We find agreement between models of atmospheric chemistry changes from ionization for the 1908 Tunguska airburst event and nitrate enhancement in GISP2H and GISP2 ice cores, plus an unexplained ammonium spike. We then consider a candidate cometary impact at the Younger Dryas onset (YD). The large estimated NO_x production and O_3 depletion are beyond accurate extrapolation, but the ice core peak is much lower, possibly because of insufficient sampling resolution. Ammonium and nitrate spikes have been attributed to biomass burning at YD onset in both GRIP and GISP2 ice cores. A similar result is well-resolved in Tunguska ice core data, but that forest fire was far too small to account for this. Direct input of ammonia from a comet into the atmosphere is adequate for YD ice core data, but not Tunguska data. An analog of the Haber process with hydrogen contributed by cometary or surface water, atmospheric nitrogen, high pressures, and possibly catalytic iron from a comet could in principle produce ammonia, accounting for the peaks in both data sets.