2002
IRSA, 2002
CERS 2007 summary
Neele & Lepre 2013, Japanese Alpine News (April 2014, p. 108)
Neele's photos of the 2013 expedition

History of First Descents of the rivers of Tibet and western China

More info about rivers in China