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 and 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. Information on the status of the Project, the scientific investigations, as well as access to the data is available on our WWW server whose URL is http://www.gong.noao.edu.
The GONG 1999 annual meeting was held in Tucson, 22-24 March. The goal of the workshop was to provide a forum for the discussion of the work in progress to understand the physics of the moderate and high spherical harmonic degree modes, the data analysis methods needed to extract the appropriate description of their properties, and the physics of the solar interior that can be derived from them. But with nearly ninety participants, the open forum also presented the opportunity to discuss all aspects of helioseismology. The meeting presented the opportunity to stimulate the work of the GONG community in anticipation of the high-resolution data that GONG+ should start providing in mid-2000. The conference abstracts are available at the GONG web site.
The Meeting was also the occasion for the Scientific Advisory Committee and the DMAC Users' Committee to meet. Representatives from all of the sites were in attendance, providing an opportunity, with everyone present, to discuss the continued GONG operations, the installation and certification of the GONG+ camera system, and the day-to-day responsibilities associated with operating the network.
During the first quarter of 1999, the GONG network operations have been generally successful despite several instances of component failure. Cerro Tololo experienced two instances of component failure. One was a faulty signal generator card that was diagnosed and replaced in less than two hours. The other was a more unusual and troublesome problem, which prevented the instrument from guiding. After a thorough checkout of the system, a faulty circuit card was identified, replaced, and the instrument was restarted, but not until nearly sixteen hours of images were lost. There were also two occasions when the data acquisition stopped, requiring a remote system reboot (from Tucson), which resulted in an additional seventeen hours of lost data.
During a routine site check, some unusual features were noticed in the images from Big Bear. At first it was thought that the camera had developed a problem, but after a few tests the problem pointed to the data acquisition system. With some clever troubleshooting by the Big Bear staff, it was determined that a faulty power supply was generating noise. About seven hours of lost images resulted from troubleshooting and repair. In looking back through recent site-checking information, it appears that the problem began about ten days prior to its detection. Consequently, these days are lost as well, as it is not likely that a correction can be applied to the images. In addition, the Mauna Loa data computer suffered intermittent power loss due to a poor connection in a power supply connector (nearly thirteen hours of images were lost), and an ADC board failed in the El Teide instrument but was quickly diagnosed and replaced. Downtime was also incurred at several of the sites due to sporadic tape drive failures, but the number of images lost amounted to only a few hours and was relatively minor compared to many previous quarters.
Several GONG stations have been affected by severe weather during the last three months. Perhaps our greatest concern was the Learmonth instrument when cyclone Vance swept past with wind speeds of 267 km/hour on 22 March. First reports to the GONG operations team included, "The GONG building is standing, but its situation is not known beyond that." Surprisingly, the damage amounted to a puddle of water and some dirt on the floor of the shelter. The instrument came up without a hitch after a four-day shutdown. Some damage was sustained at the Learmonth Solar Observatory and considerable damage occurred in the nearby town of Exmouth, but thankfully there were no casualties. Across the network, the El Teide site was hit by their worst storm in 40 years piling up in excess of three feet of snow, causing a nine-day shutdown of the GONG instrument in January. The GPS antenna was damaged, but site staff mounted the spare and the system was back online. The Big Bear site has seen numerous snow storms this spring, but none harsh enough to cause damage.
With all the hardware and weather-related incidents affecting the GONG network in the last quarter, we are once again reminded how important it is to have concerned, committed, and skilled staff in the field, at the front-line of the effort sustaining a reliable telescope network--hats off to all our site personnel!
The reduction of Classic GONG data (three years of nearly continuous images from the six-site network) has been completed, including the data that were recorded during the deployment phase (with three observing stations in operation). A total of 35 GONG months (36-day months) from 950507 to 981017 were acquired: p-mode power spectra have been reduced from month-long time series, four-minute averages have been computed, and various magnetogram products produced. In addition, the DMAC produced low-l power spectra (for l's up to 45) from the full Classic GONG 1280-day time series and presented a poster displaying these spectra at the GONG `99 workshop (available on the GONG web site).
