Three analytical diffusion models are used to describe the distribution of a nonconservative tracer in the hypolimnion of a lake: The one‐dimensional vertical (1‐DV) model, developed for tracers in the deep sea, leads to a significant overestimation of vertical diffusivity Kz in lakes; the one‐dimensional topographic vertical (1‐DTV) model, applicable if horizontal mixing is fast compared to in situ decay, demonstrates that, except for the deepest layers, vertical tracer distributions are insensitive to Kz; the one‐dimensional topographic horizontal (1‐DTH) model, in which horizontal mixing is incomplete and vertical diffusion is disregarded, is the appropriate model for the interpretation of vertical excess 222Rn profiles a few meters above the lake bottom. The radon flux F from the bottom is calculated from 226Ra in the sediments and corrected for depth variation of porosity and radium activity. In Baldeggersee (Switzerland), F is 390 ± 70 dpm·m‒2·d‒1. Radium activities in the sediments are 0.7 ± 0.2 dpm·g‒1 dry mass, similar to those in other Swiss lakes and in coastal areas of the ocean but 20–50 times smaller than in the deep Pacific. Horizontal diffusivity in the hypolimnion decreases from June (5,000 m2·d‒1) to October (1,000 m2·d‒1); vertical diffusivity is between 1 and 10 m2·d‒1, roughly 10 times larger than values calculated from temperature in the upper hypolimnion. Vertical buoyancy flux is between 10‒11 and 10‒10 m2·s‒3, i.e. between values for the deep ocean and the small ELA lakes in Canada. The large Kz and the transient, often irregular shape of the radon profiles are the result of horizontal bottom currents previously measured by other methods.
Read full abstract