Geoscience Reference
In-Depth Information
adopted. A particular problem can be due to the fact that surface tension of water
causes the tracer particles to coalesce, with the formation of groups. In fact, if two
particles are attracted, each of them has a residual velocity that is inconsistent with
the true flow velocity, which can produce a significant source of error. Because of
the surface tension, tracer particles can also adhere to solid boundaries and rough-
ness elements that protrude from the flow. The simplest way for reducing this
phenomenon is the use of smaller tracer concentrations, but other methods are also
available (Admiral et al.
2004
). Moreover a good contrast between the tracers and
the background color is desirable to easily analyze particle images. Whenever
possible, like in laboratory investigations, the model or the water can be colored.
In this way, with an appropriate threshold, background noise can be filtered in a
preprocessing phase. Moreover, tracers need to be cheap and environmentally
acceptable. Investigations on the effect of tracer material, seeding procedure, and
concentration are reported by Muste et al. (
2004
).
Field measurements are made with natural light, while for indoor recordings
strong illumination is required, usually realized with halogen, sodium-vapor, or
other kinds of lamps. In this case, the lamps are to be strategically positioned in
order to have uniform illumination of the area and to avoid light reflection on the
free surface. Since in LSPIV the illumination is much less compared with the high
intensity laser sheet of a conventional PIV, a preprocessing phase with image
enhancement procedures is often required. Also in this case the use and the effects
of such procedures should be carefully considered.
More than one camera can be used for image recording, if the size of the
investigation area is large. In this case, the velocity fields obtained with different
cameras have to be assembled in a postprocessing phase.
The large distance and sometime the oblique angle between the camera and the
image plane can introduce lens and geometrical distortions, which should be
removed before the processing phase. Since the lens distortion effects are usually
negligible, simple correction procedures, based on a geometrical transformation
applied to recorded images, can be used (Fuijta et al.
1998
). Like for standard PIV,
cross-correlation algorithms can be used to obtain instantaneous velocities and a
post-processing phase for discarding erroneous vectors can be performed.
Given the large size of LSPIV imaged area and consequently of the interrogation
areas, the spatial resolution of velocity measurements is limited. The flow scales
smaller than the interrogation area cannot be solved, so LSPIV cannot measure
small-scale turbulence, but it is intended to capture large-scale flow structures.
LSPIV was used to measure the surface velocity distribution in entire rivers
(Fuijta et al.
1998
), also during extreme events, using a mobile LSPIV unit (Kim
et al.
2008
), able to rapidly develop the measurement equipment in a desired site, or
using video images from helicopters (Fuijta and Hino
2007
b). In fact, the knowl-
edge of surface velocities is useful for many purposes in fluvial hydraulics. They
furnish important information in shallow water flows, for example, they were used
for investigating on large coherent structures (Weibrecht et al.
2002
) and mixing
processes between a river and dead-water zones created by groin fields (Weibrecht
et al.
2008
). It is even possible to use surface velocities to estimate indirect
