Paleoclimate from Sediment Records

March 2016 – the first coring expedition in Biển Hồ

Our group of 7 (i.e. three professors and one graduate student from Vietnam National University Hanoi; three from Indiana, namely Anrdt, Minh and Jan who is applying for Bremen University) spent three days of hard work in and around Pleiku in the central highland of Vietnam. Dương Nguyễn-Thùy ‘s enthusiasm and insistence made it possible to overcome repeated resistance from local authorities who were suspicious about the involvement of foreigners in work concerning the precious freshwater supply of Pleiku. Equally important, Dương’s former adviser Prof. Phương Hòa Tạ’s excellent connections to some regional officials (in part his former students) provided critical support. Finally we broke the ice and received a research permit. Everywhere we go we are accompanied by a friendly local official to ensure a smooth operation.

Biển Hồ – hồ T’Nưng crater lake (maar) bathymetric map and sampled and trap set locations on March 2016 (Pleiku, Gia Lai).

The sensitive nature of the maar lakes (Biển Hồ – hồ T’Nưng – Đôi mắt Pleiku – Pleiku’s eye or Pleiku’s pearl) for Pleiku’s water supply prohibits most private boating. There is only one skipper authorized to operate his private boat. He is paid by the authorities to occasionally recover the bodies of drowned people. We rented his boat with a crew of 2. Only when we showed up the owner asked for $2000 rental fee. He had refused to spell out a price tag earlier. Luckily Professors Dương and Phương were able to bargain him down a little.

Topmost part of a sediment core from a lake.

There are three interconnected maar lakes (Biển Hồ) about 10 km north of Pleiku. The maximum water depth is about 25 m in the deepest maar. The second maar is about 22 m deep, and the third one is even shallower. The maars have slightly different water temperatures. We were able to run various transects with an echolot.  Hướng used his computer graphing skills to establish a preliminary colorized bathymetric map of all three lakes. There is a strong thermocline. Deeper water is oxygen-depleted. Our echolot “fish finder” did not indicate any fish near the bottom. The sediment is an extremely organic-rich, black sapropel with no trace of hydrogen sulfide. Upon warming of sediment, some small methane bubbles exsolved. Jan used the Gopro camera to take underwater movies during coring. The extremely water-rich sapropel remains very difficult to core with a traditional core catcher down to sediment depths of more than 5 m. It simply oozes back out of the core catcher when you try to recover a core. We estimate the water content at 5 m to 70 %. That bad! These maar sediments are very mineral-starved. Drilling of lateral holes in the upper part of the core liner helped a little, but still we could not recover a core longer than 0.5 m.

Comming soon...
Jan and Hướng with a “perfect” bouy designed for a sediment trap.

Most importantly, there is a bacterial mat covering the sediment/water interface. We could not see laminations below the mat through the plexiglass because the sediment smeared too much during its travel up and down the core liner. The sediment is way too unconsolidated to structurally withstand any transport. We are storing our short core upright in the house of a local fisherman for the next few months. With the help of the same fisherman and his son, we set a sediment trap with 3 collection chambers above a common mooring. We trained the fisherman to service the trap and collect samples for us in a 3-month rhythm.

The water level of the interconnected maars fluctuates by a few meters depending on the severity of the rainy season.

The surrounding crater rims occasionally suffer wildfires. We saw plenty of charcoal and charred plant remains when walking around the maars. Jan and Anrdt are taking a sediment sample from a core catcher (stored inside a ca. 8 mm diameter plastic straw) with us to Indiana where we plan to use a plasma asher to see whether there is any rhythm in the inorganic content in the black sediment. We can also try to use epoxy impregnation and microscopy. Dương also has a short straw from the same core catcher and will do some mineralogical work in comparison with a basalt weathering profile that she sampled along a crater slope.

In the remaining days we are planning to lay the groundwork for a Vietnamese research proposal (officially spearheaded by Hướng) to attract funds for repeat coring and bolstering the Vietnamese efforts towards paleoclimatic maar research.

