Impact of thermal property variability and structural layering in the lower crust on the continental geotherm and heat flow estimates

The thermal structure of the lower continental crust (LCC) is poorly constrained due to the lack of sufficient data on thermal properties, such as thermal conductivity (TC) and radiogenic heat production (A). However, this is essential for modelling the continental geotherm and heat flow as well as associated temperature-dependent processes and properties. Therefore, the few existing models that calculate the geotherm beneath the upper crustal level simplify the natural variability of TC and A of the lower crust by using averages of only few lithologies or even for an entire crustal section. Notably, individual crustal sections are defined as single thick layers, sometimes tens of km thick, which mismatch the evidence regarding the structure of the LCC. This makes heat flow and temperature calculations prone to errors and may lead to inaccurate and/or biased estimates of absolute values. A recent comprehensive study on the thermal properties of lower crustal lithologies (Lemke et al., 2026), carried out as part of the ICDP-DIVE (Drilling the Ivrea-Verbano zonE) project (Greenwood et al., 2025), fills this data gap and enables us to assess the impact of the variability of thermal properties in the LCC on geothermal and heat flow estimates.

 

To this end, we set up a 1D steady-state heat flow model of the continental crust, which is divided into an upper crust with constant properties and a lower crust with variable properties. The thermal property structure for the LCC is randomly drawn from lithology-specific A and TCdistributions (Lemke et al., 2026). Similarly, the thicknesses of the individual layers (d) are also drawn from predefined statistical distributions. Based on the available evidence, these distributions are typically Gaussian for the thermal properties (A, TC) and hyperbolic for the layer thicknesses (d), but we also test uniform distributions for these parameters. To assess the influence of the variabilities of TC, A, and d, model types are defined, for which each individual parameter as well as the combined effects of all three parameters are assessed. We compute geotherms upwards and downwards, starting with basal and surface heat flow values, respectively.  We test various lower crustal compositions: an intermediate one based on project DIVE as well as mafic and felsic endmembers. By performing numerous realisations for each model setup, the variability, as quantified by two standard deviations of temperature and heat flow is assessed.

 

The results show that the variability of thermal properties and layer thicknesses has a significant impact on temperature and heat flow. TC variability has the greatest influence on temperature uncertainties, while A variability has the greatest influence on heat flow uncertainties. Thicker layers, and layers with more widely varying thicknesses cause increasing uncertainties. The uncertainties reach ~10% for the heat flow while the temperature uncertainties are comparable to common corrections e.g. related to paleoclimatic signals. The chemical composition of the LCC determines the absolute value of the geotherm, but there is no significant impact on temperature variability. This work thus provides the basis for assessing geotherm and heat flow uncertainties in future models.