This dissertation uses a combination of techniques to investigate the rheological properties of the lithosphere in continental areas and subducting slabs. In these settings, the relationship between temperature and the distribution of seismicity, particularly in the mantle, is examined. The dissertation is divided into three studies, which are largely self-contained and deal with different geographical and seismotectonic settings.

To begin, the properties of the lithosphere in the Lake Baikal region are examined in a multidisciplinary study incorporating observations of seismicity, gravity anomalies, topography, and seismic velocity and thermal structures. Teleseismic waveform-modelling is used to constrain the focal depths of 18 earthquakes, which are used along with published teleseismic and local-event studies to examine the seismogenic thickness (T_s) in the Baikal region. All sets of earthquake data show that the mantle in this region is not a significant source of seismicity. Estimates of elastic thickness (T_e) in the areas surrounding Lake Baikal are everywhere less than the seismogenic thickness, consistent with the simple interpretation that the long term strength of the lithosphere resides in its seismogenic layer, which in this region involves the whole crust but not, to any significant extent, the mantle. The apparent weakness of the mantle in the Baikal region is explained by its high temperature.

The thermal conductivity of mantle rocks is strongly dependent on temperature, and has been shown to have a significant effect on the modelled thermal structure of cooling oceanic plates. The second study in this dissertation models temperatures in subducting slabs, taking into account the temperature-dependence of the relevant physical parameters, and investigates the maximum temperatures, potential temperatures, and homologous temperatures up to which intraslab earthquakes occur. An analysis of the world’s subduction zones reveals that intraslab seismicity is consistent with being limited to material having potential temperatures less than 600 °C. Apparent exceptions to this pattern occur in regions where the Nazca Plate subducts subhorizontally beneath South America, with which the final study in this dissertation is concerned.

The unusual subduction geometry in the Peruvian and Pampean segments of the South American subduction zone keeps the overriding plate and the subhorizontal subducting slab in contact for several hundreds of km. Thermal modelling including this unusual geometry shows the subducting slab to be relatively cold in these regions, as compared with typical subduction zones where more immediate contact with the mantle wedge results in a hotter slab temperature structure. In the Peruvian and Pampean regions, intraslab seismicity is found to occur at potential temperatures less than 600 °C. Furthermore, the continental mantle above the slab is shown to be cold enough to generate earthquakes. Teleseismic waveform-modelling is used to constrain 39 earthquake focal depths, and confirms that the continental mantle of Peru is indeed seismogenic.