Module gdess.data_source.models.e3sm.calculation

Functions

def add_global_mean_vars(xr_ds_: xarray.core.dataset.Dataset, variable_list, prefix='glmean_', weighting_var='area_p', averaging_dims='ncol') -> xarray.core.dataset.Dataset

For each variable in the input list, we add an accompanying variable that represents a global mean

Expand source code

def add_global_mean_vars(xr_ds_: xr.Dataset,
                         variable_list,
                         prefix='glmean_', weighting_var='area_p', averaging_dims=('ncol')
                         ) -> xr.Dataset:
    """ For each variable in the input list, we add an accompanying variable that represents a global mean
    """
    for var in variable_list:
        xr_ds_[prefix + var] = xr_ds_[var].weighted(xr_ds_[weighting_var]).mean(averaging_dims)
        xr_ds_[prefix + var].attrs["units"] = xr_ds_[var].units
        xr_ds_[prefix + var].attrs['long_name'] = xr_ds_[var].long_name + ' (globally averaged)'

    return xr_ds_

def calc_change_in_mass(dataset: xarray.core.dataset.Dataset, varname='glmean_TMCO2_FFF', prefix='deltas_') -> xarray.core.dataset.Dataset

Change in mass of $CO_2$ from timestep to timestep using backwards difference - $\Delta{mass_{CO_2}}$

Parameters

dataset : xarray.Dataset

 

varname : str

 

prefix : str

 

Notes

$$\Delta \texttt{glmean_TMCO2}_{t} = \texttt{glmean_TMCO2}_{t} - \texttt{glmean_TMCO2}_{t-1} \forall t$$ Returns

---

xarray.Dataset

Expand source code

def calc_change_in_mass(dataset: xr.Dataset,
                        varname='glmean_TMCO2_FFF', prefix='deltas_'
                        ) -> xr.Dataset:
    """Change in mass of \\(CO_2\\) from timestep to timestep
    using backwards difference - \\(\\Delta{mass_{CO_2}}\\)

    Parameters
    ----------
    dataset : ``xarray.Dataset``
    varname : `str`
    prefix : `str`

    Notes
    -----
    .. math::
        \\Delta \\texttt{glmean_TMCO2}_{t} = \\texttt{glmean_TMCO2}_{t} - \\texttt{glmean_TMCO2}_{t-1} \\forall t

    Returns
    -------
    ``xarray.Dataset``
    """
    dataset[prefix + varname] = xr.DataArray(calc_var_deltas(dataset[varname]),
                                             coords={'time': dataset['time']},
                                             dims=['time'])
    return dataset

def calc_global_weighted_means(dataset: xarray.core.dataset.Dataset, variable_list=None, prefix='glmean_', weighting_var='area_p', averaging_dims='ncol') -> xarray.core.dataset.Dataset

Global means with weighting by area

Parameters

dataset

 

variable_list

 

prefix

 

weighting_var

 

averaging_dims

 

Notes

xarray has recently introduced a weighting method (http://xarray.pydata.org/en/stable/examples/area_weighted_temperature.html). The area of each grid cell for the ne4pg2/ne30pg2 grids is called “area_p”. For the ne4/ne30 grids it is just called “area”. Not sure why they changed it.

$$\texttt{glmean_CO2var}_{t}= \frac{\sum_{i=1}^{nlat}\sum_{j=1}^{nlon} \texttt{CO2var}_{i,j,t} * \texttt{area}_{i,j}}{\sum_{i=1}^{nlat}\sum_{j=1}^{nlon} \texttt{area}_{i,j}} \forall t, with\,units: \{kg/m^2\}_t= & \frac{\sum_{i=1}^{nlat}\sum_{j=1}^{nlon} \{kg/{m^2}\}_{i,j,t} * \{m^2/m^2\}_{i,j}}{\sum_{i=1}^{nlat}\sum_{j=1}^{nlon} \{m^2/m^2\}_{i,j}} \forall t.$$ We would need to multiply by the radius of Earth ((R^2_{E}) in units of $\{m^2\}$) to get to $kg$ because area is in steradians rather than $m^2$. The above multiplication by area ((m^2/m^2)) is taken care of in the xarray.DataArray.weighted() method.

Returns

xarray.Dataset

Expand source code

def calc_global_weighted_means(dataset: xr.Dataset,
                               variable_list=None,
                               prefix='glmean_',
                               weighting_var='area_p',
                               averaging_dims='ncol'
                               ) -> xr.Dataset:
    """Global means with weighting by area

    Parameters
    ----------
    dataset
    variable_list
    prefix
    weighting_var
    averaging_dims

    Notes
    -----
    xarray has recently introduced a weighting method
        (http://xarray.pydata.org/en/stable/examples/area_weighted_temperature.html).
    The area of each grid cell for the ne4pg2/ne30pg2 grids is called “area_p”.
    For the ne4/ne30 grids it is just called “area”.  Not sure why they changed it.