Working here is not without risks. The rented boat was not really safe. When carrying an ‘anchor stone’ from the shore onto the rickety ramp of the boat, Jan slipped, fell, and cut his chin on sharp metal. Dương inquired where a cosmetic surgeon was available. Jan was brought to a local Vietnamese Army Hospital (open to the public) and received 6 expert stitches at a very reasonable cost of only $35 (including medication). The same evening Anrdt suddenly felt extremely sick with tummy cramps. Apparently the dinner did not agree with his digestive system. He have not been in such pain for more than 30 years! Luckily, after a few agonizing hours all became normal again and Anrdt had a good sleep.

Mud coring from a lake
A sediment core has been recovered from a lake.
December 2016-January 2017 – the second coring expedition in Biển Hồ

We cored maar sediments near Pleiku, Vietnam. Our newly designed “autonomous sediment coring device with a pneumatically operated core catcher” worked well in the ca. 20-m deep maar lake. An alternate design using an extended set of metal rods pushing a coring device into the soft sediment also shows promise. We have half a dozen sediment cores that are about 50 to 70 cm in length. We are optimistic that a combination of both coring methods will yield several meters of sediment from the maar lake when we will return next time.

This time there was a distinct sediment layer on the surface that was fluffy and reddish-brown in color, very similar to the the clay that has been washing into the lake during the recent rainy season. We are now at the tail end of the wet season. We witnessed deep gullies that had been recently eroded along points of seasonal inflow along the perimeter of the maar lake. Jan often deployed the GoPro underwater-camera during coring. We never saw any fish, although we saw hobby fishermen with fishing poles catching some fish from shallow waters. The sediment cores often had quite clear water above the sediment, so we could very well observe the sediment/water interface. We never saw anything moving in the water and in the sediment. No worms, no nematodes.

As the GoPro descended, there was abundant light available for some time until there was a rapid transition into darkness before hitting the sediment surface. Arndt interpret this as a deep nepheloid layer that hasn’t had enough time to settle out. In contrast, in March 2016 we were able to see the sediment surface with the GoPro very well during coring. March is well into the dry season when there hasn’t been any significant for some time. Also, The March water level in the maar lake was significantly below the current water level.

Jan was also able to take a few short cores from ‘dry’ maars in the area, which are actually swampy at this time. One maar had a cement-like, very hard, dark layer at a depth of less than a meter below soft sediment. We suspect this to be a diagenetically hardenend tephra layer. Dương has a sample and will find out using the mineralogical composition. A tephra layer should have a very low TOC content.  Jan could not penetrate the layer. In another swampy maar he was attacked by leeches and donated plenty of blood. Our next sampling trip to Pleiku will need to improve our core recovery from dry maar lakes. Maybe we really need vibracoring equipment.

All cores are now in Hanoi in the dark at room temperature in their core liners.

Sediment collected over several weeks in sediment traps.
Dry residue from the contents of a lacustrine sediment trap.
November 2017 – the third coring expedition in Biển Hồ

We returned to Biển Hồ maar lake in November 2017 with with new home-built platform and sediment coring devices. We were delighted to welcome Dr. Antti Ojala from the Geological Survey of Finland as a new member to our team. Antti’s expertise in lake coring and core-to-core correlation using paleomagnetic sediment properties proved invaluable. Two days of sediment coring in Biển Hồ in November 2017 yielded 17 wonderful piston and gravity cores from up to 21 m water depth with up to 3.5 meter penetration into sediment. Our mobile coring platform and the newly designed coring devices performed better than expected.
In addition to coring in Biển Hồ maar lake, we also considered some shallower lakes and dry maars which can record bio-geochemical shifts relating to paleoenvironmental changes. We took short cores in Ia Bang lake and in the Plei Ốp swampy maar near Pleiku City.
The logistics worked out better than during previous coring expeditions, because our experienced team members know their duties and don’t require orders from the “boss”. We worked like a well-tuned professional coring team! See detail of our third coring expedition in Biển Hồ maar lake and core documentation and analyses in the laboratory at VNU in the session “I cored”.

Team members carry our coring platform from an assembly point near the shore to Biển Hồ maar lake in November 2017.

