From Fringes to the USNO Navy Prototype Optical Interferometer Astrometric Catalog J. A. Benson'a, D. J. Huttera, R. T. Zavalaa, H. C. Harrisa, P. D. Shanklanda K. J. Johnstonb au. S. Naval Observatory, Flagstaff Station, 10391 W. Naval Observatory Rd., Flagstaff, AZ 86001; bU. S. Naval Observatory, 3450 Massachusetts Ave. NW, Washington, DC 20392-5420 ABSTRACT We report progress on the United States Naval Observatory, Navy Prototype Optical Interferometer, Astrometric Catalog (UNAC). This catalog uses observations from eight astrometric observation runs (Jan. 2005 - Nov. 2009) at the Navy Prototype Optical Interferometer (NPOI). The goal of the first release of the UNAC is to provide an astrometric catalog of at least 100 bright (V < 6) stars with precise positions accurate to < 16 milliarcseconds. In this paper we report on some of the data processing methods used to obtain absolute astrometric positions from optical interferometer data. We also discuss plans for assessing the accuracy of OUf interferometrically derived absolute astrometric positions. Keywords: Optical interferometry, astrometry, NPOI 1. INTRODUCTION The Navy Prototype Optical Interferometerl (NPOI), is a six element optical interferometer located on Anderson Mesa, AZ. The interferometer includes arrays that are used for imaging and for astrometry. The astrometric array consists of four 50-em siderostats that feed 12-cm beams into a vacuum beam relay system. The astrometric array includes baselines ranging from 19 m to 38 m. The astrometric stations have extensive station metrology: that includes a nearly end-to-end "constant-tenn" metrology system'. The NPOI has rapid tip-tilt star tracking and uses active group-delay fringe tracking over 16 wavelength channels ranging from - 550 - 850 nm. In this paper we describe progress with the NPOI towards obtaining a high precision astrometric catalog of a small number of bright stars. Since the NPOI, like all interferometers, is a pointed instrument rather than a wide field survey instrument, we have concentrated on obtaining high precision astrometric measurements of slightly more than 100 stars. In November 2009, we completed observations for the first release of the USNO Navy Prototype Optical Interferometer Astrometric Catalog (UNAC). The goal of this catalog is to produce precise, accurate « 16 milliarcseconds) astrometric positions for - 100 bright (V < 6) stars. 2. DATA AND DISCUSSION The path from fringes to precise and accurate astrometric positions for an optical interferometer is by no means trivial. At the astrometric baselines of the NPOI, one micron of baseline uncertainty corresponds to - 15 milliarcseconds (mas) of positional uncertainty on the sky. For comparison, the atmosphere inserts hundreds of microns of random delay into the predictable geometric delays. In addition, the interferometer stations themselves move around by hundreds to a few thousand microns during the course of a night of observations. These two effects conspire to make astrometric measurements much more challenging than a simple geometry problem. The NPOI is fully outfitted with internal metrology systems that measure the motions of the stations in the interferometer as well as motions in the vacuum feed system that serves to transport the light from the stations to the beam combiner. These data are used during the data reductions to properly account for the internal motions of the stations in the interferometer. *[email protected]; phone I 928779-5132 x261; fax 1 928774-3626; www.tlsno.navy.mil Optical and Infrared Interferometry II, edited by William C. Danchi, Fran90ise Delplancke, Jayadev K. Rajagopal, Proc. of SPIE Vol. 7734, 77343K· © 2010 SPIE· eee code: 0277·786X110/$18· doi: 10.1117/12.858244 Proc. of SPIE Vol. 7734 77343K-1 Downloaded from SprE DIgital Library on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spledl.orglterms Report Documentation Page Form Approved OMB No. 0704-0188 Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. 