The campaign to identify the mode frequencies in the three-year data set added eleven sets during the past quarter, bringing the total number of mode frequency tables to nineteen. These mode frequencies were extracted from power spectra generated from 108-day time series and centered every 36 days. When the campaign is completed (July 1999), 33 sets of mode frequencies will have been produced, one for each Classic GONG month, less the first and last months.
Recently, the data between the beginning of GONG's three-site network operations (950507) and the original magnetogram reduction campaign with Carrington Rotation 1906, (early 1996) were inspected, and it appears that some of these data will produce magnetogram products of acceptable quality. Thus, data will be processed in the near future extending the series of GONG synoptic maps back to the beginning of the three-site network operations.
Using the entire 1280-day time series of calibrated velocity images, an attempt was made to detect some of the very weak velocity fields on the surface of the Sun. Although the detection of these types of velocity fields was not a high-priority objective in the design of the GONG observing equipment, it appears that a marginally detectable signature of residual differential rotation (torsional oscillation) can be extracted.
The frequencies of solar oscillations have been determined for 27 of the 33 108-day blocks covering the period 7 May 1995 to 29 September 1998. During this time, solar activity has progressed from a minimum in early 1996 and into a rapidly ascending phase of cycle 23. The GONG measurements continue to show the expected increase in the basic frequencies of the oscillations, as well as more subtle variations (the so-called even splittings) which are well correlated with the activity level, and which appear to be linked to the latitudinal distribution of the magnetic field. We have recently pushed the expansion of these subtle variations up to eighth order, and have found that the correlation with the corresponding higher-order latitudinal components of the magnetic field still holds true. (The results are in press in the ApJ.)
The random-restart peak-fitting test of the multi-tapered spectra is temporarily on hold while we work around a problem with the random number generator. Once this test is completed, we will begin to use multi-tapering on the GONG spectra routinely.
The proposed correction of the estimated widths and amplitudes to reduce the effects of an un-deconvolved temporal window is proving to be robust as we add more data. We have attempted to detect solar cycle related variations in the widths and amplitudes, and we do see preliminary evidence of such changes.
Progress on the leakage matrix calculation has slowed due to other NSO demands for resources. The modulation transfer function and the vector nature of the oscillatory velocity field still remain to be installed. We have computed the matrix for two three-month time series.
The multi-dimensional fitting now provides very nice fits of peaks in the
m-averaged l- diagram, along with an empirical estimate of the leakage
matrix. However, it is still computationally very expensive.
At the recent DMAC Users' Committee meeting it was recommended that the planned development of a new peakfind may be postponed, and we will first develop a realistic simulation of time series for use in testing peak-finding algorithms.
Data for a two-day time series have been merged at the image level, and high-quality ring diagrams were produced. This test has conclusively demonstrated that it is vital to organize the processing such that only a single interpolation is used. When two interpolations were applied, strong artifacts appeared in the spectra.
Two scientific visitors, Sushant Tripathy (Udaipur Solar Observatory) and Markus Roth (Kiepenheuer-Institut), are working with GONG project staff. Sushant is pushing the temporal resolution of solar-cycle variations by analyzing short (eighteen-day, nine-day, etc.) time series, while Markus is going in the opposite direction, analyzing one-year time series to search for his theoretical frequency shifts arising from large-scale velocity fields. Pier Francesco Moretti (University of Rome), who was studying power spectra of magnetic field measurements, has just departed.
A GONG+ high-resolution camera, installed at the Tucson facility, is now providing full- resolution solar images to a prototype video data acquisition system. Although we are in the beginning stages of analyzing the data, the performance of the prototype system is very promising. A prototype data caching system, consisting of two DLT7000 tape drives and four 18GB hard-disk drives, has also been assembled at the Tucson station. The caching scheme is new to the data storage concept and design of the original GONG system, but should improve the efficiency of the tape drives and solve the increased data flow (a factor of sixteen improved areal resolution, plus continuous line-of-sight magnetograms). The DMAC is gearing-up as well, beginning to address the challenge of pumping 10242 images through the data reduction pipeline.
John Leibacher