astro-ph0606193/nz.tex
1: %  article.tex (Version 2.81, released 24 September 2003)
2: %  taken from template (article) for SPIE Proceedings
3: %
4: %  060331 create
5: %  060407 version 1 = submit draft to all co-authors
6: %  060418 version 2 = update Epps comments
7: %  060419 incl. CaF system = version submitted on 060424
8: %  060427 updates for USNO Ed.Board comments
9: %  060508 insert reference 2nd curved spider paper -> re-submit SPIE
10: 
11: 
12: \documentclass[]{spie}  %>>> use for US letter paper
13: %%\documentclass[a4paper]{spie}  %>>> use this instead for A4 paper
14: %% \addtolength{\voffset}{9mm}   %>>> moves text field down
15: 
16: \usepackage[]{graphicx}
17: 
18: \title{URAT: astrometric requirements and design history} 
19: 
20: \author{Norbert Zacharias\supit{a}, Uwe Laux\supit{b},
21:         Andrew Rakich\supit{c}, and Harland Epps\supit{d}
22: \skiplinehalf
23: \supit{a}U.S.~Naval Observatory, 3450 Mass.~Ave.~NW, Washington DC 20392\\
24: \supit{b}Th\"uringer Landessternwarte, Sternwarte 5,  07778 Tautenburg, Germany\\
25: \supit{c}EOS Space Systems, 111 Canberra Ave, Griffith, ACT 2603, Australia\\
26: \supit{d}UC, Santa Cruz, Lick Observatory, Nat.Sc.2, Rm 191, Santa Cruz, CA 95064
27: }
28: 
29: \authorinfo{Further author information: (Send correspondence to N.Z.)\\
30:   N.Z.: E-mail: nz@usno.navy.mil, telephone: 1 202 762 1423\\
31:   U.L.: E-mail: laux@tls-tautenburg.de \\
32:   A.R.: E-mail: arakich@eos-aus.com \\
33:   H.E.: E-mail: epps@skewray.ucolick.org}
34:  
35: 
36: %>>>> uncomment following for page numbers
37: % \pagestyle{plain}    
38: %>>>> uncomment following to start page numbering at 301 
39: %\setcounter{page}{301} 
40:  
41:   \begin{document} 
42:   \maketitle 
43: 
44: \begin{abstract}
45: The U.S.~Naval Observatory Robotic Astrometric Telescope (URAT)
46: project aims at a highly accurate (5 mas), ground-based, all-sky
47: survey.
48: Requirements are presented for the optics and telescope for this
49: 0.85 m aperture, 4.5 degree diameter field-of-view, specialized
50: instrument, which are close to the capability of the industry.
51: The history of the design process is presented as well as
52: astrometric performance evaluations of the toleranced, 
53: optical design, with expected wavefront errors included.
54: \end{abstract}
55: 
56: \keywords{optics design, astrometry, wide-field, sky-survey, URAT, USNO}
57: 
58: 
59: \section{INTRODUCTION}
60: 
61: From 1998 to 2004 the U.S.~Naval Observatory (USNO) conducted an
62: astrometric, all-sky survey to 16th magnitude with its 20 cm aperture
63: Twin Astrograph, resulting in the USNO CCD Astrograph Catalog (UCAC),
64: with its second release (UCAC2) made public in 2003 \cite{ucac2}.
65: Around the year 2000 the USNO began preparations for a follow-up project,
66: the USNO Robotic Astrometric Telescope (URAT), a dedicated, 
67: astrometric, 1-meter size telescope with a wide field of view for
68: an all-sky survey going much deeper than UCAC and on a positional
69: accuracy level aiming at 5 milliarcsecond (mas) standard error per 
70: coordinate.
71: 
72: The idea of such a telescope goes back to the late 1980s \cite{cdv89},
73: then envisioned for photographic plates.
74: The purpose of such a telescope is threefold. 
75: First URAT aims at the densification of the
76: celestial reference frame, providing a large number of accurate reference 
77: stars \cite{a-surveys} to support, for example, general, deep sky surveys
78: like PanSTARRs and LSST, and to support astrometry in our solar system,
79: as outlined in the recent decadal survey \cite{decadal}.
80: Second, URAT provides absolute proper motions on an inertial system,
81: enabling significant galactic dynamics studies in the pre-Gaia era.
82: Third, URAT will be able to observe trigonometric parallaxes of
83: many thousands of stars unbiased by selection effects like
84: high proper motion targets.  More details about the URAT project 
85: in general can be found elsewhere \cite{potsdam,lowell1}.
86: 
87: 
88: \section{REQUIREMENTS}
89: 
90: \subsection{General} 
91: 
92: The basic parameters and requirements of the proposed new telescope
93: are summarized in Table 1.
