Assimilation of subsurface temperature data to improve the thermal model of the European lithosphere

Abstract

As part of the IMAGE-FP7 project, we developed a methodology that resulted in an improved, physics-based thermal model of the European lithosphere. The main structure of our model is based on geophysical data, dividing the model into a four layer geometry consisting of sediments, upper crust, lower crust and lithospheric mantle. Temperature-dependent thermal properties, including radiogenic heat production and temperature- and pressure-dependent bulk thermal conductivity, are assigned on the base of the broad-scale lithological variation within the European crust. The temperature-dependent thermal properties are corrected with a 1-D steady-state temperature approximation, assuming only vertical heat flow. Using these corrected thermal properties, the 3-D thermal field is calculated with a conjugate-gradient method, assuming fixed temperatures at the surface and at the base of the lithosphere as boundary conditions, respectively. Further improvements of the thermal model, aiming at consistency between temperatures and heat flow observations and tectonic model predictions, are obtained by applying data assimilation. An Ensemble Kalman Filter (EnKF) is used to assimilate temperature data and improve the prior estimates of the thermal properties and the thermal field. Borehole temperature data can directly be used for this procedure. Since large regions in Europe lack (public) borehole temperature data, we applied data assimilation using regional thermal models – originally based on borehole data – to calibrate our European thermal model.