THE GLOBAL PROJECT
Adv. Space Res. Vol. 24, No. 2. pp. 173-176. 1999 Q 1999 COSPAR. Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain 0273-l 111199$ZO.OQ + 0.00 PII: 50273-I 177(99)00497-4
John W. Leibacher and the GONG Project Team
National Solar Observatory, National Optical Astronomy Observatories Pod Ofice Boz 26732, ficson, Arizona 85726, USA
ABSTRACT The Global Oscillation Network Group ( GONG ) is an international, community-based project to conduct a detailed study of solar internal structure and dynamics using helioseismology. In order to exploit this technique, GONG has developed a network of six extremely sensitive and stable solar velocity imagers to obtain nearly continuous observations of the Sun’s “Eve-minute” oscillations. GONG also provides magnetograms nominally every twenty minutes. The system became operational in October 1995, and will operate for an eleven-year solar cycle. The observation duty cycle has averaged about 90%, and single-site data loss due to instrument down time is less than 2%. Data is processed at pace with the collection rate. Progress is underway to replace the original 256 x 242 rectangular pixel cameras, with 1024 x 1024 square pixels cameras to obtain the optimal scientific return from the instruments from the eleven-year run. There are 175 individual members of the GONG, from 70 different institutions, and 20 nations. 0 1999 COSPAR. Published by Elsevier Science Ltd. INTRODUCTION The Global Oscillation Network Group ( GONG ) Project grew out of community discussions in the early 1980’s, got its start with a workshop in 1984, ~88 approved for funding by the National Science Foundation in 1987, and commenced full network, and data management and analysis center, operations in October 1995. It is scheduled to observe for eleven years. The nine-year-long site survey suggested that the six sites would deliver a duty cycle close to 90%, and that has been achieved. More significantly, the daily sidelobes caused by periodic data losses - are negligible. The instrument, a Fourier Tachometer, displays robustness, good repeatability, sensitivity and linearity. It has proven to be extremely reliable in practice. The data processing pipeline is based upon a network of workstations and is able to keep up with the data received from the sites with a margin for reprocessing needs. The data storage and distribution system provides data to GONG members over the Internet and via removable media. The overall status of the Project has been described by Harvey et al. (1992), Leibacher et al. (1995a), and Leibacher et al. (1995b). The initial scientific results were presented in the May 31, 1996 issue of Science. SITES The six sites that comprise the GONG network are: Big Bear Solar Observatory, California, USA; Mauna Loa Solar Observatory, Hawaii, USA; Learmonth Solar Observatory, Australia; Udaipur Solar Observatory, 173
J. W. Leibacher and the GONG Project Team
India; Observatorio de1 Teide, Tenerife, Canary Islands, Spain, and Cerro To1010 Interamerican Observatory, Chile. These sites were selected after an extensive survey of 15 candidate sites for periods of two to six years; see Hill et al. (1994a), and Hill et al. (1994b) . INSTRUMENT The GONG instrument has been described by Harvey et al. (1988) and the optical qualities have been presented by Harvey et al. (1995). The heart of the GONG instrument is a Michelson-interferometer-based “Fourier Tachometer” which measures the Doppler velocity of the Ni I 6768 A line from the entire solar disk simultaneously, with a 256 x 242 pixel CCD. Each go-second integration produces three intensity images, differing in modulation phase by 120”, which are processed to yield a velocity image, as well as intensity and line strength images. Magnetograms are acquired hourly at each site, with adjacent sites shifted by 20 minutes, so that magnetograms with a 20 minute cadence are available. The data is stored on Exabyte cartridges for processing at the central data reduction and analysis facility. The entire instrument resides in an environmentally controlled shelter building, which houses an external light feed, the instrument itself, control and acquisition computers, and data recording equipment. The light feed is an automatic system, which turns itself on each day, acquires the Sun, continues to track using an ephemeris during cloudy periods, and supervises and reports its environment and operational status. The instruments are continuously intercompared and coalignment of the images to a small fraction of a pixel is readily achieved. Thanks to our dedicated partners at the sites and routine preventative maintenance visits by project staff, downtime has been negligible. The prototype instrument has been refurbished to be functionally identical to the field instruments, and it provides a testbed for trouble shooting problems in the field and evaluating the new camera system. DATA MANAGEMENT
The Data Management and Analysis Center ( DMAC ), described by Pintar et al. (1988), incorporates a pipeline of processing steps, and a Data Storage and Distribution System ( DSDS ). The DMAC processes 1 GB of data per day ( 12 kB/s ) to maintain cadence with the data acquisition rate. The DMAC’s processing capacity incorporates a margin to allow reprocessing without falling behind the input data stream. The p-mode data reduction pipeline consists of several steps: instrument correction ( camera and optical calibration ) including the determination of the modulation transfer function ( MTF ); transformation of the calibrated velocity images to spherical harmonic coefficients as a function of degree ( .t?) and azimuthal order ( m ); the combining, or merging, of simultaneous data obtained at multiple site including image restoration ( seeing and scattering corrections ); assembling month-long time series ( GONG “months” are 36 days long ); Fourier transformation of the coefficients to power as a function of temporal frequency ( v ); and extraction of mode parameters from the power spectra for all distinct values of e, m, and n out to e of 250. The fitting is currently being carried out for three GONG month intervals. Magnetograms am extracted from the data stream when the oscillation images are calibrated. The calibrated images are resampled to a four-minute cadence for low frequency and steady flow analysis. A science data product is associated with each of these processing steps. These data products along with the raw data and the ancillary data ( e.g., instrument header parameters and processing histories ) now total over 3 TB. The processing pipeline activities are distributed among approximately a dozen high-end workstations, and the data is transported between the processing stages by Exabyte tapes, which are also the archival medium within the DSDS. The data is checked into and out of the DSDS, EISthe processing proceeds, thus each processing stage ensures that the day products have been successfully archived. Oracle is utilized as the relational database management system underlying the DSDS, and a DSDS user interface using the World Wide Web ( www.gong.noao.cdu ) provides data to the community.
