Previous
abstract 
Next
abstract
We are investigating the pumping mechanism behind water masers in long-period variables by continuing to monitor the 22 GHz emission lines of 17 Miras and 5 semi-regular variables. Data obtained periodically since 1987 from the Haystack 37m telescope have shown all stars to exhibit highly variable water maser emission, with most Miras, minus a few exceptions, having phase lags of 0.3 phase between optical maximum and peak water maser intensity. Since the IR maximum lags the visual maximum by about 0.1 phase, peak maser intensity appears to lag luminosity maximum by about 0.2 phase. Such lags, on the order of 30-100 days, lend support to some form of collisional pumping. Direct radiative pumping of the maser, on the order of hours to days, would show no observable lag, and outflowing heated gases from near the photosphere would take on the order of years to reach the masing regions. Surprisingly, the observed lags appear to be independent of optical period, suggesting a possible correlation between optical period, hence luminosity, and distance of the maser spots from the photosphere.
The maser emission patterns of semi-regular variables appear more complex. In general, we observe more maser features extending over a larger velocity range with more varying degrees of lag than those observed for Miras.
