![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
Detection and reconstruction of large scale flow structures in a river by means of empirical mode decomposition combined with Hilbert transform |
VerfasserIn |
Mário J. Franca, Ulrich Lemmin |
Konferenz |
EGU General Assembly 2014
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 16 (2014) |
Datensatznummer |
250093213
|
Publikation (Nr.) |
EGU/EGU2014-7740.pdf |
|
|
|
Zusammenfassung |
The occurrence of large scale flow structures (LSFS) coherently organized throughout the
flow depth has been reported in field and laboratory experiments of flows over gravel beds,
especially under low relative submergence conditions. In these, the instantaneous velocity is
synchronized over the whole vertical profile oscillating at a low frequency above or below the
time-averaged value.
The detection of large scale coherently organized regions in the flow field is often difficult
since it requires detailed simultaneous observations of the flow velocities at several levels.
The present research avoids the detection problem by using an Acoustic Doppler Velocity
Profiler (ADVP), which permits measuring three-dimensional velocities quasi-simultaneously
over the full water column. Empirical mode decomposition (EMD) combined with the
application of the Hilbert transform is then applied to the instantaneous velocity data to detect
and isolate LSFS.
The present research was carried out in a Swiss river with low relative submergence of
2.9, herein defined as h/D50, (where h is the mean flow depth and D50 the bed grain size
diameter for which 50% of the grains have smaller diameters). 3D ADVP instantaneous
velocity measurements were made on a 3x5 rectangular horizontal grid (x-y). Fifteen velocity
profiles were equally spaced in the spanwise direction with a distance of 10 cm, and in the
streamwise direction with a distance of 15 cm. The vertical resolution of the measurements is
roughly 0.5 cm. A measuring grid covering a 3D control volume was defined. The
instantaneous velocity profiles were measured for 3.5 min with a sampling frequency of 26
Hz.
Oscillating LSFS are detected and isolated in the instantaneous velocity signal of the 15
measured profiles. Their 3D cycle geometry is reconstructed and investigated through phase
averaging based on the identification of the instantaneous signal phase (related to the
Hilbert transform) applied to the original raw signal. Results for all the profiles are
consistent and indicate clearly the presence of LSFS throughout the flow depth
with impact on the three components of the velocity profile and on the bed friction
velocity. A high correlation of the movement is found throughout the flow depth, thus
corroborating the hypothesis of large-scale coherent motion evolving over the whole
water depth. These latter are characterized in terms of period, horizontal scale and
geometry.
The high spatial and temporal resolution of our ADVP was crucial for obtaining
comprehensive results on coherent structures dynamics. EMD combined with the Hilbert
transform have previously been successfully applied to geophysical flow studies. Here we
show that this method can also be used for the analysis of river dynamics. In particular, we
demonstrate that a clean, well-behaved intrinsic mode function can be obtained from
a noisy velocity time series that allowed a precise determination of the vertical
structure of the coherent structures. The phase unwrapping of the UMR and the
identification of the phase related velocity components brings new insight into the flow
dynamics
Research supported by the Swiss National Science Foundation (2000-063818).
KEY WORDS: large scale flow structures (LSFS); gravel-bed rivers; empirical mode
decomposition; Hilbert transform |
|
|
|
|
|