 |
| Titel |
Single-ping ADCP measurements in the Strait of Gibraltar |
| VerfasserIn |
Simone Sammartino, Jesus Garcia Lafuente, Cristina Naranjo, José Carlos Sánchez Garrido, Ricardo Sánchez Leal |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250134208
|
| Publikation (Nr.) |
 EGU/EGU2016-14908.pdf |
|
|
|
|
|
| Zusammenfassung |
| In most Acoustic Doppler Current Profiler (ADCP) user manuals, it is widely recommended
to apply ensemble averaging of the single-pings measurements, in order to obtain
reliable observations of the current speed. The random error related to the single-ping
measurement is typically too high to be used directly, while the averaging operation
reduces the ensemble error of a factor of approximately √N, with N the number
of averaged pings. A 75 kHz ADCP moored in the western exit of the Strait of
Gibraltar, included in the long-term monitoring of the Mediterranean outflow, has
recently served as test setup for a different approach to current measurements. The
ensemble averaging has been disabled, while maintaining the internal coordinate
conversion made by the instrument, and a series of single-ping measurements has been
collected every 36 seconds during a period of approximately 5 months. The huge
amount of data has been fluently handled by the instrument, and no abnormal battery
consumption has been recorded. On the other hand a long and unique series of very
high frequency current measurements has been collected. Results of this novel
approach have been exploited in a dual way: from a statistical point of view, the
availability of single-ping measurements allows a real estimate of the (a posteriori)
ensemble average error of both current and ancillary variables. While the theoretical
random error for horizontal velocity is estimated a priori as ∼2 cm s−1 for a 50 pings
ensemble, the value obtained by the a posteriori averaging is ∼15 cm s−1, with an
asymptotical behavior starting from an averaging size of 10 pings per ensemble. This result
suggests the presence of external sources of random error (e.g.: turbulence), of higher
magnitude than the internal sources (ADCP intrinsic precision), which cannot be
reduced by the ensemble averaging. On the other hand, although the instrumental
configuration is clearly not suitable for a precise estimation of turbulent parameters,
some hints of the turbulent structure of the flow can be obtained by the empirical
computation of zonal Reynolds stress (along the predominant direction of the current) and
rate of production and dissipation of turbulent kinetic energy. All the parameters
show a clear correlation with tidal fluctuations of the current, with maximum values
coinciding with flood tides, during the maxima of the outflow Mediterranean current. |
|
|
|
|
|
|