Comparison of Archived Dataset with AIR
GONG data from a 108 day period (GONG months 77-79: 021102 - 031117) were processed using AIR to select images for inclusion. Helioseismology products generated from this data set were compared with the archived products.
During month 77 UD was not operating and ML was having a scheduled PM (Preventive Maintenance) during the first two weeks. However, LE was operating normally and experienced generally good weather, so there were no major gaps in coverage. LE was not operating for much of months 78 and 79. UD resumed operation early in month 78.
The duty cycles based on the window functions from the merged monthly time series are:
Month Archived AIR 77 87% 92% 78 86% 88% 79 86% 87%
Helioseismology products generated from this data set were compared with products from the original, normally processed data set
The AIR data set had a slightly higher net duty cycle: ~89% vs ~86%. The difference in duty cycle at individual sites was generally slightly larger before allowing for station coverage overlap.
The results from comparing mode frequency peaks (shown in Fig. 1 and 2), l-nu diagrams, and a Regularized Least Squares inversion (Fig. 3) were all very similar for the two datasets.
Ring diagram derived flows are shown in Fig. 4.
Figure 1 PEAKFIND was used to identify p-modes from the time series produced from each version of the data set. The ratios of mode widths, amplitudes, and strengths of individual modes are shown.
Figure 2 Here are histograms of the distribution of mode widths, amplitudes, and strengths. There are systematic differences in the widths and amplitudes reflecting the different duty cycles but using the product of width * amplitude as a measure of mode strength shows very little difference between the two data sets.
Figure 3 The results of a regularized least squares inversion to find rotation rate vs depth for several latitudes. The line width is proportional to the uncertainty in the inversion.
Figure 4 A ring diagram analysis was used to obtain flow velocities. Here the x component of velocity is shown. The velocities and error bars from the standard dataset are plotted with the velocities from the AIR dataset over plotted in red.