94: 
95: \begin{table}[h]
96: \caption{Basic data of URAT.} 
97: \label{tab:basic}
98: \begin{center}       
99: \begin{tabular}{|l|rll|} 
100: %% use of \rule[]{}{} below opens up each row
101: \hline
102: \rule[-1ex]{0pt}{3.5ex}  aperture           & 0.85 & m & \\
103: \rule[-1ex]{0pt}{3.5ex}  effective aperture & 0.60 & m & equivalent unobstr.~area \\
104: \rule[-1ex]{0pt}{3.5ex}  focal length   & 3.60 & m & \\
105: \rule[-1ex]{0pt}{3.5ex}  image scale    & 57.3 & "/mm & = 57.3 mas/$\mu$m \\
106: \hline
107: \rule[-1ex]{0pt}{3.5ex}  field of view  & 4.00  & deg  & diameter design goal \\
108: \rule[-1ex]{0pt}{3.5ex}                 & 4.50  & deg  & usable (vignetted)\\ 
109: \rule[-1ex]{0pt}{3.5ex}  passband       & 650 $-$ 800  & nm& astrometric \\
110: \rule[-1ex]{0pt}{3.5ex}                 & 500 $-$ 950  & nm& photometric \\
111: \rule[-1ex]{0pt}{3.5ex}  stray light, ghost images & \multicolumn{3}{l|}
112:                    {10 magnitudes fainter than direct light / surface area}\\
113: \hline
114: \rule[-1ex]{0pt}{3.5ex}  focusing & \multicolumn{3}{l|}
115:                            {by moving the backend of the telescope} \\
116: \rule[-1ex]{0pt}{3.5ex}  telescope flip:& \multicolumn{3}{l|}
117:                    {2 telescope orientations, $180^{\circ}$ w.r.t.~sky} \\
118: \rule[-1ex]{0pt}{3.5ex}  mount & \multicolumn{3}{l|}
119:                            {equatorial} \\
120: \rule[-1ex]{0pt}{3.5ex}  guiding & \multicolumn{3}{l|}
121:                            {active, with 2 guide detectors in focal plane} \\
122: \rule[-1ex]{0pt}{3.5ex}  operation & \multicolumn{3}{l|}
123:                            {robotic} \\
124: \rule[-1ex]{0pt}{3.5ex}  detector  & \multicolumn{3}{l|}
125:                            {4 CCDs, each 10.6k $\times$ 10.6k pixel} \\
126: \hline
127: \end{tabular}
128: \end{center}
129: \end{table} 
130: 
131: The goal is to reach about magnitude 21 in a few minutes exposure
132: time and cover an entire hemisphere about 6 times per year
133: in a relatively narrow astrometric passband.
134: In order to cover a large amount of sky, critical sampling 
135: near 2 pixel per full width at half maximum (FWHM) is required. 
136: The limiting magnitude requirement leads to a 1-meter class aperture.
137: In order to ease the f/ratio requirement, a small pixel size 
138: (about 9 $\mu$m) was adopted.
139: These basic project parameters then determine the focal length
140: and field-of-view requirements.
141: 
142: The passband needs to be in the V to I range (peak detector quantum
143: efficiencies, large flux of ``normal" stars).
144: Exclusion of the $H_{\alpha}$ line is important to avoid 
145: complications with galactic nebulae; the emphasis is on stars.
146: The UCAC project chose a bandbass just blue of $H_{\alpha}$,
147: while for URAT a passband just red of $H_{\alpha}$ has been adopted.
148: This will mitigate the differential refraction effects, both from
149: the Earth's atmosphere and from refractive optics design considerations.
150: The width of the passband is chosen as a compromise between reaching
151: a deep limiting magnitude and minimizing differential color effects.
152: The optics design goal was set to cover 650 to 800 nm, while the
153: passband to be used will likely be about 670 to 770 nm.
