isotopylog.HeatingExperiment.forward_model

HeatingExperiment.forward_model(kd, nt=300, z=6, **kwargs)

Forward models a given kDistribution instance to produce predicted evolution.

Parameters:

• kd (isotopylog.kDistribution) – The ipl.kDistribution instance containing the rate model used to fit the data.
• nt (int) – The number of time points to use in the forward-modeled data estimates. Defaults to 300.
• z (int) – The number of neighbors in the carbonate lattice. Only used if he.model == 'SE15'. Defaults to 6, as desribed in Stolper and Eiler (2015).

Returns:

he – The updated ipl.HeatingExperiment instance, now containing forward-modeled clumped isotope and reaction progressestimates.

Return type:

isotopylog.HeatingExperiment

Notes

Uncertainty is calculated using the Jacobian of the model fit function containing the derivative of this function with respect to each modeled parameter. Jacobians are calculated by perturbing each parameter and calculating the finite difference derivative. If uncertainty bands appear too noisy, pass a lower value of eps, e.g., eps = 1e-5.

See also

isotopylog.kDistribution()

The class containing all rate data that are to be forward-modeled.

Examples

Assuming some kDistribution instance kd exists and contains the rate data for the model of interest:

he.forward_model(kd)

To change the number of forward-modeled time points to 1000:

he.forward_model(kd, nt = 1000)

References

[1] Passey and Henkes (2012) Earth Planet. Sci. Lett., 351,

223–236.

[2] Henkes et al. (2014) Geochim. Cosmochim. Ac., 139, 362–382.

[3] Stolper and Eiler (2015) Am. J. Sci., 315, 363–411.

[4] Hemingway and Henkes (2021) Earth Planet. Sci. Lett., 566,