The Global Oscillation Network
THE FUTURE The initial operations and data analysis have convincingly demonstrated that a network of imaging helio seismic instruments can achieve the required performance. With the significant variation of the pmode frequencies through the eleven-year cycle of solar activity, it is now clear that the network should be run for eleven years to achieve its potential for studying the structure aud dynamics of the steady solar interior, as well as to study the activity cycle itself, and how it arises below the visible surface of the Sun. The initial three years’ data have been acquired with a 256 x 242 pixel CCD, which ia far from optimal for the science achieveable from the ground - for global and local helioseismology, as well as for surface measurements of flows and the magnetic field - and we are undertaking the replacement of the existing cameras with a 1024 x 1024 pixel CCD, which should be operational by mid-2000. This will eliminate a sign&ant spatial undersampling and difference in resolution in latitude and longitude in the current instrument, as well. The new system will also provide magnetograms every minute;which will be summed for five minutes to improve the signal-tonoise ratio and decrease the volume of data. Initially, we will be unable to reduce all of the data from the new camera system, which is known as GONG+, as it produces 32 times more data than the current DMAC is accomodating. At a minimum, after merging the velocity images, we will convolve the merged velocity time series with a smoothing kernel and sample it appropriately to produce a roughly 256 x 256 pixel image; that is, the same as we currently process. While producing the same basic scientific data products, they will be free of spatial a&sing, have the same resolution in both spatial dimensions, have continuous magetograms, and higher signal-to-noise ratio. In addition, the full spatial resolution data will be available for local helioseismology, and campaigns of full e analysis. We plan to provide full analysis capabilities shortly. ACKNOWLEDGMENTS The GONG project critically depends on the skills, talent, and dedication of a large number of persons at the observing sites around the world ( Big Bear Solar Observatory, High Altitude Obseratory, Learmonth Solar Observatory, Udaipur Solar Observatory, Instituto de Astro6sico de Canarias, and Cerro To1010 Interamerican Observatory ) and project staff in Tucson. The Project is managed by the National Solar Observatory, a Division of the National Optical Astronomy Observatories, which is operated by the Association of Universities for Research in Astronomy, Inc. under a cooperative agreement with the National Science Foundation. REFERENCES Harvey, J.W. & the GONG Instrument Development Team, 1988, in Proc. Symp Seismology Sun-like Stars, ed. E-J. Rolfe, ESA SP-286 203.
of the Sun and
Harvey, J.W., Hill, F., Kennedy, J., and Leibacher, J. 1993, in GONG 1992. Seismic Investigation Sun and Stars, ed. T.M. Brown, PASP Conf. Ser. 42, 397. Harvey, J.W. aud the GONG Instrument Team, 1995, in GONG ‘94: He&o- and Astero-Seismology Earth and Space, ed. R-K. Ulrich, E. J. Rhodes, and W. Dlippen, PASP Conf. Ser. 76,432. Hi,
of the from the
F., Fischer, G., Grier, J., Leibacher, J.W., Jones, H.P., Jones, P., Kupke R., aud Stebbins, R.T. 1994a, Solar Phys., 152, 321.
Hill, F., Fischer, G., Grier, J., Leibacher, J.W., Jones, H.P., Jones, P., Kupke R., Stebbins, R.T., Clay, D.W., Ingram, R.E.L., Libbrecht, K.G., Zirin H., Ulrich, R.K., Webster, L., Hieda, L.S., LaBonte, B.J., Lu, W.M. T. Sousa, E.M., Garcia, C.J., Yasukawa, E.A., Kennewell, J.A., Cole, D G., Zhen, H., Su-Min, X., Bhatnagar, A., Ambastha, A., Al-Khashlan, A.S., Abdul-Samad, M.-S., Benkhaldoun, Z., Kadiri, S., Sanchez, F., Palle, P.L., Duhalde, O., Solis, H., Saa, O., and Gonzalez, R. 1994b, Solar Phys., 152, 351. Leibacher, J.W. and the GONG Project Team 1995a, in GONG ‘94: He&o- and Astero-Seismology
J. W. Leibacher-and
the GONG Project Team
Earth and Space, ed. R.K. Ulrich, E.J. Rhodes, and W. Diippen, PASP Conf. Ser. 76, 381. Leibacher, J-W. and the GONG Project Team 1995b, in Fouth Soho Workshop: Helioseismology, ed. J.T. Hoeksema, V. Domingo, B. Fleck, and B. Battrick, ESA SP-376, Volume 1, 247. Pintar, J., and the GONG Data Team 1988, in Proc. Symp. Seismology of the Sun and Sun-Like Stars, ed. E.J. Rolfe, Paris: ESA SP-286, 217.