Related pubblications

  • Nguyễn-Văn, Hướng, Dương Nguyễn-Thùy, Jan P. Schimmelmann, Bernd Zolitschka, Thân Tạ-Văn, Nguyệt Nguyễn-Ánh, Phương Tạ Hòa, Dương Nguyễn-Thùy, Thắng Lê-Quyết, Quỳnh Nhi Phạm-Nữ, Vũ Huỳnh-Kim and Arndt Schimmelmann (2017) Exploring the paleoenvironmental potential of laminated maar sediment in central Vietnam: An archive of regional paleo-flooding? PAGES Zaragoza 2017 5th Open Science Meeting ”Global challenges for our common future – a paleoscience perspective”, 9-13 May, 2017, Zaragoza, Spain, Session #16: Multidisciplinary reconstruction of paleofloods, p. 325 in abstract book[Abstract PDF] [Poster JPEG]
A poster at PAGES Zaragoza 2017 5th Open Science Meeting ”Global challenges for our common future – a paleoscience perspective”, 9-13 May, 2017, Zaragoza

Exploring the paleoenvironmental potential of laminated maar sediment in central Vietnam: An archive of regional paleo-flooding?


Global warming increases atmospheric humidity and will likely affect the monsoon system and the impact of typhoons in Vietnam. Lacustrine sediment from Biển Hồ maar lake in central Vietnam is a long-time archive of erosion from a small catchment that had been limited to the interior of the volcanic crater until two dams were built in 1978-1983 that established an additional freshwater reservoir in the Northeast from where turbid and nutrient-rich flood water occasionally enters the maar lake.

Local precipitation history and statistics

The available short instrumental records from Pleiku indicate a seasonal monsoonal rainfall pattern with a peak in August, as well as strong interannual variability causing occasional flooding and droughts.

Preliminary coring of sediment

Short exploratory sediment cores were retrieved from deep and shallow parts. A microbial mat was ‘curling up’ on the sediment surface in March 2016 (top right). The  topmost sediment in January  2017 was a reddish-brown clay flood layer that matched the freshly eroded sediment at the spillover sill in the NE corner of the maar (bottom right). Sediment cores and benthic environment in the deep part of the maar gave no indication for bioturbation.

Post-dam rise of lake level and reforestation

Images from the 1960s show exposed terraces at low water level in the shallow SE part of the maar (e.g., U.S. Army trucks, 1967). The water level rose a few meters after the construction of dams ca. 1980. Reforestation and protection improved the maar’s rim after 1990.

Short cores suggest laminated sediment

The distinct topmost 2017 clay flood layer has a sharp lower boundary. Kaolinite, goethite, and groutite are abundant weathering products of volcanic rock. Minor components are hydrozincite, witherite, and plagioclase. Calcium carbonate is rare or absent. Intervals with coarser sediment may relate to flood years, e.g., the 2009 ‘Ketsana’ tropical cyclone and 1996 flooding in central Vietnam. 210Pb and  137Cs ages are needed to test our hypothesis.

Steel shrapnel at ~13 cm depth may date to the Vietnam war in the early 1970s. A relatively sharp decrease in the abundance of diatoms at ~18 cm depth may derive from dilution by minerals during enhanced erosion.

The top section of a dried and peripherally sliced core documents laminae that are partially compromised due to disturbance near the core liner. The 2017 clay layer has a distinct color and sharp boundary in an anoxically stored, wet core top.

Preliminary pollen data at 2-cm resolution indicate a pine pollen increase (purple) in upper sediment due to reforestation. The appearance of coffee pollen (red) near the surface agrees with the recent spread of coffee plantations nearby.


Sedimentary archives in central Vietnam’s maars likely extend from the Anthropocene deep into the Pleistocene. Maar lakes with suboxic bottom waters may offer well-preserved and high-resolution laminated sediments with paleoenvironmental information, including a record of flood layers. Some sediments may be varved. The anthropogenically influenced last few decades of lake history can provide proof-of-concept when interpreting earlier flood layers of similar composition. Similarly, the recent reforestation efforts around the maar crater’s rim offer a test case for evaluating the sedimentary pollen response to rapid environmental change. In addition to Biển Hồ maar lake, the region offers numerous other maars with potentially valuable sedimentary records.

Geosciences are improving the lives of people in Vietnam