This chapter demonstrates that least-squares inverse modeling is possible in 2-D thermal and maturity modeling using fast approximate forward (FAF) solutions to the transient heat equation. The variability in the thermal reconstruction, induced by data and heat flow uncertainties, is investigated using a Monte Carlo–based sensitivity analysis. This results in a large number of scenarios, which are all consistent with data and a priori knowledge. Two thermal approximations with different parameterizations of inversion variables are investigated.

The results show that past sediment temperatures are not well constrained by borehole temperature and vitrinite reflectance when sediment temperatures are at their maximum at the present day. The a priori choice of heat flow is decisive for large periods of the time interval considered. Sensible heat flow solutions in 2-D and 3-D inverse modeling call for the use of a well-specific inversion parameter affecting the estimated temperature gradient to account for diverging temperature and vitrinite reflectance depth trends in the inversion data.

The method is applied to three profiles on the Norwegian shelf. At one well, a reaction kinetic network was applied to the source rock temperature histories, yielding an estimated 22 4% realization of the oil potential and 18 3% realization of the gas potential.