Monthly report for September 2001 by Christian A. Hummel: Hummel wrote a summary of events at the IAU working group meeting on optical interferometry in France and sent it to NPOI colleagues. Hummel analysed the geometry for delay corrections based on varying feed directions in combination with the off-center feed position and found that the correction does not depend on the mirror angle. This means that the internal constant term monitoring already includes feed direction changes and that the observed systematic errors must be related to tracking trade-offs between the narrow angle tracker and the siderostat. Hummel finished preliminary design of an observation planning tool for NPOI. Trip report: This is my report on the meeting of the Working Group on Optical/IR interferometry held August 30 and 31, 2001, at the OHP in France. Even though I took as many notes as I could, there will be some details missing in this report. I have embedded in this report my own personal comments. The home page for this meeting has this URL: http://olbin.jpl.nasa.gov/iau/ General impression: ------------------- This was a meeting worthy of its many well known interferometrists, including Labeyrie, McAlister, Ridgway, Lena, Quirrenbach, Haniff, and Townes. All major interferometers were represented, although JPL/ISC disappointed with their failure to send more than one representative. If there was one important belief among the participants worth mentioning it was that our field has only a window of some 5 years or so to proof it's worthy of more funding in the future. It is exactly for this purpose that coordination and exchange of ideas will play important roles, which is why Peter Lawson, the chairman of this working group, organized this meeting. Peter's goal is to perhaps raise the working group to a level of commission at the IAU at the 2003 meeting in Sydney. It would help, amongst other things, to provide a roadmap for NSF and the IAU on technical programs worth pursuing in the field of optical interferometry. The worry is that too many promises may have been made by our community for example regarding images of stars, but on only a few we have delivered, and on top of that stellar science that few "outsiders" regard as exciting. The Keck and VLTI interferometers must deliver results soon or the astronomical community will conclude that our programmes have little impact on what's really important these days. Notes on items of the agenda: Viewpoints of major players in this field: ------------------------------------------ NSF: McAlister pointed out some basic facts related to how NSF works and what this means for those of us who are seeking funding from NSF. His presentation essentially re-iterated points made by NSF's Wayne van Citters at the Dana Point meeting, i.e. that NSF provides roadmaps to advise congress, and that for big projects the lead time is some 10 years or so. For us this means that if we were to plan a major new interferometric facility, we would have to start early, and in particular, would not get on our way before the ALMA project (mm-interferometer in the Atacama desert) was operational. We would also have to define carefully what we could do from the ground versus what needs to go into space. Ultimately, our community has to gain more visibility again in the decadal report after the last one did not receive significant input from us and therefore had no specific recommendations for interferometric projects, except for infrared interferometry. ESO: Ballester started out on topics somewhat unrelated to what Peter has had in mind for his presentation, instead informing us about the service mode operation planned for the VLTI, the conditions required to be met for this mode, and the data flow concept. This is quite ambitious, since what this means is the implementation of interferometry for novice users who can walk home with calibrated visibilities. When Ballester finally came back to ESO interests in collaboration he mentioned observation preparation tools, OPTICON (an abstract concept I did not understand), and imaging and (narrow angle) astrometry. JMMC: Chelli reported on activities of the Jean-Marie Mariotti Center in France, which was formed to bundle the somewhat divergent french interferometric research groups. He identified for us areas of service to be provided by JMMC and persons responsible, e.g. observation preparations (Duvert), calibrator catalogs (Berio), modeling (Chelli), imaging (Thiebaud), and software environment (Valiron). He also provided time lines, by which the last item (imaging) would be done in 2004. I usually have problems believing ambitious predictions like this, especially when it turned out later that many people talking about imaging have never done this themselves and don't even understand the concept of self-calibration. However, all that Chelli wanted us to know is that the french forces have been unified in support of VLTI and by limiting the scope of JMMC to the french community initially they would not cross over into ESO territory. FRINGE: As became obvious quickly, the VLTI project has spawned a number of national efforts to staff and support centers of interferometric excellence ("excellence" as a goal I presume). This is of course a welcome development needed to make the most out of a superlative project like the VLTI (or Keck to make it more general), but one mustn't forget that these entities need to collaborate. FRINGE is the german equivalent to JMMC, and was presented by Leinert. NEVEC: Finally, Jaffe gave a report on the Dutch center of excellence, pointing out that they were the only ones with anything close to what amounts to contracts with ESO to provide interferometric services for the VLTI. In the following discussion on how to coordinate the European centers, it was quickly resolved that the most important concerted effort has to be lobbying of ESO to purchase two more auxilliary telescopes for the VLTI to bring to five the number of ATs. Three had been built, but five had been anticipated, and if no action is taken now, we might never get the remainder the the ATs. (The VLTI can accomodate up to 8 ATs.) Discussion points raised included the recommendation to develop a scientific vision and tie it to a technical development (interferometry of course), and planet searches were identified as one such possible vision. People warned us not to make promises in the field of cosmology, as if to remind us that before becoming too optimistic we should concentrate on the more down-to-earth problems as to how to make 3-telescope arrays work well (and there are still only two of them!). Ridgway sees our arrays limited in capabilities, but I opposed his viewpoint since I believe that many of our arrays are working nowhere near their design specs (see NPOI!). Antoine Labeyrie completed this discussion with a presentation on arrays as a viable alternative to the extremely large single-dish telescopes planned by ESO and Caltech (CELT). He envisions spherically shaped volcanic craters as sites for arrays of his hyper-telescope design ("densified pupil"), with blimps floating at or near the beam combination point above the crater. No comments were diplomatically offered at this point, but later in the programm Labeyrie got into more detail and I felt that he must