Field data, scaling analysis, and three-dimensional numerical experiments demonstrate that the dominant large-scale circulation pattern during the southeast trade winds in Lake Tanganyika (East Africa) is a downwind flow in the metalimnion, with a returning southward upwind flow in the upper region of the epilimnion. This is in the opposite direction to all prior literature available on the lake, which has assumed that the southeast trade winds must drive a northward surface flow due to momentum induced by the wind. The gradient in wind speed and humidity along the length of the lake and the resulting gradient in heat exchange with the atmosphere result in warm water accumulating at the northern end of the lake. The north basin of Lake Tanganyika, thus, acts as a heat sink while the south basin is a net source of heat to the atmosphere. The resulting variation in buoyancy flux is strong enough to drive a convective flow in the upper layer in the opposite direction to the wind. A large-scale convective circulation results in an annual exchange of 6.2 × 1012 m3 of water and an annual net transport of heat (averaging 0.2 terawatt) from the northern to the southern region of the lake. Currents in the upper 100-m layer, with a mean flow of 0.2 Sverdrup and flow speed of 0.08 m s−1, can be described by the superposition of a parallel convective flow with the internal seiche motion.

Differential cooling drives large-scale convective circulation in Lake Tanganyika

Spremberg Reservoir is a small and shallow water body where wind sheltering along steep and densely vegetated sections of the shoreline is expected to produce small-scale inhomogeneities of the wind field across the water surface. The potential impacts of these inhomogeneities on the quasi-steady circulation in the main subbasin of the reservoir are investigated in this study by means of a three-dimensional hydrodynamic model. ADCP measurement profiles at three locations are available for comparison. Agreement between characteristic features of the measurements and the simulation data cannot be achieved if uniform winds are applied as boundary conditions, but improves distinctly when sheltering effects are implemented. This is done qualitatively by means of analogy to the separation of flow across a backward facing step. For south-westerly winds, the modification of the flow tends from a three-dimensional to an essentially horizontal, one-cell circulation. The action of the wind sheltering on the flow is found to rely on (1) the addition of a curl to the background wind field and (2) the weakening of the topological moment along the windward shore. In Spremberg Reservoir the first mechanism appears to prevail for south-westerly winds. The study underlines the need for consideration of small-scale inhomogeneities of the wind field induced by sheltering as a potential source of vorticity in hydrodynamic simulations. Further investigations should be directed towards a physically founded parameterization of the actual wind stress distribution and towards quantifying lake dimensions at which such influences on the circulation pattern tend to be negligible.

Measurements and three-dimensional simulations of flow in a shallow reservoir subject to small-scale wind field inhomogeneities induced by sheltering

Examples of numerical modelling of surface water hydrodynamics and water quality are presented. Their meaning for the development of the EIA 3D model system is analyzed from the almost 70 tracer studies (mainly 1971–1974) until the 3-dimensional (3D) solutions which became dominant since 1982. Up to summer 2008, the number of 3D applications has increased to almost 230 while the number of all applications exceeds 300. The specific applications considered are from: Porttipahta reservoir in Northern Finland (I), Porttipahta and Lokka reservoirs (II), combination of reservoirs, lake, river sections, Kemi estuary and the sea (III), Kemi estuary and other coastal applications in the Gulf of Bothnia (IV), Lake Haukivesi in Eastern Finland (V), Lake Nasiselka in South-West Central Finland (VI), and Kymi River and Kotka estuary in Southern Finland (VII). A detailed description of the 3D model system is given in the application of Nasiselka (VI). It is completed with drastically changing water levels, drying and wetting of immersed areas, characteristics of pulsing system, and internal loading in the application of Porttipahta (I). The application of the Kymi River and Kotka estuary (VII) shows the sensitivity of tracers as transport indicators and the validity of the transport model. In Lokka and Porttipahta (II) the methods of validity tests are extended to include field tests, laboratory experiments, and comparisons with expert evaluations and analytical solutions. A strong indication of model validity is obtained in Nasiselka (VI). When the loading after the model work was changed according to a plan included in the computation, the observed changes of water quality corresponded closely with those predicted by the model. Another severe indication of the model validity is from Porttipahta and Lokka (I, II). With the parameter values based on data from 1967–1986, a recent application indicated a close agreement with the observed data from 2000–2006. In Haukivesi (V) and Nasiselka (VI) the agreement between the model and observed results is extended to the biological indicators of the algal biomass. The integrated application to the network of a planned and two existing reservoirs, a strongly regulated lake, river sections, Kemi estuary and the sea (III) shows the usability of the model system to all types of water bodies. As a practical result it highlights the decay of the effects of a new impoundment with time and distance, including fast dilution in the estuary and the sea.

/pdf/mathematical-modelling-of-flow-and-transport-as-link-to-4lglk4ws3w.pdf

Mathematical modelling of flow and transport as link to impacts in multidiscipline environments

Weak population structure and recent demographic expansion of the monogenean parasite Kapentagyrus spp. infecting clupeid fishes of Lake Tanganyika, East Africa.

Seepage rate of hydrothermally generated petroleum in East African Rift lakes: An example from Lake Tanganyika

A three-dimensional numerical circulation model of Lake Tanganyika, developed within the Lake Tanganyika Research Project (LTR-FAO/FINNIDA) is described. The model is based on full non-linear hydrodynamics equations, modified according to hydrostatic and Boussinesq approximations. A space-splitting scheme is used for numerical approximation of governing equations. Numerical experiments with different wind loading revealed the rather complicated flow structure in Lake Tanganyika. Data for model validation have been collected since May 1993. Main hydrological and meteorological parameters such as water level, wind speed and direction, air temperature, humidity and water temperature are recorded with automatic devices. A simple Lagrangian approach with floating buoys and GPS system is used for current measurements. The obtained surface velocities in different locations lake-wide were compared with the calculated surface currents. Agreement between measured and simulated velocities was satisfactory. Accuracy of the simulations depends on the accuracy of the wind forcing reconstruction. Regular strong winds are observed over Lake Tanganyika during the dry season, a period of trade winds that starts at the end of May and lasts until late August or early September. Numerical experiments with a mesoscale meteorological model (Savijarvi, H., 1995. Sea breeze effects on large-scale atmospheric flow. Contr. Atmos. Phys. 68 335–344) showed that the lake–land breeze system is intensified during this period. The collected data and the numerical experiments with the circulation model provide a solid base for understanding the main hydrophysical processes on a lake-wide scale. Important knowledge has also been obtained on the spatial and temporal dynamics of currents.

T Huttula

Papers

Numerical simulation of wind-driven circulation in Lake Tanganyika