The 485 m deep Lake Kivu (Rwanda, DR Congo) is a fascinating and unique aquatic system: The water contains ~60 km3 methane and ~300 km3 carbon dioxide and the water column is strongly density-stratified. The aim of this project is to understand the evolution and the maintenance of this unusual natural stratification of the lake. Due to sub-aquatic inputs of salty and carbon dioxide-rich water, Lake Kivu is strongly and permanently stratified. This stratification is unique, as the lake water contains four components affecting water density. Whereas salt and carbon dioxide increase the density in deeper layers, the temperature and methane reduce density in the layers of greater depth. This close competition leads to a so-called double-diffusive layering regime, which causes an astonishing staircase of about 300 well-mixed layers (of ~m large scale), separated by thin but very stable interfaces (of ~dm large scale).

Lay summary

(1) Goal: The aim of this project (continuation) is to show whethernumerical simulations are capable to reproduce the double-diffusive structuresfound in Lake Kivu (see abstract). If successful, simulations could provide adetailed description of the ongoing physical processes. Furthermore we want toexplain the interaction between the small-scale vertical structures with basin-scaleprocesses to understand the importance of double diffusion for the lake systemand its methane reservoir. The project will make a significant contribution to researchinto double-diffusive regimes as well as to an optimized planning of themethane extraction.

(2) Achievement: We have showed that Direct Numerical Simulations in two dimensions are capable toreproduce interface thicknesses in Lake Kivu and that molecular transportthrough undisturbed interfaces captures the total transport as long as the interfacestability is above a critical threshold. This requires profiles pf highresolution through the interfacial structures which are not possible usingstandard CTDs. Using specialized microstructure instruments, we testedlaboratory-based laws of the fluxes through the interfaces and suggested amodification of the most commonly used parameterization.

(3) Importance: On one hand, quiescent Lake Kivu can be seen asnatural laboratory to test flux laws which are relevant globally, such as inthe Arctic Ocean to quantify the heat transport to the overlying ice, whichhave so far been based on small-scale laboratory experiments. On the otherhand, the results are also of economic interest: To avoid building-up of therisk of a lake gas eruption, the two governments of the riparian countries havedecided to use the methane, worth more than 20 billion US dollars. Therefore,it is of great interest to better quantify the methane production and thevertical fluxes of the related water constituents (especially the nutrients).With this research we envisage to integrate new field observations on turbulentand double-diffusive mixing with carbon/methane cycling in Lake Kivu, as investigatedduring the previously funded SNF/SDCproject.

(4) For background information on Lake Kivu


Lake Kivu - turbulence and double diffusion inpermanent stratification

More information at:


or Tobias.Sommer @eawag.ch or alfred.wueest@eawag.ch; Wüest Alfred,Eawag, Kastanienbaum, Switzerland

For an introduction to double diffusion