Upload
anonymous-ec9pxm
View
217
Download
0
Embed Size (px)
Citation preview
8/18/2019 SAS 2014 Proceedings
1/252
Proceedings for the 33rd Annual Conference of the
Society for Astronomical Sciences
The Symposium on Telescope Science
Editors:Brian D. Warner
Robert K. BuchheimJerry L. Foote
Dale Mais
June 12-14, 2014Ontario, CA
A Joint M eeting of
The Society for
Astronomical Sciences
The American Association of
Variable Star Observers
The Center for Backyard
Astrophysics
8/18/2019 SAS 2014 Proceedings
2/252
DisclaimerThe acceptance of a paper for the SAS proceedings does not imply nor should it be inferred as an endorsement
by the Society for Astronomical Sciences of any product, service, method, or results mentioned in the paper.The opinons expressed are those of the authors and may not reflect those of the Society for Astronomical Sci-ences, its members, or symposium sponsors.
Published by the Society for Astronomical Sciences, Inc.Rancho Cucamonga, CA
First printing: May 2014
Photo Credits:
Front Cover: NGC 2024 (Flame Nebula) and B33 (Horsehead Nebula)Alson Wong, Center for Solar System Studies
Back Cover:Center for Solar System Studies (CS3) site, Landers, CA,Robert D. Stephens, Center for Solar System Studies
8/18/2019 SAS 2014 Proceedings
3/252
Table of Contents
Table of Contents
PREFACE I
CONFERENCE SPONSORS III
SYMPOSIUM SCHEDULE V
PRESENTATION PAPERS
A CROWD SOURCED LIGHT CURVE FOR SN 2014G 1 J. C. M ARTIN
OBSERVATIONS OF NOVAE FROM ROAD 7 F RANZ -J OSEF H AMBSCH
IMPACT OF OBSERVING PARAMETERS ON 17 NIGHTS WITH NOVA DEL 2013 13 W AYNE L. G REEN
R ECOVERING FROM THE CLASSICAL-NOVA DISASTER 17 J OSEPH P ATTERSON
PUSHING THE ENVELOPE: CCD FLAT FIELDING 23 J AMES V AIL
TOWARD MILLIMAGNITUDE PHOTOMETRIC CALIBRATION 29 E RIC DOSE
MINING YOUR OWN DATA 41 M AURICE C LARK
A STRATEGY FOR URBAN ASTRONOMICAL OBSERVATORY SITE PRESERVATION:
THE SOUTHERN ARIZONA EXAMPLE 45 E RIC R. C RAINE , B RIAN L. C RAINE , P ATRICK R. C RAINE , E RIN M. C RAINE , S COTT F OUTS
USING A CCD CAMERA FOR DOUBLE STAR ASTROMETRY 55
J OSEPH C ARRO
MEASURING DOUBLE STARS WITH A DOBSONIAN TELESCOPE
BY THE VIDEO DRIFT METHOD 59 R ICK W ASSON
8/18/2019 SAS 2014 Proceedings
4/252
Table of Contents
K ITT PEAK SPECKLE INTERFEROMETRY OF CLOSE VISUAL BINARY STARS 77 RUSSELL G ENET , D AVID ROWE , T HOMAS C. S MITH , A LEX T EICHE , R ICHARD H ARSHAW
D ANIEL W ALLACE , E RIC W EISE , E DWARD W ILEY , G RADY BOYCE , P ATRICK BOYCE , D ETRICK B RANSTON , K AYLA C HANEY , R. K ENT C LARK , C HRIS E STRADA , R EED E STRADA , T HOMAS F REY , W AYNE G REEN , N ATHALIE H AURBERG , G REG J ONES , J OHN K ENNEY ,
S HERI LOFTIN , I ZAK M C G IESON , R IKITA P ATEL , J OSH P LUMMER , J OHN R IDGELY ,
M ARK T RUEBLOOD , DON W ESTERGREN , P AUL W REN
ORION PROJECT: A PHOTOMETRY AND SPECTROSCOPY
PROJECT FOR SMALL OBSERVATORIES 93 J EFFREY L. H OPKINS
SIMPLIFIED COLOR PHOTOMETRY USING APASS DATA 105 N ICHOLAS DUNCKEL
DETECTING PROBLEMATIC OBSERVER OFFSETS IN SPARSE PHOTOMETRY 111 T OM C ALDERWOOD
SOFTWARE BASED SUPERNOVA R ECOGNITION 115 D R. S TEPHEN M. W ALTERS
