mixsea.overturn.eps_overturn

mixsea.overturn.eps_overturn(depth, t, SP, lon=0.0, lat=0.0, dnoise=0.0005, alpha=0.95, Roc=0.2, background_eps=nan, use_ip=False, N2_method='teos', overturns_from_t=False, pbinwidth=1000, EOS='gsw', linear_EOS_params={'SP0': 35, 'a': 0.0002, 'b': 0.0007, 'rho0': 1025, 't0': 15}, return_diagnostics=False)[source]

Calculate turbulent dissipation based on the Thorpe scale method. This function cannot handle NaNs in the input data, but there is another called `nan_eps_overturn’ that attempts to.

Parameters:

  • depth (array-like) – Depth [m]

  • t (array-like) – Temperature [°C]. If using gsw equation of state, it should have ITS90 °C units.

  • SP (float or array-like) – Practical salinity [g/kg]. Can be a single constant value.

  • lon (float, optional) – Longitude of observation (improves accuracy of TEOS-10 EOS)

  • lat (float, optional) – Latitude of observation (improves accuracy of TEOS-10 EOS)

  • dnoise (float, optional) – Noise level of density [kg/m^3] or temperature [°C], depending on overturns_from_t. Default is 5e-4.

  • alpha (float, optional) – Ratio of Ozmidov scale to Thorpe scale, alpha = Lo/Lt. Default is 0.95. Care must be taken to choose a value appropriate for the setting, e.g. Dillon 1982 [1], Ferron et al. 1998 [2]. Convert to Thorpe 1977 [3] conventions with C0 = alpha**2. Not to be confused with alpha in Equation 4 from Thorpe 1977, which is the inverse of our alpha.

  • Roc (float, optional) – Critical value of the overturn ratio Ro. An overturn will be considered noise if Ro < Roc.

  • background_eps (float, optional) – Background epsilon where no overturn detected. Defaults to NaN.

  • use_ip (bool, optional) – Sets whether to use the intermediate profile method. Default is False. If True, the dnoise parameter is passed as the `accuracy’ argument of the intermediate profile method.

  • N2_method (string, optional) – Method for calculation of buoyancy frequency. Default is ‘teos’. Options are ‘bulk’, ‘endpt’, ‘teos’ and ‘teosp1’.

  • overturns_from_t (bool, optional) – If true, overturning patches will be diagnosed from the temperature or conservative temperature, depending on the equation of state. Default is False.

  • pbinwidth (float, optional) – Potential density is not valid far from its reference pressure. For deep profiles, we loop over pressure bins. The pbinwidth parameter [dbar] sets the width of bins used for looping. Default is 1000, e.g. a 2000 dbar profile will use two bins and two different reference densities.

  • EOS (string, optional) – Equation of state, which can either be ‘gsw’ denoting TOES-10 or ‘linear’. The default is ‘gsw’. If you choose ‘linear’, the N2_method must be either ‘bulk’ or ‘endpt’. For the linear equation of state, density is calculated as rho0*(1 - a*(t-t0) + b*(SP-SP0)) (see parameter definitions below).

  • linear_EOS_params (dict of floats, optional) – Dict of parameters rho0, a, b, SP0 and t0, where rho0 is a constant density [kg/m^3], a is the thermal expansion coefficient [1/°C] and b is the saline expansion coefficient [kg/g] and SP0 and t0 are constants for salinity [g/kg] and temperature [°C]. The defaults are dict(rho0=1025, a=2e-4, b=7e-4, SP0=35, t0=15).

  • return_diagnostics (bool, optional) – Default is False. If True, this function will return a dictionary containing variables such as the Thorpe scale Lt, etc.

Returns:

  • eps (ndarray) – Turbulent dissipation [W/kg]

  • N2 (ndarray) – Background stratification of each overturn detected [s^-2]

  • diag (dict, optional) – Dictionary of diagnositc variables, set return with the `return_diagnostics’ argument.

References