190
A. Accornero et al.
duction rate of settling particle is not constant throughout the year. We chose to
explore both the above assumptions (i.e. constant and seasonal cyclic particle
production) due to the fact that particles involved in the vertical flux are not only
the resuit of primary and secondary production processes, but also can be produced from transformation of suspended fine particles through mechanisms of
aggregation, adsorption, microbial colonization, ingestion and egestion and loss
of buoyancy, and in this case the total production rate and the seasonal trend are
impossible to estimate. In both cases the response of the numerical transmissometer was similar to the real one. In Fig. 8 the évolution of the numerical transmissometer record for the second run is shown. Even if a strong decrease occurs
in particle production, transmissometer signal shows a behaviour that is similar
to the real one, shown in Fig. 7. In the run, assuming constant particle production,
the évolution of the 423-m-trap simulated behaviour (Fig. 9a) exhibits a very high
variable signal after Julian day 120, which means thath the probability of a particle settling in the trap is highly variable. It is évident that the huge probability of
quick displacement of the isopycnal surfaces do not foster particle déposition.
This effect might, moreover, be enhanced by the increased horizontal velocity that
is not considered in this simple simulation. Figure 9b shows the évolution of the
numerical sédiment trap flux relative to the run with cyclic production of particles. In this case a very good agreement with the real data is achieved. Almost no
sedimented particles are shown in the trap, even if the transmissometer signal
shows the presence of some particles.
5 Summary and Conclusions
Owing to the limited equipment available and to problems in the functioning of
some of the mooring array instruments this experiment must be considered as
preliminary. Its results, however, allow us to identify some features that characterize the investigated area and to make some considérations that can constitute
a good starting point for future studies:
1. The existence of a characteristic hydrodynamic behaviour along the RIS, modulated by an inflow-outflow intermittence well represented by the 5.02-day
periodicity shown by the V-velocity trend (FFT analysis); this 5.02-day periodicity is, moreover, found in ail the measured physical variables at every depth.
2. The superimposition of short-term (2-4 days) puise events resulting in the outflowing of supercold water from below the RIS towards the Ross Sea and sometimes associated with recirculation processes.
3. The low particle downward fluxes and the major rôle of physical forcing in
affecting sédimentation processes at the investigated site.
Fluxes of sedimenting particles are undoubtedly influenced by ecological factors, such as the seasonal cycles of primary production and related secondary
production, strictly dépendent on environmental conditions such as seasonal patterns of solar irradiance, ice extent, type and thickness, water column stability
and meteorological conditions [31, 33, 49-56]. In the first period, from the starting of the experiment through April, moderated current velocity (< 20 cm s1) and
limited vertical displacement of the isopycnal surfaces (< 20 m day'1) do not seem
A. Accornero et al.
duction rate of settling particle is not constant throughout the year. We chose to
explore both the above assumptions (i.e. constant and seasonal cyclic particle
production) due to the fact that particles involved in the vertical flux are not only
the resuit of primary and secondary production processes, but also can be produced from transformation of suspended fine particles through mechanisms of
aggregation, adsorption, microbial colonization, ingestion and egestion and loss
of buoyancy, and in this case the total production rate and the seasonal trend are
impossible to estimate. In both cases the response of the numerical transmissometer was similar to the real one. In Fig. 8 the évolution of the numerical transmissometer record for the second run is shown. Even if a strong decrease occurs
in particle production, transmissometer signal shows a behaviour that is similar
to the real one, shown in Fig. 7. In the run, assuming constant particle production,
the évolution of the 423-m-trap simulated behaviour (Fig. 9a) exhibits a very high
variable signal after Julian day 120, which means thath the probability of a particle settling in the trap is highly variable. It is évident that the huge probability of
quick displacement of the isopycnal surfaces do not foster particle déposition.
This effect might, moreover, be enhanced by the increased horizontal velocity that
is not considered in this simple simulation. Figure 9b shows the évolution of the
numerical sédiment trap flux relative to the run with cyclic production of particles. In this case a very good agreement with the real data is achieved. Almost no
sedimented particles are shown in the trap, even if the transmissometer signal
shows the presence of some particles.
5 Summary and Conclusions
Owing to the limited equipment available and to problems in the functioning of
some of the mooring array instruments this experiment must be considered as
preliminary. Its results, however, allow us to identify some features that characterize the investigated area and to make some considérations that can constitute
a good starting point for future studies:
1. The existence of a characteristic hydrodynamic behaviour along the RIS, modulated by an inflow-outflow intermittence well represented by the 5.02-day
periodicity shown by the V-velocity trend (FFT analysis); this 5.02-day periodicity is, moreover, found in ail the measured physical variables at every depth.
2. The superimposition of short-term (2-4 days) puise events resulting in the outflowing of supercold water from below the RIS towards the Ross Sea and sometimes associated with recirculation processes.
3. The low particle downward fluxes and the major rôle of physical forcing in
affecting sédimentation processes at the investigated site.
Fluxes of sedimenting particles are undoubtedly influenced by ecological factors, such as the seasonal cycles of primary production and related secondary
production, strictly dépendent on environmental conditions such as seasonal patterns of solar irradiance, ice extent, type and thickness, water column stability
and meteorological conditions [31, 33, 49-56]. In the first period, from the starting of the experiment through April, moderated current velocity (< 20 cm s1) and
limited vertical displacement of the isopycnal surfaces (< 20 m day'1) do not seem