1. REPORT DATE 3. DATES COVERED JUL 2010 2. REPORT TYPE 00-00-2010 to 00-00-2010 4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER From Fringes To The USNO Navy Prototype Optical Interferometer 5b. GRANT NUMBER Astrometric Catalog 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION U.S. Naval Observatory,3450 Massachusetts Avenue, REPORT NUMBER N.W.,Washington,DC,20392-5420 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S) 11. SPONSOR/MONITOR’S REPORT NUMBER(S) 12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited 13. SUPPLEMENTARY NOTES Proceedings of the SPIE, Volume 7734, pp. 77343K-77343K-9 (2010) 14. ABSTRACT We report progress on the United States Naval Observatory, Navy Prototype Optical Interferometer, Astrometric Catalog (UNAC). This catalog uses observations from eight astrometric observation runs (Jan. 2005 - Nov. 2009) at the Navy Prototype Optical Interferometer (NPOI). The goal of the first release of the UNAC is to provide an astrometric catalog of at least 100 bright (V < 6) stars with precise positions accurate to < 16 milliarcseconds. In this paper we report on some of the data processing methods used to obtain absolute astrometric positions from optical interferometer data. We also discuss plans for assessing the accuracy of our interferometrically derived absolute astrometric positions. 15. SUBJECT TERMS 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF ABSTRACT OF PAGES RESPONSIBLE PERSON a. REPORT b. ABSTRACT c. THIS PAGE Same as 9 unclassified unclassified unclassified Report (SAR) Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18 The effects of the atmosphere are detennined by careful analysis of the curvature of the fringe phase across the 16 channel wavelength band. Note that the lower order linear phase dependence with wave number component is completely confounded with the inevitable imperfect fringe tracking. i.e. With our. group delay fringe tracker, if there is a slight error in the position of the delay cart with respect to the true zero optical path, then this imparts a linear slope to the phase of the fringes as a function of wave number. In addition to adding phase curvature, the atmosphere also adds varying phase offsets. This phase offset is however a N*2n wrapped quantity, where N can have values from zero to up to plus or minus several hundred, hence the atmospherically induced phase offset is not of much use. However, by detennining the degree of phase curvature, we are able to correct for the. random additional delay that the atmosphere adds. This gives us the "dispersion corrected" delays or in other words, the delays due only to geometry and any residual (over the measured) motion of the interferometer stations. We first use the dispersion corrected delays and internal motion measurements to solve for the interferometer station positions. We allow for up to a third order polynomial time dependence (due to thennal effects throughout the night) for the station locations. We then use Markov chain Monte Carl04 with simulated annealing to determine the star position offsets from their apparent topocentric calculated positions. In optical interferometery, the atmospherically induced phase changes of hundreds of2n wraps, makes it impossible to do phase-offset measurements from distant (e.g. quasars) sources with well known positions, as is often done with radio interferometers5.6. This means that for full astrometric accuracy, we must use the full topocentric calculations, including the fully interpolated UTI - UTC values throughout the night. We use the USNO NOVAS' routines for the stellar position calculations. We use code derived from interp.f', obtained from the International Earth Rotation & Reference Systems Service' (IERS) for interpolating UTI-UTC values between the daily tabulated values as provided by the IERS Rapid ServicelPrediction Center. 