Figure 6. Time series plots of chromatographic CFC-12 peak areas from air and calibration samples injected at the Barrow, Alaska between 06/10/91 and 11/10/91 (i.e. days 161 - 314 of the year 1991). Air samples were injected from two seperate sample streams shown here in green (AIR1 in the right-hand margin of the plots) and black (AIR2). Calibration samples were injected from two seperate working standard tanks shown here in red (CAL1) and blue (CAL2). The line plot of unflagged peak analyses (top frame) shows a considerable degree of chromatographic instability resulting from a poorly functioning gas sample valve that allowed intermittent injections of super and sub pressurized samples. Chromatography problems stemming from equipment malfunctions were common occurrences throughout the areas/heights database and uncorrectable by nature. In intermittent cases such as illustrated here, the ill effects were compounded by the fact that a single calibration sample might be involved in calibration response interpolations to several nearby air samples. This compounding effect was minimized by flagging the data as shown in the bottom frame. Here, line-connected data remain available for usage in atmospheric concentration calculations, while off-line data points are unavailable.