The motional Stark effect diagnostic reliably measures significant deviations in safety factor profile during DIII-D sawteeth

Motional Stark effect (MSE) data acquired during large fast-ion stabilized sawteeth are critically reexamined. The safety factor at the sawtooth crash changes by $\Delta q\simeq0.15$, much more than any likely errors, indicating that substantial reconnection occurs at these sawtooth crashes. The absolute magnitude of the central safety factor after the crash is less certain: $q_0\simeq0.90$-0.97 with an uncertainty of $\sim0.05$.