154: 
155: \begin{table}[hb]
156: \caption{Specific astrometric requirements for URAT.} 
157: \label{tab:astrom}
158: \begin{center}       
159: \begin{tabular}{|l|l|} 
160: \hline
161: \rule[-1ex]{0pt}{3.5ex}  general:  & \\
162: \rule[-1ex]{0pt}{3.5ex}  good image quality  & $\ge$ 70 \% Strehl ratio \\
163: \rule[-1ex]{0pt}{3.5ex}  entrance pupil & circular symmetric, no spider structure \\
164: \rule[-1ex]{0pt}{3.5ex}  flat field      & no significant field curvature\\
165: \rule[-1ex]{0pt}{3.5ex}  stability       & variations of ``high order terms" \\
166: \rule[-1ex]{0pt}{3.5ex}                  & on $\le$ 100 nm level (thermal, flexure)\\
167: \hline
168: \rule[-1ex]{0pt}{3.5ex}  distortion: & \\
169: \rule[-1ex]{0pt}{3.5ex}  geom.~optical distortion  & $\le$ 1 arcsec / degree$^{3}$ \\
170: \rule[-1ex]{0pt}{3.5ex}  deviation from 3rd order dist.& $\le$ 300 nm over $4^{\circ}$
171:                                                                diameter field \\
172: \hline
173: \rule[-1ex]{0pt}{3.5ex}  lateral color: & \\
174: \rule[-1ex]{0pt}{3.5ex}  center-of-mass centroids & within 400 nm over astrom.~band \\
175: \rule[-1ex]{0pt}{3.5ex}  peak location  centroids & within 400 nm over astrom.~band \\
176: \hline
177: \rule[-1ex]{0pt}{3.5ex}  symmetric images: & \\
178: \rule[-1ex]{0pt}{3.5ex}  all PSFs + 0.7" seeing & centroids 90\% and 1\% within 400 nm \\
179: \hline
180: \end{tabular}
181: \end{center}
182: \end{table}
183: 
184: The ``flip" of the telescope is a required to perform astrometric
185: calibration observations.  The same area of the sky will be
186: observed with the entire telescope tube assembly plus camera
187: in 2 possible orientations which are rotated by $180^{\circ}$
188: around the optical axis with respect to each other. 
189: At the astrograph we achieved this by observing with the 
190: telescope from the East and West side of the pier (B\&C equatorial mount).
191: An equatorial fork mount would be an option for URAT if it
192: can be rotated by a sufficiently large angle and if the telescope
193: can swing through the ``fork" to both sides.
194: 
195: A requirement which entered the process at a late stage was
196: the ability of URAT to also perform photometric surveys in at
197: least 2 colors within the V to I range of the spectrum.
198: Color information is required for many applications to predict
199: the brightness of stars in non-standard passbands often used
200: in DoD instrumentation.
201: No strict astrometric performance is required for the 
202: 500 to 900 nm band; a general ``good" image quality, like
203: ``nearly diffraction limited" for the expected medium
204: seeing of 1 arcsec FWHM is sufficient for photometric
205: surveys, with re-focusing of the instrument between filters 
206: allowed.
207: The detector development is in progress\cite{sta} with a funded phase II
208: Small Business Innovation in Research (SBIR) program.
209: A complete prototype camera including a single 10.6k by 10.6k
210: thinned CCD is expected to be delivered to USNO by the end of 2006\cite{mike}.
211: 
212: 
213: \subsection{Astrometric Requirements}
214: 
215: The requirements specifically relevant for astometry are summarized
216: in Table 2.
217: The general image quality, which is typically defined by a Strehl ratio
218: is often the most important requirement for telescope designs, demanding
219: a nearly diffraction limited design over a specified field of view and passband.
220: For URAT this is of minor importance and generally ``sharp" images follow from
221: the other, more specific requirements for astrometric mapping.
222: 
223: Particularly problematic is the requirement of a circular symmetric
224: entrance pupil without secondary mirror spider structures which 
225: prohibited the use of most traditional optical designs.
226: The reason for this requirement is to be able to measure very bright
227: and very faint stars simultaneously.  It is assumed that the detector
228: for URAT does not ``bleed" charge when saturation is reached.
229: Accurate positions could then be obtained even from overexposed
230: stellar images by just rejecting the central, saturated pixels of the
231: image profile.  Conventional designs with diffraction spikes from 
232: secondary mirror support structures would prevent this.
233: Curved spiders were suggested \cite{spider1,spider2} which reduce the effect 
234: of ``spikes", but a rigorous approach was adopted here particularly for
235: DoD applications.
236: 
237: At first it seems that a general pattern of geometric optical distortion, 
238: even if very large, could be tolerated for an astrometric telescope as long
239: as the stability is guaranteed.  However, extensive calibrations would be
240: required to achieve satisfactory astrometric results from such a telescope,
241: which is rendered impossible in practice when considering the limited number
242: of reference stars available with typically unknown colors or considering
243: the number of ``bins" needed to map out the focal plane geometric distortions 
244: to within a specified, tight tolerance and bridge gradients between bins.
245: Thus a relatively small optical distortion is required to begin with,
246: which then will need to be calibrated to the measure accuracy level
247: with as few parameters as possible.  That is why only a 3rd order term
248: is allowed here with higher order deviations smaller than 300 nm 
249: in the focal plane, which corresponds to 17 mas in extreme cases 
250: (usually at the edge of the field), with much better performance on average.
251: 
252: Because of the generally unknown spectral energy distribution of the stars 
253: to be observed, the images of stars must have ``the same" centroid regardless
254: of wavelength, which leads to a strict lateral color requirement.