indeed have a very deep understanding of optics. The common data format: ----------------------- This was a refreshing course on what we need to implement a common format for calibrated data. Tom Pauls and Bill Cotton led the presentation and discussion, and we were happy that apparently everybody realized the important benefit of such a concept for data exchange and were ready to subscribe to whatever a small task force would finalize by this year's end. Most details are already hammered out, though a brief but fierce discussion flared up about whether complex numbers and their imaginary and real errors, as well as their correlation, were sufficiently well represented in a format which implements only amplitudes and phases and their errors. With the help of Buscher we tried to convince Chelli that one of the axes of the uncertainty ellipse is always parallel to the phasor, unless one had a pathological case (which Buscher jokingly claimed needed pathological software to deal with too). Chelli didn't give up and we vowed to clear this up with him off-line. I felt that this is part of the work we needed to invest to have the community behind us on this. What we cannot afford is spending time on something which is not adopted by (almost) everybody. Common software: ---------------- A real washout, even though Quirrenbach thought I did a reasonably good job at leading this discussion. After Marson, Duvert and I coordinated our presentations the previous day, I introduced the topic as one which deserved attention due to the significant resources going into software development. I offered to discuss common and shared software as alternatives to consider for groups who wanted to coordinate their efforts. (The details were included in my previous mail to you on this topic.) Marson pointed out that aips++ is a mature software in the sense that it already provides environment and tools to start coding interferometric software. It was agreed that imaging was probably the number one reason to have common software, but no one really seemed to have thought about these aspects, and a culmination of this frustrating effort to get a sense of what people had in mind softwarewise was Brummelaars comment that the CHARA guys were waiting for the software to fall from the sky. This is obviously not what they are really planning to do, but in any case, I concluded that interest in common software was practically non-existent, and that for example the JMMC group was well on their way to develop GILDAS into their package for imaging, modeling, and observation preparation. I'm not surprised about this since, as I found out, there can be unexpected fundamental differences in preference for example on whether or not to use dynamic typing, i.e. i=0 defines an integer, i=0. defines a float. IDL, Python (chosen by COAST as an environment for their new softare), and glish use dynamic typing, whereas the GILDAS group strongly prefers defined variables assigments. Various comments were made related to establishing common libraries of functions, like the IDLASTRO library and the Numerical Recipes, and using meta-language (a concept I didn't understand). No action items were adopted. Science: -------- Haniff laid out scientific objectives defining technical requirements which in his view have to be met for any new interferometer project. The critical areas include AGN, YSO, and peculiar stars, like cepheids, lpv, symbiotics, WR, and BE. He claimed that it does not make sense to embark on any project which is not capable of addressing at least one of this objectives. I think I can agree with that. I pointed out that while pioneering observations of cepheids for example were exciting, their astrophysical impact was minimal due to the insufficient data quality. We must not stop at this point but improve the calibration. We need to get at least a, say 10 by 10 pixel image of a stellar surface to raise interest in what we are doing. The general consensus then became that we need to make the current interferometers work well, something which seems obvious but still is not pursued aggressively enough perhaps. Townes raised neglected issues as more atmospheric research (make use of LIDAR for seeing calibrations), and interferometry on spectral lines. Combination of methods was strongly encouraged. Calibrator catalog: ------------------- Several groups tried to sell their calibrator catalog, but luckily it was realized that actual measurements and a system to organize this information in the most useful way probably should be the first stab at providing a database available for the community. Especially the giant interferometers like Keck and VLTI, where every observing hour costs thousands of dollars cannot affort wasting time on surveying for calibrators. Groups were asked to send their lists to Peter to post on the OLBIN page. Imaging: -------- My favourite session, moderated by Quirrenbach, with a nice introduction by Buscher who is the only one who has anything near a workable program to derive images from squared visibilities and closure phases. Lannes, a french imaging guru, gave a highly mathematical viewpoint of the imaging problem, which might not have been that useful for some, but was most challenging for those of us who like to translate each others terms for the same thing. One example: what Lannes decribed as "critical image reconstruction modes" I believe are called "sidelobes" by VLBI folks and "point spread functions" by single dish astronomers. He likened the closure phase problem to the "nearest lattice point" problem in mathematics, at which point I lost him. Later however, it turned out that he never spend any time on our actual problem of not having complex visibilities. He mentioned parametric imaging which I pointed out as something I believed was a very important procedure well suited for us, since it combines the advantages of model fitting and imaging. It was decided that COAST and NPOI should come up with imaging algorithms and send each other fake data from a mutually unknown object for imaging (imaging "bake-off"). Hardware: --------- Various participants came forward with problems they are wrestling with and welcomed collaborations: Townes: heard of gold-alloy coating by Hepner which are very good Quirrenbach: thinks that Lawrence-Livermore makes best optical coatings Seneta (SUSI): has problems with far-field wavefront quality Townes: finds seeing layer 150 - 200 feet above ground, measure with pole and LIDAR Delplanck: infrared CCDs promised by companies, but then not manufactured Brummelaar: how to build beam combiners Hummel (NPOI): we need single-mode fibers Anonymous: chromatic aberrations causing astrometric errors Swain: analysis of seismic vibrations (Keck has BIG airconditioners) Summary: -------- Peter is moving to NASA headquarters in DC, and may rename OLBIN and move under a different umbrella; folks at ISC/JPL might object. He will look for ways to expand the functionality of OLBIN. We should have a presence at every meeting somehow related to interferometery. Quirrenbach will draft a letter to ESO signed by our SOC to support purchase of 2 more ATs. Meet again next year. Everone agreed that Peter did a wonderful job in bringing this meeting about and us together. Only time can tell how much we really accomplished.