9.7 Frequently asked questions

• How often is the phase correction applied? Statistics have been calculated for all receiver 1 CORR scans in the preliminary data reduction files between September 1997 and March 1999 (a total of 3120 hours). Over this period, the phase correction was diagnosed ``good'' during 78.8% of the observing time.
• Does one gain something on a self calibrated source? Yes, the monitor corrected spectra are improved on a sub-scan scale, i.e. on one second. An example for this case can be found in Fig. 9.6.
• How fast can the phase change? Phases at 3mm can turn by more than 360$^\circ$ over 30 seconds. Even under calm conditions, it was found that a time resolution of one second for the correction gave better results than an average correction over four seconds.
• Is dry weather the same as low phase noise weather? Unfortunately not. Even with 3mm precipitable water, one can have very bad phases if the wind is high (about 10 m/s).
• Why is the phase correction sometimes disabled under clear sky conditions? Each correction system has its intrinsic noise. If the atmospheric phase noise is below 10$^\circ$ at 3mm or 30$^\circ$ at 230 GHz (which happens under stable winter conditions), the added noise will undo the benefits, and the diagnostics will switch the correction off. The same can happen under less favorable conditions if a receiver has an unstable total power signal (this can happen at some frequencies).
• How does the corrected phase depend on baseline length? Basically, it becomes independent. Over time scales longer than the monitor time interval, the structure function dependence stays.
• Are there still changes / improvements in the system? Yes, we are working on it.