255: For all field points within the $4^{\circ}$ diameter field of view
256: and for all wavelengths in the astrometric passband (650--800 nm)
257: the maximal difference in image centroids of any 2 monochromatic 
258: point spread functions (PSF) shall not exceed 400 nm for a 90\% confidence
259: level of the ``as-built" system.\footnote{The wording ``as-built"
260: used in quotes is used in this paper for an optical system which includes
261: expected wave-front errors from manufacturing and alignment errors
262: as derived from simulations of the toleranced system.}
263: When dealing with such tight tolerances the ``centroid" needs to
264: be defined more precisely.
265: Due to possible small image profile asymmetries (coma aberrations)
266: there are different definitions of ``centroid".
267: Here the lateral color requirement needs to be obeyed by 2 definitions:
268: center-of-mass centroid of a PSF (taking the entire flux) and the
269: peak location of the PSF, as defined by the center-of-mass
270: position of the 90\% and above contour level of the PSF, after
271: folding with a 2-dimensional Gaussian seeing profile of 0.7 arcsec FWHM.
272: The first lateral color criterion could be checked by ZEMAX
273: directly, while the latter required a dedicated astrometric
274: evaluation (see below).
275: 
276: \begin{figure}
277: \begin{center}
278: \includegraphics[height=6cm]{layout_refr.eps}
279: \end{center}
280: \caption[ex1] {Optics layout of pure refractive solution.}
281: \end{figure} 
282: 
283: \begin{figure}
284: \begin{center}
285: \includegraphics[height=6cm]{layout_rs2b.eps}
286: \end{center}
287: \caption[ex2] {Optics layout of RS2b design.}
288: \end{figure} 
289: 
290: \begin{figure}
291: \begin{center}
292: \includegraphics[height=6cm]{layout_1caf.eps}
293: \end{center}
294: \caption[ex3] {Optics layout of 1CaF design.}
295: \end{figure} 
296: 
297: \begin{figure}
298: \begin{center}
299: \includegraphics[height=6cm]{layout_rs3.eps}
300: \end{center}
301: \caption[ex4] {Optics layout of RS3 design.}
302: \end{figure} 
303: 
304: \begin{figure}
305: \begin{center}
306: \includegraphics[height=6cm]{layout_fasa.eps}
307: \end{center}
308: \caption[ex5] {Optics layout of FASA2 v4 design.}
309: \end{figure} 
310: 
311: Similarly, the image asymmetry is bound to give a difference in centroid 
312: positions nowhere exceeding 400 nm for the ``as-built" system and folded
313: with 0.7 arcsec seeing when comparing the 90\% and 1\% flux level of any PSF
314: in the astrometric field-of-view and passband.
315: At 0.7 arcsec seeing the PSFs will be undersampled with about 1.5 pixels
316: per FWHM and ``pixel-phase" centroid bias effects will be taken into
317: consideration similarly to the UCAC reduction procedures.\cite{ucac2}
318: 
319: \begin{figure}[ht]
320: \begin{center}
321: \begin{tabular}{c}
322: \includegraphics[height=14cm]{rs3_1.ps}
323: \end{tabular}
324: \end{center}
325: \caption[example] {Spot diagrams of RS3 design.}
326: \end{figure} 
327: 
328: \begin{figure}[h]
329: \begin{center}
330: \begin{tabular}{c}
331: \includegraphics[height=14cm]{fasa_1.ps}
332: \end{tabular}
333: \end{center}
334: \caption[example] {Spot diagrams of FASA2 v4 design.}
335: \end{figure} 
336: 
337:  
338: \section{Design Options}
339: 
340: Design work specifically for URAT began around the year 2001.
341: Modified Richter-Slevogt systems were proposed by one of us (U.L.),
342: which were presented e.g.~at the Lowell astrometry meeting
343: \cite{lowell2}
344: followed by a single full-aperture lens approach from EOST (A.R.).
345: Initial contacts with various vendors seemed to indicate that
346: a 1-meter-class telescope even with ``corrector plates" is not
347: a problem. 
348: However, after having presented details (highly aspheric RS3 system 
349: for example) and the required tolerances, we quickly reached the 
350: state-of-the-art engineering limits with high price tags.
351: 
352: The 5 designs investigated are summarized in Table 3
353: with optical layout diagrams shown in Figures 1 to 5.
354: All but 1 design include only 2 optical elements close to the focal
355: plane: a filter and dewar window, which are given as ``+2" on
356: the ``number of elements" line in Table 3.
