DISCREPANCIES BETWEEN DATA INDICATING TIME- AND SPACE-VARIABLE DEFORMATION

Discrepancies between data are often more interesting than agreement, following Niels Bohr's observation “How wonderful that we have met with a paradox. Now we have some hope of making progress.” Increasingly, discrepancies in crustal or plate motions measured by different techniques appear to be real and reflect the fact that the data sample differently in space and time. Recognizing these differences can also explain apparently-discordant geomorphic, structural, and earthquake data. At plate boundaries, motions change on scales of tens of millions of years, as evidenced by the slowing of Nazca-South America motion associated with the uplift of the Andes. The geometry and rate of motion of fault systems and microplates within plate boundary zones can change dramatically on million-year timescales although motion between the major bounding plates remains constant. For example, the motion of the Adria microplate within the Nubia-Eurasia convergent zone changed dramatically in the past two million years. The apparent contradiction between a thrust belt on the eastern coast of Italy and active normal faulting in the Appennines, illustrated by the April 2009 L'Acquila earthquake, reflects the change in the motion of Adria from westward subduction beneath Italy to eastward divergence from Italy. The locus of the very slow deformation within plates migrates on time scales as short as thousands of years, as fault systems turn on and off. Such migration is illustrated by GPS data across the New Madrid Seismic Zone showing no motion despite large earthquakes within the past 2000 years. This effect, combined with the fact that earthquake aftershock sequences within continents can extend for hundreds or thousands of years, explains why some faults can appear seismically active with little topographic expression, or the reverse, as indicated by the May 2008 Sichuan earthquake on a fault that appeared seismically inactive.