Zoe K Mildon , Gerald Roberts, Joanna Faure Walker, Shinji Toda

Successive locations of individual large earthquakes (Mw>5.5) over years to centuries can be difficult to explain with simple Coulomb stress transfer (CST), because seismicity can miss out nearest-neighbour along-strike faults where coseismic CST increases are greatest. We show that “Coulomb pre-stress” may explain this, because magnitudes are >±50 bars if interseismic loading and local stress amplification at fault bends are included, so coseismic CST, in the range of ±2 bars, will rarely overwhelm the Coulomb pre-stress. To illustrate this, we calculate the Coulomb pre-stress prior to 34 earthquakes from 1349-2016 A.D. in central Italy and use this to discuss the location of subsequent earthquakes. We show that earthquakes tend to occur where the cumulative coseismic and interseismic CST is positive. Ruptures propagate both across faults that are positively stressed, and in a few examples, from positions where highly stressed patches associated with along-strike fault bends are surrounded by negatively stressed fault surfaces. Coulomb pre-stress calculated for strike-variable faults is an ignored yet vital factor for earthquake triggering.