357: 
358: \begin{table}[h]
359: \caption{Comparison of initial optical design options for URAT.} 
360: \label{tab:comp}
361: \begin{center}       
362: \begin{tabular}{|l|ccccc|} 
363: \hline
364: \rule[-1ex]{0pt}{3.5ex}                    &  refractive  &  RS2b   &1CaF & RS3 & FASA2 \\
365: \hline
366: \rule[-1ex]{0pt}{3.5ex} number of elements &      4 + 2   &   4 + 2 &5 + 2& 4 + 2 & 6 + 2\\
367: \rule[-1ex]{0pt}{3.5ex} number of mirror aspheres&  0     &   1     & 2   &  2   &    1  \\
368: \rule[-1ex]{0pt}{3.5ex} number of lens   aspheres&  1     &   1     & 1   &  2   &    1  \\
369: \rule[-1ex]{0pt}{3.5ex} required glass     & special& fus.~sil.&CaF  &BK7, F2&fus.~sil.\\
370: \hline
371: \end{tabular}
372: \end{center}
373: \end{table}
374: 
375: Due to the large central obstruction of the catadioptric designs
376: (about 50\% in area) an all-refractive, classical astrograph lens
377: is an option, giving an equivalent throughput with an aperture of 
378: about 60 cm.  
379: The advantages and disadvantages of the individual design options are:
380: 
381: \begin{description}
382: \item[All-refractive.] PRO:
383:   \begin{itemize}
384:     \item no central obstruction
385:   \end{itemize}
386:    CON:
387:   \begin{itemize}
388:     \item narrow passband (670--750 nm)
389:     \item need folding to fit in existing domes, asymmetric mechanical setup
390:     \item availability of optical glass with desired specs is questionable
391:   \end{itemize}
392: 
393: \item[RS2b.] PRO:
394:   \begin{itemize}
395:     \item spherical primary mirror
396:     \item only fused silica for all refractive elements
397:   \end{itemize}
398:    CON:
399:   \begin{itemize}
400:     \item 2 full-aperture lenses with 1 aspheric surface
401:     \item image symmetry not meeting requirements (residual coma)
402:   \end{itemize}
403: 
404: \item[1CaF.] PRO:
405:   \begin{itemize}
406:     \item only 1 full-aperture lens, spherical
407:   \end{itemize}
408:    CON:
409:   \begin{itemize}
410:      \item issues with CaF lens (thermal and refractive gradients)
411:      \item limited field-of-view, color correction
412:   \end{itemize}
413: 
414: \item[RS3.] PRO:
415:   \begin{itemize}
416:     \item few elements, smallest amount of glass
417:   \end{itemize}
418:    CON:
419:   \begin{itemize}
420:     \item 2 full-aperture corrector plates 
421:     \item 2 highly aspheric surfaces on corrector plates
422:     \item F2 optical glass for corrector plate may be problematic
423:   \end{itemize}
424: 
425: \item[FASA2.] PRO:
426:   \begin{itemize}
427:     \item only fused silica for all refractive elements
428:     \item only 1 refractive asphere on sub-aperture lens 
429:     \item spherical secondary mirror
430:     \item nearly no geometric optical distortion at all
431:   \end{itemize}
432:    CON:
433:   \begin{itemize}
434:     \item number of surfaces is large, sub-aperture lenses double-pass
435:     \item largest central obstruction of presented design options
436:   \end{itemize}
437: \end{description}
438: 
439: 
440: \section{Selection of Design}
441: 
442: The all-refractive option was discarded when the requirement for
443: a photometric survey option arose.
444: The RS2b design has a residual amount of higher order coma which
445: gives rise to stellar image profile asymmetries exceeding the
446: strict requirements set for URAT.
447: The 1CaF design did not meet the later requirements for a larger
448: field of view, has larger color errors than RS3 or FASA2, and
449: the CaF lens is problematic due to the large gradients in
450: thermal expansion and refractive properties combined with the
451: alignment tolerances and stability requirements.
452: The RS3 and FASA2 designs have almost equivalent optical performance.
453: Spot diagrams of the ideal, as designed RS3 and FASA2 systems
454: are given in Figures 6 and 7, respectively.
455: 
456: However the RS3 design has 2 highly aspherical surfaces on the 
457: corrector plates and requires BK7 and F2 (or similar) optical glass.
458: By the end of 2005 the FASA2 design \cite{fasa2} was adopted as URAT
459: baseline due to easier manufacturability as compared to the RS3 
460: design.
461: 
462: \begin{figure}
463: \begin{center}
464: \includegraphics[width=6.0cm,height=15cm,angle=-90]{rs3_2.ps}
465: 
466: \includegraphics[width=6.0cm,height=15cm,angle=-90]{fasa_2.ps}
467: \end{center}
468: \caption[example] {Lateral color diagrams for RS3 (top) 
469:              and FASA2 (bottom) ideal designs.}
470: \end{figure} 
471: 
472: \begin{figure}[h]
473: \begin{center}
474: \includegraphics[height=17cm]{asb2_cont1.ps}
475: \end{center}
476: \caption[example] {``as-built" PSFs contour plots of FASA2 design.}
477: \end{figure} 
478: 
479: 
480: \section{Astrometric Performance}
481: 
482: \begin{table}[h]
483: \caption{Lateral color from ``as-built" FASA2 design PSFs 
484:          for sample points in the focal plane and extreme 
485:          monchromatic wavelengths (650$-$800 nm) as well
486:          as polychromatic (blue$-$red star, see text).