    .. math::
        \\texttt{glmean_CO2var}_{t}= \\frac{\\sum_{i=1}^{nlat}\\sum_{j=1}^{nlon} \\texttt{CO2var}_{i,j,t} * \\texttt{area}_{i,j}}{\\sum_{i=1}^{nlat}\\sum_{j=1}^{nlon} \\texttt{area}_{i,j}}
            \\forall t,
        with\\,units: \\{kg/m^2\\}_t= & \\frac{\\sum_{i=1}^{nlat}\\sum_{j=1}^{nlon} \\{kg/{m^2}\\}_{i,j,t} * \\{m^2/m^2\\}_{i,j}}{\\sum_{i=1}^{nlat}\\sum_{j=1}^{nlon} \\{m^2/m^2\\}_{i,j}}
            \\forall t.

    We would need to multiply by the radius of Earth (\\(R^2_{E}\\) in units of \\(\\{m^2\\}\\))
        to get to \\(kg\\) because `area` is in steradians rather than \\(m^2\\).
    The above multiplication by `area` (\\(m^2/m^2\\)) is taken care of in the xarray.DataArray.weighted() method.

    Returns
    -------
    ``xarray.Dataset``
    """
    dataset = add_global_mean_vars(dataset, variable_list=variable_list, prefix=prefix,
                                   weighting_var=weighting_var, averaging_dims=averaging_dims)
    return dataset

def calc_time_deltas(xr_ds_: xarray.core.dataset.Dataset) -> xarray.core.dataarray.DataArray

Time deltas in seconds

Expand source code

def calc_time_deltas(xr_ds_: xr.Dataset
                     ) -> xr.DataArray:
    """Time deltas in seconds
    """
    seconds_per_day = 24 * 60 * 60
    return (xr_ds_['time_bnds'].diff('nbnd') * seconds_per_day).astype('float').isel(nbnd=0).round()

def calc_time_integrated_fluxes(dataset: xarray.core.dataset.Dataset, prefix: str = 'timeint_') -> xarray.core.dataset.Dataset

Time-integrated flux of $CO_2$ from surface emissions and aircraft $\int{flux_{CO_2}}$

$$\int_{t-1}^{t} \texttt{SFCO2} \approx \texttt{glmean_SFCO2}_{t} * \quad (time_t - time_{t-1}) \forall t, with\,units: \{kg/m^2\}_t= \{kg/m^2/s\}_t * \quad \{s\}_{t} \forall t.$$ Parameters

---

dataset : xarray.Dataset

 

prefix : str

 

Returns

xarray.Dataset

Expand source code

def calc_time_integrated_fluxes(dataset: xr.Dataset,
                                prefix: str = 'timeint_'
                                ) -> xr.Dataset:
    """Time-integrated flux of \\(CO_2\\) from surface emissions and aircraft
    \\(\\int{flux_{CO_2}}\\)

    .. math::
        \\int_{t-1}^{t} \\texttt{SFCO2} \\approx \\texttt{glmean_SFCO2}_{t} * \\quad (time_t - time_{t-1})
            \\forall t,
        with\\,units: \\{kg/m^2\\}_t= \\{kg/m^2/s\\}_t * \\quad \\{s\\}_{t}
            \\forall t.

    Parameters
    ----------
    dataset : ``xarray.Dataset``
    prefix : str

    Returns
    -------
    ``xarray.Dataset``
    """
    dtime = calc_time_deltas(dataset)
    #
    dataset[prefix + 'SFCO2'] = xr.DataArray(dataset['glmean_SFCO2'] * dtime,
                                             coords={'time': dataset['time']}, dims=['time'])
    #
    dataset[prefix + 'TAFCO2'] = xr.DataArray(dataset['glmean_TAFCO2'] * dtime,
                                             coords={'time': dataset['time']}, dims=['time'])
    #
    dataset[prefix + 'TOTALFLUX'] = xr.DataArray(dataset[prefix + 'SFCO2'] + dataset[prefix + 'TAFCO2'],
                                                 coords={'time': dataset['time']}, dims=['time'])

    return dataset

def calc_var_deltas(xr_da_: xarray.core.dataarray.DataArray) -> numpy.ndarray

Backward difference calculation

Expand source code

def calc_var_deltas(xr_da_: xr.DataArray
                    ) -> np.ndarray:
    """Backward difference calculation
    """
    return np.insert(np.diff(xr_da_), 0, 0)