10 In the following series of plots and tables we show our current very preliminary distribution of UNAC offsets from the positions predicted from the Hipparcos catalogll• 12 (epoch 1991.25) updated to our UNAC epoch 2008. Within a few months we will be running our data through an improved data processing pipeline. Hence, we consider our current analysis very preliminary. At this preliminary stage, we are showing our results only in a statistical sense. We are not including the actual star designations. ~Or-________~ U~N~Ac~oe~I~~~~~~~~ __________, UNAC Delta Ra vs Ra 200 II 100 ! 100 " 'I I~ ~ • 'f I fit' il,l 1 f '1.1l ,,' • I I ! f> t !~ J "'/ ,,I ! §~~ -100O 'l', 'i f' ;II i~fU H, t II f ,lb. [ . • f, l " .' f II. \ lI'I u<0z~~ 0 1,' 1 'AI f "1 I I n -100 I -200 -< -200'--____" 7'--__- --;;;--'--__- .7_ ___- ,;;--____ -~~~----~-----"Oo-----"S~----'20~----~~ ~ 10 15 20 2S Ra (Hours) Ra (Hours) Figure 1. These panels show the distribution of the UNAC right ascension (Ra) offsets in mas from Hipparcos as a function of Ra. The right panel is a zoomed in view of the left panel. Figure 1 shows the right ascension (Ra) differences in mas between our observed UNAC positions and the Hipparcos predicted positions plotted as a function of the star's Ra. There does not appear to be any obvious dependence of the position offsets with Ra. If such dependence were apparent, this would be an indication of systematic errors. There are, however some clear "by eye outliers" (BEOs). Table 1 describes the largest BEOs in the left panel of figure 1. Proc. of SPIE Vol. 7734 77343K-2 Downloaded from SPIE DIgital Library on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spledl.org/terms Table \. Descriptions of the BEOs in Figure \. Delta Ra (mas) Ra (Hours) Comments -II I 0.8 Spectroscopic binary 195 3.2 Spectroscopic binary -114 5.1 Low Dec -214 5.9 Binary -142 7.7 Binary -198 I\.9 Seems normal 174 17.7 Binary Stars in binary systems are generally "difficult" astrometric candidates. They may have non-linear proper motion tenns that need to be accounted for. Depending on the parameters of the binary system (including its distance), a full orbital solution may also be needed to predict the positions. At any rate, the binaries require special treatment in our analysis. At present, we have not provided them with the needed individualized analysis. Hence, it is not surprising that the outliers are primarily the few binary systems that we have in our data set. 'oor-____- -"U"NA"C"'o"'.""ta"'-"Ra=~"O"'."C _____. .., UNAC Delta Ra vs Dec 200 , 200 . .. 100 [ u;- 100 , I ~ ~ •~ 0 ~ ,•uz -100 . -200 -200 . 20 .., 60 80 20 20 40 60 80 Ceo;: (Degrees) Dec (Degrees) Figure 2. These panels show the distribution of the UNAC Ra offsets in mas from Hipparcos as a function of declination (Dec). The right panel is a zoomed in view of the left panel. Figure 2 shows the Ra differences in mas between the observed UNAC positions and the Hipparcos predicted positions plotted as a function of the star's declination (Dec). Here, by eye, there is a slight indication ofa negative sloping trend. After we do the final processing that we discuss in Section 4, we will do rigorous statistical tests to investigate any possible trends in these types of plots. Table 2 describes the largest BEOs in the left panel of figure 2. Proc. of SPIE Vol. 7734 77343K-3 Downloaded from SPIE Oigltall.lbrary on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spicdl.orglterms Table 2. Descriptions of the BEOs in Figure 2. Delta Ra (mas) Dec (Degrees) Comments -116 -5.1 Low Dec +174 12.6 Binary -142 24.5 Binary -214 39.1 Binary +195 44.8 Binary -198 53.7 Seems nonnal As shown in Table 2, the BEOs are again dominated by the (same) binaries. '" ,-___- 'U"'NA"'C"O"'..''.''" o,,'''-''''-'" ''''-___- , UNAC Delta Dec vs Ra .,m 200 '" '" I !" E [ [ 100! f I. t "'t I."" II 11 I!I!