487:          Maximal position differences in $\mu$m are listed
488:          for original PSF and after folding with 0.8 arcsec 
489:          FWHM seeing, separately for the $x$ and $y$ axis.}
490: \label{tab:latcol}
491: \begin{center}       
492: \begin{tabular}{|l|cccc|cccc|} 
493: \hline
494: \rule[-1ex]{0pt}{3.5ex}        & \multicolumn{4}{c|}{monochromatic} 
495:                                & \multicolumn{4}{c|}{polychomatic} \\
496: \rule[-1ex]{0pt}{3.5ex} field  & \multicolumn{2}{c}{orig.~PSF}
497:                                & \multicolumn{2}{c|}{0.8" seeing}
498:                                & \multicolumn{2}{c}{orig.~PSF}
499:                                & \multicolumn{2}{c|}{0.8" seeing} \\
500: \rule[-1ex]{0pt}{3.5ex} point  &   $x$   &  $y$  &   $x$   &  $y$ 
501:                                &   $x$   &  $y$  &   $x$   &  $y$ \\ 
502: \hline
503: \rule[-1ex]{0pt}{3.5ex} on axis& 0.38& 0.54& 0.11& 0.16&  &  &  &  \\
504: \rule[-1ex]{0pt}{3.5ex}  -1.118& 0.03& 0.59& 0.20& 0.30&  &  &  &  \\
505: \rule[-1ex]{0pt}{3.5ex}  +1.118& 0.77& 0.49& 0.13& 0.01& 0.35& 0.23& 0.07& 0.01\\
506: \rule[-1ex]{0pt}{3.5ex}  -1.500& 0.17& 0.64& 0.20& 0.27&  &  &  &  \\
507: \rule[-1ex]{0pt}{3.5ex}  +1.500& 0.91& 0.49& 0.11& 0.01& 0.22& 0.15& 0.01& 0.01\\
508: \hline
509: \end{tabular}
510: \end{center}
511: \end{table}
512: 
513: Figure 8 shows the lateral color for the RS3 and FASA2 designs.
514: Both are acceptable for URAT.
515: 
516: For the FASA2 design ``as-built" point spread functions (PSFs)
517: were generated with the adopted tolerances and manufacturing
518: errors.
519: Figure 9 shows example contour plots of these PSFs at 2
520: worst-case locations in the focal plane (left, right), 
521: for 650 nm (top), 800 nm (middle) monochromatic, and 
522: flat-weighted polychromatic data (bottom).
523: The contour levels are 90\%, 70\%, 50\%, 30\%, 10\%, and 5\%.
524: 
525: After folding with a symmetric Gaussian function representing
526: realistic seeing conditions from 0.4 to 1.6 arcsec FWHM
527: the image profiles look even more symmetric.
528: For a quantitative analysis, image centroids were calculated
529: with the simple ``center-of-mass" algorithm cutting the
530: profile at various contour levels.
531: For lateral color investigations the maximal position differences
532: (in $\mu$m) are summarized in Table 4.
533: The PSF peaks are defined by the 90\% flux level and above.
534: On the left hand side the position differences for the
535: extreme, monochromatic colors (650 nm and 800 nm) are listed,
536: separately along the $x$ and $y$ axis (meridional and sagittal,
537: respectively). 
538: On the right side the differences between 2 
539: polychromatic PSFs are shown, with weighting
540: of 1,1,2,5 for 650, 700, 750 and 800 nm respectively
541: for a red star and weighting 5,2,1,1 for a blue star.
542: %\clearpage
543: 
544: \begin{table}[h]
545: \caption{Position differences ($\mu$m) between extreme (90\% and 1\%)
546:          contour levels for the ``as-built" FASA2 design PSFs.
