On the basis of elementary energy considerations and the nonlinear temperature dependence of experimental flow laws, a model is developed for the interplay of thermal mechanical processes along megathrust systems. The model is based on the observation that thrust systems involving large portions of the lithosphere have a two-part history. During the early history of movement, dry lower crustal/upper mantle rocks are deformed in a shear zone where, because of the nonlinear effect of temperature-dependent diffusion, strain-heating rates are directly controlled by the initial thermal state; that is, heating is insignificant if initial temperatures are high but where temperatures are low, large temperature increases are possible. As a result, strain heating acts as a ''thermal buffer'' such that upon arrival of upper crustal rocks at depths of 20 km or more, the temperatures along the thrust would be controlled largely by the rheology of hanging-wall rocks rather than the initial thermal state.

Moreover, the relatively high melting temperatures for dry, lower crustal/upper mantle rcks implies that the temperatures achieved along the thrust should be relatively high. As thrusting progresses, this early history is successively eliminated as wet, upper crustal rocks are emplaced beneath the ''preheated'' hanging wall. The granitic melts that characterized many natural megathrust systems are probably generated when these wet metasedimentary rocks move beneath the hanging wall, either by hydrous fluxing of the hot hanging wall or by melting in the highest-grade footwall rocks. Rapid uplift implied by both metamorphic studies and preservation of the inverted metamorphic sequence is interpreted as a late phase even created by ''tectonic pileup'' beneath the thrust.

Thorough testing of the model will require thorough assessment of the time-dependent problem. However, one-dimensional, steady state models together with energy considerations indicate that the general model is allowable. The concept of ''thermal buffering'' by strain heating is important because it implies that megathrust systems are a natural system where metamorphic temperatures are controlled by rock rheology; hence they represent real-world systems where a thermal and mechanical systems are intimately linked. This rationale is used to extend the model by indicating a method in which metamorphic conditions along exhumed megathrusts might be used as a natural constraint on experimentally determined flow laws.