~ I) I." ''\''1' .,1 -'0 """I,,H, • I I. t f··.,1I' . I If; f I. J. l. J -100f T " -200 -300 ~ (Hours) -200 '--'--~---;';;"--...,;1----;:;---i 0-- 10 15 20 Ra (Hours) Figure 3. These panels show the distribution of the UNAC Dec offsets in mas from Hipparcos as a function of Ra. The right panel is a zoomed in view of the left panel. Figure 3 shows the Dec differences in mas between the observed UNAC positions and the Hipparcos predicted positions plotted as a function of the star's Ra. Table 3 describes the largest BEOs in the left panel of figure 3. Table 3. Descriptions of the BEOs in Figure 3. Delta Dec (mas) Ra (Hours) Comments -336 2.5 Binary -200 3.2 Binary -174 13.4 Low Dec -199 15.2 Low Dec +440 23.6 Binary Proc. of SPIE Vol. 7734 77343K-4 Downloaded from SPIE Digllal LIbrary on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spfedl.orgllenns ---------- .._ . ...._ -_ .. Binaries are again prevalent in Table 3. UNAC Delta Dec vs Dec UNAC Delta Dec vs Dec SOO 200 400 ;00 100 I 200 I , ~~• 1000 l • , 10-I, 0· ~ OI J.'........ ojf; . I. •'' Itl • f .... '1 ~~ 0 It 't' ,'II ' I , '1111\. I 1I I I f I ' u<~ -100 ~: i. 1f I I , fJ ,~ ~; III -100 f [ -200 'I -300 ---;o:----;,,----;;;-'----;,,-~ -40020 0 20 40 • so -20,"_0 Ii o-_1L1 20 40 • so Dec (Degrees) Dec (Degrees) Figure 4. These panels show the distribution of the UNAC Dec offsets in mas from Hipparcos as a function of Dec. The right panel is a zoomed in view of the left panel. Figure 4 shows the Dec differences in mas between the observed UNAC positions and the Hipparcos predicted positions plotted as a function of the star's Dec. This plot strongly indicates that we have a systematic problem for stars below a Dec of - 0 degrees. Above zero Dec, there is no obvious systematic pattern. Table 4 describes the largest BEOs in the left panel of Figure 4. Table 4. Descriptions of the BEOs in the left panel of Figure 4. Delta Dec (mas) Dec (Degrees) Comments -200 44.8 Binary -336 67.4 Binary 440 77.6 Binary In this case all the BEOs are binaries. Note that all but one of the 7 binaries appear in more than one of the previous tables. 3, MEANS OF ASSESSING ACCURACY 3:1 Simple consistency check The zeroth order sanity check on the accuracy of our offsets is to see how they compare on average with the Hipparcos position and proper motion uncertainties propagated to the current epoch. We show this very preliminary comparison in Figure 5. Proe. of SPIE Vol. 7734 77343K-5 Downloaded from SPIE Dlgilal Library on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spledl.orglterms 100 ,--,--__- --'U"'N"'A"-C"O""lta"'-'0O""-'""""O,,',,'ta"-"'"''---___- --, 400 I, •0 • • + ~ -l~l'~OO,----3-,~O~--~O---~,~O---~IOO Delta Ra (mils) Delta Ra (mas) Figure 5. These panels show our positional offsets with respect to the calculated Hipparcos positions at the current epoch. The right panel is a zoomed in view of the left panel. The red dashed square on the right panel indicates the average (over our UNAC stars) 3-sigma limits for the Hipparcos positions and proper motions at the current epoch. At this very preliminary stage, our offsets appear roughly consistent with Hipparcos errors propagated to the current epoch. 3.2 Checks with other high precision accurate catalogs The FK613•14 catalog combined data from the "Fifth Fundamental Catalog" (FK5)15· 16 and the Hipparcos data in order to obtain improved proper motion precisions. The claimed improvements are factors of ~ 2-3 over the original Hipparcos results .. A complete reanalysis of the Hipparcos data has also been performed by van Leeuwen (Hipparcos, the new reduction) 17, With the two catalogs, we can therefore plot the difference in catalog predicted positions with respect to Hipparcos predicted positions for one catalog vs. the other and expect to see a high degree of correlation. Figure 6 shows this plot for our UNAC stars. While there is fairly good correlation, it is clear that there are some statistically significant differences between the catalogs. 