hydrac.model.scattering.scattering

Scattering calculation (hydrac.model.scattering.scattering)

class hydrac.model.scattering.scattering.Scattering(particle, water, ori=-3.141592653589793, freq=None, tag_M=None, gen_model=None, mode_depl=None, am=3e-05, aM=0.0006, dsigma=None, spacing='log', UND=4)[source]

Bases: object

Scattering class.

This class computes the scattering model given the input properties of particles specified by the class hydrac.model.particle.Particle, and several other inputs such as the carrier frequency of the instruments, the water parameters (ex. sound velocity in water)…

Two major cases are considered regarding the inversion procedure:

The case of singlefrequency inversion: In this case, the mean radius of the particles is entered as input. The model is computed at high resolution for display purpose and potential averaging over size distribnution if one is indicated, but only one single value will be used.

The case of multifrequency inversion: One of the multifrequency inversion methods (NNLS, see hydrac.inversion.multifrequence.nnls) considers an undetermined inversion system, where the number a size classes is defined, the concentrations of which correspond to the system unknown. In this case, in addition to the high-resolution model, it is convenient to compute the model for each ka values corresponding to the instrument frequencies and the target sizes of the particles. As such, the model is computed at two distinct resolutions.

The “model” returned by the Scattering class is two-fold:

  • The form function \(f\)
  • The total scattering cross-section \(\chi\), from which the attenuation due to scattering can be evaluated.

This class can be used as a standalone if one is interested in computing a specific model, and export it elsewhere for other purposes. The Typical usage for a standalone package to compute models (here shown for a model parameters stored in the database, copepDWBA):

# Define the particle properties with a Particle object
>>>P = hydrac.model.particle.Particle('copepDWBA',1)
# Define the water properties with a Water object
>>>W = hydrac.model.water.Water(preset='sea')
# Compute the target strength and attenuation
>>>S=hydrac.model.scattering.scattering.Scattering(P,
                                                   W,
                                                   freq=[1e6, 4e6],
                                                   tag_M='multi',
                                                   mode_depl='Mooring',
                                                   am=0.06e-3,
                                                   aM=2.8e-3)
# Access the form function
>>>S.param.f_inf

Within hydrac processing chain, the Scattering class takes the data from the hydroacoustic object as inputs. It is solely called when performing the inversion in the package hdrac.inversion.inversion.Inversion.

All the results are stored into a scattering modified dictionnary (see hydrac.util.parameters).

Parameters:
particle : object

Particle object from hydrac.model.particle.Particle

water : object

Water object from hydrac.model.water.Water

ori : float

Angle of observation => angle between incident and oscerved scattered wave (\(\pi for backscattering\))

freq : float (numpy.array or list)

Target frequencies of the hydroacoustic instrument

tag_M : str, {‘mono’,’multi’}

Tag specifying whether the instrument is multi or single frequency

gen_model : bool

Save the model in netCDF as a standalone

mode_depl : str, {‘Mooring’, ‘Cast’, …}

Deployment strategy (see hydrac.instrument.mod_deployment)

am, aM : float, float

Minimum and maximum radius for model computation

dsigma : float

Distribution width if the user wants the model to be function of mean radius. Not used for multifrequency inversions.

spacing : str, {‘lin’, ‘log’}

Spacing type of the radii vector (log to better sample fine particles)

UND : int

Underdetermination factor for the multifrequency inversion system.
tmp_del(self)[source]

Deletes temporary model information

ensemble_average_model(self, xx, dsigma, freq=[1000000.0], na=10000.0, am=1e-06, aM=0.01, space='lin')[source]

Computes the ensemble averaged model over the size distribution if the dsigma input of the class is not None.

Parameters:
xx : float

Taget \(ka_0\) vector, with \(a_0\) the mean radius, similar to the ka variable

dsigma : float

Distribution width

freq : float

Reference frequency (single frequency case)

na : float (numpy.array or list)

Number of integration points for the size distribution

am, aM : object

Min and max radius to be considered for the size distribution

space : str

Spacing of the new ka vector
Returns:
axlng : float

Taget \(ka_0\) vector, with \(a_0\) the mean radius

aflng : float

Ensemble averaged form function, function of \(ka_0\)

achilng : float

Ensemble averaged total scattering cross-section , function of \(ka_0\)
RES_model(self, am, aM, space='log', UND=None, na=None, freq=None)[source]

Calculates the scattering model by calling the right method

Parameters:
am, aM : float

Taget \(ka_0\) vector, with \(a_0\) the mean radius, similar to the math:ka variable

space : float

Distribution width

UND : float

Reference frequency (single frequency case)

na : float (numpy.array or list)

Number of integration points for the size distribution

freq : object
Returns:
atmp : float

Radius vector

katmp : float

\(ka\) vector

f_inftmp : float

Intrinsic form function, function of \(ka\)

c_inftmp : float

Intrinsic total scattering cross-section, function of \(ka\)

aij : float

Backscattering cross-section, function of \(ka\)
scatt_func_display(self, preset='basic', *args, **kwargs)[source]

Display the scattering model. Typical usage:

# default view as a (1row, 2columns) subplot
>>>S.scatt_func_display()
# view as a (2row, 1column) subplot
>>>S.scatt_func_display(subplot=[2,1])
calc_f(self, func, ka)[source]

This method is called by each class called by the present Scattering class to compute the model for each ka vectors in case of multifrequency instruments.

Parameters:
func : class method

method used to compute the desired scattering model

ka : float

\(ka\) vector

Classes

Scattering(particle, water[, ori, freq, …]) Scattering class.

Exceptions

InstrAmbiguity