Whether by chronic or
accidental releases, the impact of a nuclear installation on the environment
mainly depends on atmospheric transfers; and as the accidents at Chernobyl and
Fukushima show, affect the contamination of surfaces and impacts in the medium
and long-term on the environment and the population. In this context, this work
focuses on the characterization and modeling of dry deposition of submicronic
aerosols on liquid surfaces in motion such as rivers.
Unlike wet deposition
which is conditioned by washout and rainout (rain and clouds), dry deposition
is a phenomenon that depends entirely on the characteristics of aerosols,
receiving surfaces, and air flow. In practice, the evaluation of dry deposition
is based on the estimation of flux modeling as the product of particle
concentration and deposition velocity which can vary over several orders of
magnitude depending on the receiving surfaces (forest, snow, urban,
grassland…). This topic is motivated by the virtual non-existence of studies on
the mechanisms of dry deposition on continental water systems such as rivers;
and respect for submicronic aerosols. They have the lowest deposition
efficiencies and filtration and the longer residence time in the atmosphere. In
addition, they are potentially the most dangerous to living beings because they
can penetrate deeper into the airway. Due to the lack of data on the dry
deposition of submicronic aerosols on a liquid surface in motion, the approach
was based on two axes: 1) the acquisition of experimental deposition velocities
and 2) the analysis and interpretation of results through modeling.
The experiments were
performed with uranine aerosols released into the IOA wind tunnel (Interface
Ocean Atmosphere) of the Institute for Research on Non Equilibrium Phenomena
which is configured to study the coupling between the air flow and water. These
experiments have given many dry deposition velocities for different
configurations characterized according to wind conditions (central wind speed:
1 , 2, 4 , 5, 7.5 and 9.5 m/s), current (co - current and counter-current,
water flow velocity: 0 , 6 and 12 cm/s ), ambient (temperature and relative
humidity of the air and water temperature), the liquid surface deformations (measured
significant wave height) and size distribution of aerosols released.
The modeling part was
to adapt the model to resistance. Slinn and Slinn (1980). This model is based
on the assumption of conservation of vertical flow in the boundary layer. This
one is divided into two layers: a very thin deposition layer near the surface
which is water-saturated and a transfer layer located above. The transfer layer
provides the particle deposition layer by turbulent diffusion and
sedimentation. In the deposition layer the particles are deposited under the
effect of several mechanisms such as Brownian diffusion, sedimentation,
impaction and phoretic mechanisms. The main adjustments made by this work have
been to take specific account of the different classes of particle size
distribution, the spectrum variation as a function of hygroscopicity, and
mechanisms of aggregation. It is integrated mechanisms of diffusiophoresis and
thermophoresis, respectively produced by the evaporation of water and the
temperature gradient at the air-water interface.
To account for
hygroscopic uranine aerosols, the deposition rates are analyzed in terms of
humidity and rescaled by the friction velocity. In all cases, the deposition
rates rescaled by the friction velocity range from 10-3 to a wind speed of 1 m/s to 10-5 - 10-4 for wind speeds above 5
m/s . It is shown that the changing speed rescaled is inversely proportional to
the wind velocity when the water surface is smooth (between 1 and 5 m/s ) and
becomes proportional to the deformation when the surface becomes significant
(over 5 m/s ) . Although the results do not clearly identify the effect of the
current on the deposition velocity, the modeling shows that turbulent diffusion
is dominant in the transfer layer and the effect of hygroscopic mechanisms of
sedimentation is insignificant. In the deposition layer, the effects are more
hygroscopic and negligible deposition of finer particles can be blocked by the
phoretic mechanisms when the difference between the temperatures of air and water
increases. In all cases, the disparities between the model and experiment can
be reconciled under conditions that properly account for the size distribution
of aerosols.