3.3 Checks with Radio Interferometry astrometry Radio interferometry is capable of very accurate and precise astrometry. It was radio interferometry observations that defined (and are still improving) the International Celestial Reference Frame'" 19.20 (ICRF). Unfortunately most optically bright stars are not bright in the radio. Hence, there have been very few recent (epochs ~ 2000) radio interferometer astrometric measurements of optically bright stars. We do, however, have 4 UNAC stars that were within a set of 46 radio starss that have radio astrometric positions detennined in 2003. We are in the process of comparing our offsets with the radio interferometer measured offsets after updating to our epoch. Proc. of SPIE Vol. 7734 77343K-6 Downloaded from SPIE Digital Library on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spiedl.org/terms ------------------------ wo r----,,,,,,,,o""',,,rn,,",,=,,,,-""'='- "H","=""H"'!:!'"",,,-".c-H"'!:l'- 7< Ra Dlfferell«s: 11'l<6 -HIP] vs rHIP new· HIP ,,, -100 -20D_~,"'OOC---_",. ".- ---"c----"",,.O- --'''''00 -20 0 20 [HIP new -HIP] In Ra (rna.) [HIP new _ HIP) In Ra (mao) H wo c. _-,O",,,,,O"'"""~"""~"''-''''''''-'.c-",H'fl'" '''!.l!!H''''''''''''"''--11''''''-71 Dec Differences: FK6 -HIP vs HIP new -HIP + ,,, .. L20 + -100 -200_~,,,OO'----_"',"OO----'O;"---",,"'O- -~'OO -20 0 20 (HIP new_ HIP1 in De<: (mu) [HIP new -HlP] In DK (rna'l Figure 6. FK6 minus Hipparcos offsets are plotted vs. Hipparcos new reductions minus Hipparcos offsets in this figure. The top row shows the Ra offsets and the bottom row shows Dec. The right panels are zoomed in views of the left panels. The one-to-one correlation line is also shown. 4. CONCLUSIONS AND FUTURE PLANS We have completed the optical interferometer observations that are needed to make a - 100 bright star astrometric catalog. At present we have 116 stars that show astrometric precisions < 16 mas. These high precision stars are a subset of stars for which we have precisions of < 50 mas. The results that we show here are very preliminary. While we are encouraged that for positive declinations our current preliminary analysis does not show any obvious systematic effect, we certainly do not consider the checks we have done so far sufficient. We will soon be running all of our data through an improved data processing pipeline. Our first run through our revision 1 data processing pipeline was meant from the beginning to primarily give us upper limit precision estimates on our offsets as we continued to collect enough data to meet our precision goal. Our revision 2 pipeline (P2) is meant to primarily give us accurate results and secondarily, we expect that it will also improve our precisions slightly. For the P2 processing run, we have: Proc_ of SPIE Vol. 7734 77343K·7 Downloaded from SPIE DIgital library on 20 Dec 2010 to 132.250.22.10. Terms of Use: http://spiedl.orglterms 1. Incorporated the recently released (31 Dec. 2009) NOVAS 3.0'. This updated NOVAS version has several improvements that, at the levels important for our UNAC catalog, significantly affect the accuracy of calculations of absolute (not differential) stellar apparent positions. Hence, due to this update alone, we expect that our offsets will significantly change when we do the P2 processing run. 2. For the P2 processing we will incorporate a more robust method that ensures that we only use data (in a per baseline manuer) recorded during the time that the fringe tracker was truly locked on the fringe. This is necessary because during periods of bad seeing, we allow data to be recorded even if less than all baselines are simultaneously phased up. This data collection strategy is used because it allows us to collect possibly high signal-to-noise-ratio (SNR) baseline data even in the presence of other possibly much lower SNR baselines. 