Earlier two-dimensional (2-D) and three-dimensional (3-D) model experiments have shown that the Limb Infrared Monitor of the Stratosphere (LIMS) nitric acid data do not behave as expected from conventional gas phase chemical reactions. As contrasted to two-dimensional (2-D) model results, the 3-D model suggests that the discrepancies are at middle latitudes outside of the polar vortex, rather than at polar latitudes. Using only the data record, the characteristics of the nitric acid behavior are further examined. The data inside the Aleutian anticyclone are examined during the January wave 1 warming. The anticyclone provides a large isolated region of air that moves from about 40¿N to 60¿N during the warming. Ozone remains approximately constant during this transit, while nitric acid increases more than 1 parts per billion by volume (ppbv). Both ozone and water vapor fields develop a wave 1 structure during the warming, as expected. Nitric acid, which is also expected to develop a wave 1 signal, develops a prominent wave 2 structure. This structure is observed between 50 and 5 mbar. A prominent feature of the nitric acid field is the persistent ''bending'' of contours due to strong meridional flow. Since these contours persist, instead of aligning with the flow, there must be chemical processes maintaining the nitric acid with timescales shorter than the advective timescale. The time constant for this chemical process ranges from approximately 1 day at 70¿N to about 4 days at 30¿N. When the time constant is used in the 3-D model, all of the basic characteristics of the observations are simulated. It is not clear what chemical mechanisms are responsible for this behavior. There is a strong relationship between the insolation and the shortcomings of the nitric acid simulations. Heterogeneous reactions on background aerosols are considered, but their spatial, temporal, and chemical characteristics are not clearly consistent with the needed changes in the chemistry. If heterogeneous processes are the mechanism, then they are much more complex than currently included in stratospheric models. ¿ American Geophysical Union 1993