Global Oscillation Network Group

The Global Oscillation Network Group (GONG) Project is a community-based activity to operate a six-site helioseismic observing network, to do the basic data reduction, to provide the data and software tools to the community, and to coordinate analysis of the rich data set that is resulting. GONG data are available to any qualified investigator whose proposal has been accepted; however, active membership in a GONG Scientific Team encourages early access to the data and the collaborative scientific analysis that the Teams are undertaking. The GONG Newsletter, information on the status of the Project, reports of scientific investigations, as well as access to the data, are available on our WWW server at www.gong.noao.edu.

The network continues to demonstrate a high degree of reliability. The vast majority of down-time is attributable to either weather or scheduled preventive maintenance (PM) visits. PM visits were made to Learmonth in January (10 days), Big Bear in February (10 days), and Mauna Loa in March (14 days).

The down-time caused by instrument failures totaled only about 27.5 hours for the three month period. This is the total time that any of the instruments failed to record data, including periods of bad weather and darkness, and is therefore the upper limit of the amount of useful data lost due to equipment failure. Nearly all of this down-time is traceable to trouble with the tape drives, either when the failure of both primary and backup systems prevented the recording of images, or when the entire system needed to be shut down for the repair or replacement of the faulty drives. It is thanks to the dedication and hard work of our on-site staff, that these problems were addressed promptly.

Caption: One of the first full-disk velocity images from the new GONG 1024 × 1024 camera.

The workstation at the El Teide site suffered a disk crash, and once repaired, failed again several weeks later. These failures did not affect the data acquisition system, but the ability to monitor the system status, either remotely or on-site, was severely compromised. Since the operator uses the workstation to perform the weekly tape change, there were some images lost when the procedure could no longer be performed in the standard way. Since this problem was anticipated, we asked the staff to perform the tape change at a time when the only data lost would be from after sundown. Considerable thanks go out to the El Teide staff who helped diagnose and repair this equipment. They have recently put the near-real-time weather data from the GONG station on the WWW at www.iac.es/weather/otdata.

Poor weather has had some effect on the observations, particularly at the Big Bear site. This winter's parade of El Niño storms has kept the Big Bear staff busy stowing and covering the GONG turret. Due to the temperature and humidity conditions near Big Bear lake, we still suffer from condensation on the turret mirrors. The attempt to dehumidify the air flowing into the turret has not been successful, and other approaches to the problem are under consideration.

During the past quarter, month-long (36-day) velocity, time series and power spectra were produced for GONG month 24 (ending 970916) with a fill factor of 0.77. The p-mode reprocessing campaign added GONG months 10 and 11. The project is also producing time series and power spectra from the intensity images. These products were generated for GONG months 18 and 19.

The Field Tape Reader (the subsystem that receives the raw-data cartridges from the observing sites) processed 85 cartridges containing 535 site-days from the six observing instruments. 390 site-days were calibrated.

The Data Storage and Distribution System (DSDS) serviced 25 data distribution requests for 2386 files totaling 1.5 Gigabytes of data. Most requests were filled within 24 hours of the receipt of the request. The DSDS performed 2,010 data cartridge transactions (library check-ins and check-outs) in response to requests from the data reduction pipeline and other internal operations.

Recently, the project completed software changes that should significantly improve the quality of the rotation-corrected, m-averaged, power spectra produced from the month-long time series. These changes include: 1) applying the rotation-correction (a shift along the temporal frequency axis) to the Fourier transform rather than to the power spectrum, 2) correcting for the two-point difference filter that detrends the time series of images after registration in heliographic coordinates, and 3) applying a rotation-correction that is a function of l, m, and n (radial node number).

The GONG mode frequencies have been approved for general release by the GONG DMAC Users' Committee, after examining the results of a random-restart test of the GONG peak-fitting algorithm. In this test, a new set of starting guesses for the fitting of a mode is created by adding a random perturbation to the initial fit results and the peak is then refitted starting from the new first guess. The added random component is uniformly distributed between twice the formal error of the fit. If the model of the peak is correct, and if the fit converges to the global minimum in the likelihood surface, the rms variation in the ensemble of "random-restart" fits should be a small fraction (2%) of the formal error estimate.

The GONG estimates pass this test after implementation of additional quality criteria. These criteria are:

1) strict numerical convergence in the minimization of the likelihood function;

2) a fitted width that is within a factor of 2 of the first guess width;

3) a test to ensure that the fit has not locked onto the first guess;

4) d/dl > twice the line width. Outside of this region, spatial leaks are blended with the target and our underlying model assumption of resolved leaks is not satisfied.

We have also revised the specification for the temporal coverage and cadence for the peak fitting. We will begin to produce frequencies for 108-day (3 GONG months) time series with centers separated by 36 days. This choice is a compromise between frequency resolution in the spectrum, which requires longer time series, and temporal resolution in the solar-cycle evolution of the parameters, which demands shorter time series. The resulting nominal frequency resolution of 0.1 µm Hz is sufficient for inversions for rotation and structure. On the other hand, the data will be correlated between successive estimates, and this must be taken into account in subsequent analyses. Finally, we are updating the first guess table, replacing the old initial guesses with estimates from the GONG data itself.

The Instrument Group is developing a higher-resolution (1024 × 1024) camera for the existing observing stations. Increasing the detector scale will provide significantly improved helioseismic resolution in the near-surface regions of the Sun that are the home of the intense magnetic fields that seem to cause much of the more dramatic aspects of solar activity, extend all aspects of "local helioseismology" dramatically, as well as enabling many non-helioseismic, diachronic solar measurements.

A Silicon Mountain Designs 1M60_20 camera has been obtained and subjected to a number of tests to verify its suitability. Although limited by our present video data acquisition electronics, the camera has been successfully used to acquire velocity images, and we are very pleased with the results.

The high-speed electronics are under development, and we look forward to a fully operational prototype system this summer.

John Leibacher