547:          Data are shown for the original PSF and after folding
548:          with 0.4, 0.8, and 1.6 arcsec seeing, separately for the
549:          $x$ and $y$ axis.}
550: \label{tab:coma}
551: \begin{center}       
552: \begin{tabular}{|l|c|cccc|cccc|} 
553: \hline
554: \rule[-1ex]{0pt}{3.5ex} field& & \multicolumn{4}{c|}{along x-axis} 
555:                                & \multicolumn{4}{c|}{along y-axis} \\
556: \rule[-1ex]{0pt}{3.5ex} point&color  & orig & 0.4 & 0.8 & 1.6  
557:                                      & orig & 0.4 & 0.8 & 1.6  \\ 
558: \hline
559: \rule[-1ex]{0pt}{3.5ex} on axis& 650&0.58&0.44&0.22&0.07& 0.95&0.68&0.31&0.05\\ 
560: \rule[-1ex]{0pt}{3.5ex}        & 800&0.25&0.25&0.17&0.10& 0.40&0.40&0.23&0.09\\
561: \rule[-1ex]{0pt}{3.5ex}        &poly&0.47&0.35&0.19&0.08& 0.80&0.53&0.38&0.07\\
562: \hline
563: \rule[-1ex]{0pt}{3.5ex} -1.118 &650 &0.10&0.17&0.17&0.10& 0.96&0.78&0.41&0.06\\
564: \rule[-1ex]{0pt}{3.5ex}        &800 &0.13&0.11&0.07&0.10& 0.40&0.38&0.22&0.11\\
565: \rule[-1ex]{0pt}{3.5ex}        &poly&0.05&0.10&0.10&0.10& 0.71&0.55&0.28&0.07\\
566: \hline
567: \rule[-1ex]{0pt}{3.5ex} +1.118 &650 &1.29&0.73&0.27&0.09& 0.96&0.64&0.29&0.05\\
568: \rule[-1ex]{0pt}{3.5ex}        &800 &0.28&0.30&0.28&0.16& 0.48&0.47&0.30&0.11\\
569: \rule[-1ex]{0pt}{3.5ex}        &poly&0.84&0.54&0.27&0.11& 0.78&0.58&0.29&0.07\\
570: \rule[-1ex]{0pt}{3.5ex}        &polB&1.06&0.61&0.28&0.10& 0.88&0.60&0.30&0.06\\
571: \rule[-1ex]{0pt}{3.5ex}        &polR&0.62&0.41&0.28&0.14& 0.64&0.53&0.29&0.09\\
572: \hline
573: \rule[-1ex]{0pt}{3.5ex} -1.500 &650 &0.04&0.24&0.22&0.09& 1.05&0.88&0.47&0.06\\
574: \rule[-1ex]{0pt}{3.5ex}        &850 &0.29&0.23&0.12&0.10& 0.42&0.38&0.23&0.11\\
575: \rule[-1ex]{0pt}{3.5ex}        &poly&0.17&0.19&0.14&0.10& 0.72&0.58&0.30&0.08\\
576: \hline
577: \rule[-1ex]{0pt}{3.5ex} +1.500 &650 &1.25&0.63&0.23&0.09& 0.99&0.67&0.31&0.05\\
578: \rule[-1ex]{0pt}{3.5ex}        &850 &0.10&0.17&0.23&0.17& 0.48&0.52&0.32&0.11\\
579: \rule[-1ex]{0pt}{3.5ex}        &poly&0.73&0.43&0.24&0.12& 0.81&0.61&0.31&0.08\\
580: \rule[-1ex]{0pt}{3.5ex}        &polB&0.67&0.41&0.24&0.12& 0.72&0.61&0.32&0.07\\
581: \rule[-1ex]{0pt}{3.5ex}        &polR&0.48&0.30&0.23&0.16& 0.64&0.57&0.32&0.10\\
582: \hline
583: \end{tabular}
584: \end{center}
585: \end{table}
586: 
587: \clearpage
588: 
589: Image symmetry has been quantified similarly using ``center-of-mass"
590: image positions derived at extreme (90\% and 1\%) contour levels.
591: A position derived from the peak location of an image profile would
592: correspond to an astrometric observation of a faint star, while
593: for a bright star the entire profile starting at a low level above
594: the background would be used for a position fit.
595: For symmetric images there would be no difference in position.
596: This is even true for astigmatic images with different profile
597: widths along 2 axes.
598: For asymmetric image profiles (coma aberration) however there is such 
599: a ``magnitude equation", a systematic error in star position as a 
600: function of the brightness of the star.
601: The dominant source of coma here comes from alignment errors.
602: Thus alignment tolerances are very tight for astrometric instruments.
603: Image asymmetry is typically not tolerable on a level where the
604: the optical design would still give acceptable performance as
605: judged by the Strehl ratio alone.
606: 
607: Table 5 lists the position differences ($\mu$m) between extreme
608: contour levels of the same PSF.
609: These numbers have been obtained from the original ``as-built"
610: FASA2 design PSFs as well as after folding with 0.4, 0.8, and 1.6
611: arcsec seeing. 
612: Again, results are presented separately for the $x$ and $y$ coordinate
613: (meridional and sagittal, respectively).
614: For each field point, monochromatic as well as polychromatic PSFs
615: are analyzed, with ``poly" meaning a flat-weighted polychromatic
616: PSF while ``polyB" and ``polyR" are for the aforementioned
617: blue and red stars, respectively.
