Effect of water on long-term weakening preceding rupture of crustal faults

Fault strength is a critical parameter in studies of crustal mechanics and for the prediction of earthquake hazards. The strengths of crustal faults inferred from borehole heat flow measurements and maximum stress orientations in the crust are less than those determined from laboratory measurements. Suggested causes of the weakening of faults include high fluid pressures, dynamic processes, or the presence of weak fault gouge. However, long-term changes of fault strength cannot be directly monitored using geophysical techniques, so an explanation for fault weakening remains an unsolved problem. We provide laboratory evidence that long-term weakening of the frictional strength of faults is caused by micro-fracturing at asperity contacts, which is a result of crack growth at subcritical stress levels. Our model suggests that long-term reductions of fault strength are related to chemical reactions that take place in the presence of water. For our measurements of friction on rupture surfaces in the presence of water, we increased temperatures to accelerate reaction processes so that they were observable at laboratory time-scales. In the presence of water, frictional strength decreased as temperature increased, whereas it changed little in the absence of water. The observed decreases in frictional strength were facilitated by chemical processes, rather than by physical processes governed by the effective pressure law. These observations suggest that chemical processes such as stress corrosion play an important role in long-term fault weakening. In addition to long-term monitoring of fault zones, we need to investigate long-term processes that cannot be observed during a human lifetime if we are to understand earthquake occurrences in the deep crust.