41
Storm Impact on the Coastal Geomorphology and Current Field
clearly identified and validated with in situ data. In addition to that publication,
Seemann et al. (2000a) presented the very first version of DiSC in the account
German Association of Pattern Recognition (DAGM) Symposium. These two efforts
inspired the worldwide radar community. In the publication by Senet et al. (2000c),
the surface current field was taken into consideration for homogeneous areas, and by
applying 3-D spectral filtering on the 3-D complex image spectrum, the interesting
spectral parts of the wave field were isolated and inverted in local scale for the determination of the bathymetry. Even though it lacked a result validation, this contribution was an important and innovative approach. Trizna (2001) extensively discussed
the observed ambiguities from the inversion of the linear theory for the determination of the bathymetry. In the investigation by Hasan and Takewaka (2007), a similar
method was presented, with its main difference to the previous investigations being
that it is the maximum entropy method that has been used for the calculation of the
wave number leading to the generation of reasonable bathymetric results. However,
some of the validation data are approximately 20 years old, which is not reasonable
for areas with such high variations according to the conclusions of the previous work
(Galal and Takewaka 2008). According to Bell et al. (2004), the wave dispersion
relationship (Hedges 1976) is inverted with significant results, and since then, this
algorithm has been ameliorated and validated several times (Bell 2008; Hessner and
Bell 2009; Flampouris et al. 2009a).
In the context of the ValDiSC (Senet and Seemann 2002a,b) and the OROMA
projects (Ziemer et al. 2004), the DiSC algorithm was approved. Several research
efforts have been published during the development (Senet et al. 2000a,b; Seemann
et al. 2000a,c; Senet 2004), and an alternative method combining the advantages of
the previous investigations has recently been presented (Senet et al. 2008) for the
determination of the bathymetry from radar image sequences. The method analyzes
inhomogeneous image sequences of dynamic dispersive boundaries to determine
the physical parameters (bathymetry and current field) based on the deformation of the
wave spectrum and its reformation in local scale by using a selected wave the ory. The
DiSC algorithm is mature enough and is currently used quasioperationally for oceanographic investigations (e.g., Chowdhury 2007; Alamsyah 2008). The accuracy of
the linear version of DiSC is of the order O(10%) in comparison with echo soundings
(Flampouris et al. 2008), and recently, DiSC has been extended with nonlinear wave
theories (Flampouris et al. 2009b). It should be emphasized that a local or nearby generated wind sea permits DiSC to give results in a more excellent manner, because the
frequency spread improves the accuracy of the current in the wave direction travel,
and the directional spread improves the accuracy of the perpendicular component of
the velocity to the wave travel direction (Senet et al. 2001). Each of the derived current vectors is an independent measurement and independent form of the neighboring
values. The research for the further development of the algorithm of DiSC has been
completed and, for the first time, is presented in the following sections.
3.3 DISPERSIVE SURFACE CLASSIFICATOR
DiSC is an algorithm that allows the calculation of water depth and surface current maps from radar image sequences of the sea surface. The method of DiSC is
Storm Impact on the Coastal Geomorphology and Current Field
clearly identified and validated with in situ data. In addition to that publication,
Seemann et al. (2000a) presented the very first version of DiSC in the account
German Association of Pattern Recognition (DAGM) Symposium. These two efforts
inspired the worldwide radar community. In the publication by Senet et al. (2000c),
the surface current field was taken into consideration for homogeneous areas, and by
applying 3-D spectral filtering on the 3-D complex image spectrum, the interesting
spectral parts of the wave field were isolated and inverted in local scale for the determination of the bathymetry. Even though it lacked a result validation, this contribution was an important and innovative approach. Trizna (2001) extensively discussed
the observed ambiguities from the inversion of the linear theory for the determination of the bathymetry. In the investigation by Hasan and Takewaka (2007), a similar
method was presented, with its main difference to the previous investigations being
that it is the maximum entropy method that has been used for the calculation of the
wave number leading to the generation of reasonable bathymetric results. However,
some of the validation data are approximately 20 years old, which is not reasonable
for areas with such high variations according to the conclusions of the previous work
(Galal and Takewaka 2008). According to Bell et al. (2004), the wave dispersion
relationship (Hedges 1976) is inverted with significant results, and since then, this
algorithm has been ameliorated and validated several times (Bell 2008; Hessner and
Bell 2009; Flampouris et al. 2009a).
In the context of the ValDiSC (Senet and Seemann 2002a,b) and the OROMA
projects (Ziemer et al. 2004), the DiSC algorithm was approved. Several research
efforts have been published during the development (Senet et al. 2000a,b; Seemann
et al. 2000a,c; Senet 2004), and an alternative method combining the advantages of
the previous investigations has recently been presented (Senet et al. 2008) for the
determination of the bathymetry from radar image sequences. The method analyzes
inhomogeneous image sequences of dynamic dispersive boundaries to determine
the physical parameters (bathymetry and current field) based on the deformation of the
wave spectrum and its reformation in local scale by using a selected wave the ory. The
DiSC algorithm is mature enough and is currently used quasioperationally for oceanographic investigations (e.g., Chowdhury 2007; Alamsyah 2008). The accuracy of
the linear version of DiSC is of the order O(10%) in comparison with echo soundings
(Flampouris et al. 2008), and recently, DiSC has been extended with nonlinear wave
theories (Flampouris et al. 2009b). It should be emphasized that a local or nearby generated wind sea permits DiSC to give results in a more excellent manner, because the
frequency spread improves the accuracy of the current in the wave direction travel,
and the directional spread improves the accuracy of the perpendicular component of
the velocity to the wave travel direction (Senet et al. 2001). Each of the derived current vectors is an independent measurement and independent form of the neighboring
values. The research for the further development of the algorithm of DiSC has been
completed and, for the first time, is presented in the following sections.
3.3 DISPERSIVE SURFACE CLASSIFICATOR
DiSC is an algorithm that allows the calculation of water depth and surface current maps from radar image sequences of the sea surface. The method of DiSC is