618: 
619: For 0.4 arcsec seeing, position offsets between peak and low-level
620: contour centroids exceed the requirement, while results for
621: 0.8 arcsec seeing and more are well below the 400 nm position
622: difference requirement.
623: Acceptable astrometric performance with respect to image
624: asymmetry is expected for about 0.7 arcsec seeing, which is sufficient
625: for the envisioned survey work of URAT.
626: 
627: Thermal and mechanical details of URAT are currently under
628: investigation.
629: A slight change of focus (scale) from exposure to exposure
630: is tolerable, however, higher order geometric distortions
631: as well as residual color and magnitude dependent systematic
632: errors need to be constant to high precision (0.1 $\mu$m level).
633: More details on the FASA2 design can be found elsewhere
634: in these proceedings \cite{fasa2}.
635: 
636: 
637: \begin{thebibliography}{99}
638: 
639: \bibitem{ucac2}
640:   N.~Zacharias et al.~``The Second U.S.~Naval Observatory CCD Astrograph 
641:     Catalog (UCAC2),'' {\em AJ} {\bf 127}, p.~3043, 2004
642: 
643: \bibitem{cdv89}
644:    Chr.~de Vegt, ``New type of astrometric telescope," 
645:     in {\em Star Catalogs: A centennial tribute
646:      to A.N.Vyssotsky}, eds.~A.G.D.~Philip \& A.R.~Upgren,
647:      L.~Davis Press, Schenectady, NY, p.~51, 1989
648: 
649: \bibitem{a-surveys}
650:   N.~Zacharais,
651:   ``Astrometric surveys in support of large telescopes,'' 
652:    in {\em title-proceedings}, T.~A. Tyson \& S.~Wolff eds., 
653:    {\em Proc. SPIE} {\bf 4836}, pp.~279, 2002
654: 
655: \bibitem{decadal}
656:   D.~Pascue et al., 
657:   ``Solar System Astrometry,''
658:    in {\em The Future of Solar System Exploration, 2003--2013},
659:    M.~V. Sykes ed., 
660:    {\em ASP Conf.~Series} {\bf 272}, pp.~361-374, 2002
661: 
662: \bibitem{potsdam}
663:   N.~Zacharias,
664:   ``Astrometric reference stars: from UCAC to URAT,''
665:   in {\em 3rd Potsdam Thinkshop on robotic telescopes},
666:   {\em AN} {\bf 325}, p.~631, 2004
667: 
668: \bibitem{lowell1}
669:   N.~Zacharias,
670:   ``The URAT Project,''
671:   in {\em Astrometry in the Age of the Next Generation of Large Telescopes},
672:    P.~K. Seidelmann \& A.~K.~B. Monet eds.,
673:    {\em ASP Conf.~Series} {\bf 338}, p.~98, 2005
674: 
675: \bibitem{sta}
676:   R.~Bredthauer, ``STA1600,"
677:    {\em technical data sheet}, Semiconductor Technology Associates (STA), 2006
678: 
679: \bibitem{mike}
680:   M.P.~Lesser \& D.B.~Ouellette, 
681:    ``Development of hybridized focal plane technologies,"
682:    in {\em High Energy, Optical, and Infrared Detectors for Astronomy II},
683:    SPIE, {\bf 6276}, 2006 (in press)
684: 
685: \bibitem{spider1}
686:   J.L.~Richter,
687:   ``Spider diffraction: a comparison of curved and straight legs,"
688:    {\em Appl.~Opt.} {\bf 23}, pp.~1907-1913, 1984
689: 
690: \bibitem{spider2}
691:   J.E.~Harvey \& C.~Ftaclas,
692:   ``Diffraction effects of telescope secondary mirror spiders on
693:     various image-quality criteria," 
694:    {\em Appl.~Opt.} {\bf 34}, pp.~6337-6349, 1995
695: 
696: \bibitem{lowell2}
697:   U.~Laux \& N.~Zacharias,
698:   ``URAT optical design options and astrometric performance,''
699:   in {\em Astrometry in the Age of the Next Generation of Large Telescopes},
700:    P.~K. Seidelmann \& A.~K.~B. Monet eds.,
701:    {\em ASP Conf.~Series} {\bf 338}, p.~106, 2005
702: 
703: \bibitem{VLZ}
704:   Chr.~de Vegt, U.~Laux, N.~Zacharias,
705:   ``A dedicated 1-meter telescope for high precision astrometric sky
706:    mapping of faint stars,''
707:   in {\em Small Telescopes in the New Millenium II.},
708:    T.~Oswalt ed., Kluwer Acad.~Publ. p.~255, 2003
709: 
710: \bibitem{fasa2}
711:   A.~Rakich,
712:  ``FASA2 optical design for URAT", SPIE, this volume, 2006
713: 
714: \end{thebibliography}
715: 
716: \end{document} 
717: