Why niobium cavities quench below the predicted superheating field
In plain words
Theory gives niobium a maximum surface field of about 2400 G at low temperature, but the best cavities stop near 2000 G. The missing margin is not understood.
Precise statement
For clean Nb ($\kappa \sim 1$) Ginzburg-Landau and Eilenberger theory give $H_{\mathrm{sh}}(0) \sim 2300\text{ to }2400\,\mathrm{G}$, while record cavities quench near $B_{\mathrm{pk}} \sim 1900\text{ to }2100\,\mathrm{G}$. Determine whether the gap is caused by local defects that lower the vortex-entry barrier, by thermal runaway at hot spots, or by an intrinsic reduction of $H_{\mathrm{sh}}$ at GHz frequencies, and give the expected achievable field for an ideal surface.
What would settle it
Quench-site imaging and pulsed high-power measurements that separate thermal from magnetic quench, compared with theory for the measured surface.