3. Finally, during our P2 processing we will do full simultaneous solutions that use all baselines present. ACKNOWLEDGEMENTS The Navy Prototype Optical Interferometer is a joint project of the U. S. Naval Observatory and the Naval Research Laboratory, in cooperation with Lowell Observatory, and is funded by the Oceanographer of the Navy and the Office of Naval Research. The authors would like to thank the NPOI observers David Allen, Jason Sanborn, Susan Strosahl and Ron Winuer. J.A.B. also gratefully acknowledges a number of very useful email discussions with Dr. George Kaplan (USNO) on various NOVAS topics, as well as more general astrometric topics. REFERENCES [I] Armstrong, J. T., Mozurkewich, D., Rickard, L. J., Hutter, D. J., Benson, J. A, Bowers, P. F., Elias II, N. M., Hummel, C. A., Johnston, K. J., Buscher, D. F., Clark III, J. H., Ha, L. Ling, L.-C., White, N. M. and Simon, R. S., "TheNavy Prototype Optical Interferometer," ApJ, 496, 550-571 (1998). [2] Hutter, D. J. and Elias II, N. M., "Array metrology system for an optical long-baseline interferometer," Proc. SPIE, 4838,1234-1245 (2003). [3] Benson, J. A., Hutter, D. J., Johnston, K. J., Zavala, R T., White, N. M., Pauls, T. A., Gilbreth, G. C., Armstrong, J. T. and Hindsley, R. B., "NPOI: Recent technology and science," Proc. SPIE, 5491,464-471 (2004). [4] Gilks, W. R, Richardson, S. and Spiegelhalter D. J.; [Markov chain Monte Carlo in practice], Chapman & Hall, London, (1996). [5] Boboltz, D. A., Fey, A L., Puatua, W. K., Zacharias, Z., Claussen, M. J., Johnston, K. J. and Gaume, R. A., "Very Large Array Plus Pie Town Astrometry Of 46 Radio Stars," AJ, 133,906-916 (2007). [6] Fey, A. L., Boboltz, D. A., Gaume, R. A and Johnston, K. J., "Merlin Astrometry of 11 Radio Stars," AJ, 131, 1084-1089 (2006). [7] Kaplan, G., Bangert, J., Bartlett, J., Puatua, W. and Monet, A., "User's Guide to NOVAS 3.0," USNO Circular 180 (2009). [8] ftp.hpiers.obspm.fr/eop-pc/models/intem.f [9] http://hpiers.obspm.fr [10] http://maia.usno.navy.mil [II] Perrymarm, M. A. C., et aI., "The Hipparcos Catalog," Astron. Astrophys., 323, 1.49-L52, (1997). [12] ESA, [The Hipparcos and Tycho Catalogs, Vol. 3], ESA, Noordwijk, (1997). [13] Wielen, R., Schwan, H., Dettbarn, C., Lenhardt, H., Jahreiiss, H. and Jiihrling, R., "Sixth Catalog of Fundamental Stars (FK6) Part I. Basic Fundamental Stars with Direct Solutions," Astronomisches Rechen-Institut, publication No. 35 (1999). [14] Wielen, R., Schwan, H., Dettbarn, C., Lenhardt, H., Jahreiiss, H. and Jiihrling, R., "Sixth Catalog of Fundamental Stars (FK6) Part III. Additional Fundamental Stars with Direct Solutions," Astronomisches Rechen-Institut publication No. 37 (1999). [15] Fricke, W., Schwan, H., Lederle, T., (in collaboration with Bastian, U., Bien, R., Burkhardt, G., du Mont, B., Hering, R., Jaehrling, R., Jahreiss, H., Roeser, S., Schwerdtfeger, H. M. and Walter, H. G.,) "Fifth Fundamental Catalogue (FK5) Part I," Astronomisches Rechen-Institut, publication No. 32 (1988). Proc. of SPIE Vol. 7734 77343K-B Downloaded from SPIE Digital LIbrary on 20 Oec 2010 to 132.250.22.10. Terms of Use: htlp:/Ispiedl.org/terms --- -------- [16] Fricke, W., Schwan, R., Corbin, T., (in collaboration with Bastian, U., Bien, R., Cole, C., Jackson, R., Jaehrling, R., Jahreiss, H., Lederle, T. and Roeser, S.,) "Fifth Fundamental Catalogue (FK5) Part 11," Astronomisches Rechen Institut, publication No. 33 (1991). [17]Van Leeuwen, F., [Ripparcos, the New Reduction of the Raw Data], Springer (2007). [18]Ma, C., Arias, E. F., Eubanks, T. M., Fey, A. L., Gontier, A. -M., Jacobs, C. S., Sovers, O. J., Archinal, B. A. and Charlot, P., "The International Celestial Reference Frame as Realized by Very Long Baseline Interferometry," AJ, 116,516-546 (1998). [19]Fey, A. L., Ma, C., Arias, E. F., Charlot, P., Feissel-Vemier, M., Gontier, A. -M., Jacobs, C. S., Li, J. and MacMillan, D. S., "The Second Extension of the International Celestial Reference Frame: ICRF-EXT. I," AJ, 127, 3587-3608 (2004). [20] Lanyi, G. E., Boboltz, D. A., Charlot, P., Fey, A. L., Fomalont, E. B., Geldzahler, B. J., Gordon, D., Jacobs, C. S., Ma, C., Naudet, C. J., Romney, J. D., Sovers, O. J. and Zhang, L. D., "The Celestial Reference Frame at·24 and 43 GRz.!. Astrometry," AJ, 139, 1695-1712 (2010). ProC. of SPIE Vol. 7734 77343K-9 Downloaded from SPIE Digital Library on 20 Oec 2010 to 132.250.22.10. Terms of Use: hltp:lfspiedl.orglterms