Daria Gladskikh: Mathematical Modeling of Lake and Reservoir Thermohydrodynamics and Biogeochemistry Using the LIMNIS Computational Technology

Daria Gladskikh’s talk presented LIMNIS, a computational technology for investigating the thermohydrodynamics and biogeochemistry of lakes and reservoirs. LIMNIS combines full three-dimensional models with reduced one-dimensional and “one-plus-dimensional” configurations in a single software system. Its coupled components describe water motion and temperature, turbulent mixing, ice and snow cover, and the transport and transformation of methane, oxygen, inorganic and organic carbon, and phytoplankton. The coupling accounts, for example, for the effect of turbulence on gas exchange and the effect of chlorophyll on the absorption of solar radiation.

A central question was how model detail should be matched to the properties of a particular water body. Using the Rybinsk Reservoir as an example, the speaker showed that even a relatively shallow and frequently mixed reservoir requires a three-dimensional configuration to represent horizontal variability, seiches, and complex bottom topography. Flooded river channels form deep areas that retain heat longer during autumn cooling, affecting the simulated production and distribution of methane. A one-dimensional model that omitted these processes produced substantial errors in temperature and biogeochemical characteristics.

The second example concerned shallow polymictic thermokarst lakes near the settlement of Tazovsky. For these lakes, the three-dimensional model reproduced the diurnal temperature cycle and observed methane fluxes, after which a one-dimensional prototype was tested. The reduced setup captured the main characteristics of these water bodies but required a correction for methane entering the water from sloping lake beds. Planned LIMNIS developments include a more physically based treatment of sediment sources, parameterization of seiche effects, simulations of water-body responses to climate change, and computationally efficient parameterizations for Earth-system models.

During the discussion, a participant highlighted the strong spatial heterogeneity of methane concentrations and fluxes in reservoirs and the limited knowledge of bottom sediments. Daria clarified that no direct methane measurements were available to validate the Rybinsk Reservoir calculation; the model was compared with oxygen observations instead. The presented methane estimates should therefore be treated as model results that require further validation with dedicated observations.