The Central Apennines in Italy have had multiple moderate-size but damaging shallow earthquakes. In 2016, three Mw > 6 earthquakes struck this region and caused severe damage. This sequence began with the Mw 6.2 Amatrice earthquake in late August, followed by the Mw 6 Visso earthquake in late October, and then the largest Mw 6.6 Norica earthquake 4 days afterwards. In the first part of this talk, I will present earthquake fault geometry and coseismic slip distribution of each major event using GPS and interferometric synthetic aperture radar (InSAR) measurements. I will also present deep-seated landslides triggered by the Amatrice earthquake. In the second part of this talk, I will show the InSAR-resolved crustal deformation in Central Apennines between the Amatrice and the Visso events. Based on the result, I characterize rapid postseismic transient as well as slip triggered by the Amatrice earthquake. Coseismic stress changes can trigger both aseismic and seismic slip along nearby faults. Faults that are velocity-strengthening (VS) are expected to always slip aseismically, while faults that are velocity-weakening (VW) can host aseismic and seismic slip. Although it is often assumed that the velocity-dependent properties, and therefore fault stability, remain constant throughout the earthquake cycle, recent laboratory experiments in carbonate rocks have shown that a switch from VS to VW can result from an increase in sliding velocity or displacement, thereby elevating the potential for unstable sliding following stress changes on otherwise stably creeping faults. Here, we use four independent InSAR time series to constrain slip following the Mw 6.2 Amatrice earthquake. We find a slow slip patch adjacent to the Amatrice coseismic asperity where aftershocks were nearly absent. This patch remained stable during the Mw 6 Visso earthquake that occurred two months after the Amatrice earthquake, but ruptured four days later in the Mw 6.6 Norcia earthquake. We hypothesize that the Norcia earthquake was triggered by slow slip when 1) an increase in aseismic sliding velocity caused the frictional properties to transition from VS to VW and 2) the slip patch size grew until it exceeded a critical nucleation size required for dynamic instability. Our geodetic data provides clear observations of an earthquake that nucleated from initial slow slip, and has greater implications for seismic hazards along fault systems worldwide.