SPECTRO-POLARIMETRY: ANOTHER NEW FRONTIER 125 J OHN M ENKE
A SURVEY OF CURRENT AMATEUR SPECTROSCOPIC
ACTIVITIES, CAPABILITIES AND TOOLS 131 T OM F IELD
SURVEYING FOR HISTORICAL SUPERNOVAE LIGHT ECHOES
IN THE MILKY WAY FIELD 139
D.L. W ELCH
THE Z CAMPAIGN: YEAR FIVE 143 M IKE S IMONSEN
A SEARCH FOR EXTREME HORIZONTAL BRANCH STARS
IN THE GENERAL FIELD POPULATION 147 DOUGLAS W ALKER , M ICHAEL A LBROW
HOW MANY R CORONAE BOREALIS STARS ARE THERE R EALLY? 159 G EOFFREY C. C LAYTON
8/18/2019 SAS 2014 Proceedings
5/252
Table of Contents
THE ASYNCHRONOUS POLAR V1432 AQUILAE
AND ITS PATH BACK TO SYNCHRONISM 163 D AVID BOYD , J OSEPH P ATTERSON , W ILLIAM A LLEN , G REG BOLT , M ICHEL BONNARDEAU , T UT AND J EANNIE C AMPBELL , D AVID C EJUDO , M ICHAEL C OOK , E NRIQUE DE M IGUEL , C LAIRE D ING , S HAWN DVORAK , J ERROLD L. F OOTE , ROBERT F RIED , F RANZ -J OSEF H AMBSCH , J ONATHAN K EMP ,
T HOMAS K RAJCI , B ERTO M ONARD , Y ENAL OGMEN , ROBERT R EA , G EORGE ROBERTS , D AVID S KILLMAN , DONN S TARKEY , J OSEPH U LOWETZ , H ELENA U THAS , S TAN W ALKER
GROUND-BASED EFFORTS TO SUPPORT A SPACE-BASED EXPERIMENT:
THE LATEST LADEE R ESULTS 177 B RIAN C UDNIK , M AHMUDUR R AHMAN
METHODS FOR DISTANCE DETERMINATION VIA DIURNAL PARALLAX 183 M ICHAEL G ERHARDT
AN EXPERIMENT IN PHOTOMETRIC DATA R EDUCTION
OF R APID CADENCE FLARE SEARCH DATA 191 G ARY A. V ANDER H AAGEN , L ARRY E. OWINGS
THE STRANGE CASE OF GSC 05206:01013 201 M AHFUZ K RUENG , M AURICE C LARK
MODERN V PHOTOMETRY OF THE ECLIPSING TRIPLE SYSTEM B PERSEI 205 DONALD F. C OLLINS , J ASON S ANBORN , ROBERT T. Z AVALA
PAPERS WITHOUT PRESENTATION AND POSTERS
SKYGLOWNET SKY BRIGHTNESS METER (ISBM) NODES:
CERRITOS OBSERVATORY STATION, TUCSON, AZ, AND
COLORADO STATE UNIVERSITY, FORT COLLINS, CO 215 ROGER B. C ULVER , E RIN M. C RAINE , H EATHER M ICHALAK
UNDERGRADUATE OBSERVATIONS OF SEPARATION AND POSITION ANGLE
OF DOUBLE STARS ARY 6 AD AND ARY 6 AE AT MANZANITA OBSERVATORY 225 M ICHAEL J. H OFFERT , E RIC W EISE , J ENNA C LOW , J ACQUELYN H IRZEL , B RETT L EEDER , S COTT M OLYNEUX , N ICHOLAS S CUTTI , S ARAH S PARTALIS , C OREY T OKUHARA
GOT SCOPE? THE BENEFITS OF VISUAL TELESCOPIC OBSERVING
IN THE COLLEGE CLASSROOM 233 K RISTINE L ARSEN
8/18/2019 SAS 2014 Proceedings
6/252
Table of Contents
8/18/2019 SAS 2014 Proceedings
7/252
Preface
I
Preface
Welcome to the 33rd annual SAS Symposium on Telescope Science. This meeting is a “triple conjunction” inthat it involves not only the Society for Astronomical Sciences (SAS), but the American Association of Varia-
ble Star Observers (AAVSO) and the Center for Backyard Astrophysics (CBA). This year’s symposium offers
the rare opportunity of bringing together in one location at the same time some of the more accomplished ama-teur researchers from around the world. SAS is honored by the presence of members from these two organiza-tions and extends a warm welcome to all.
It takes many people to have a successful conference, starting with the Conference Committee. This year theregular committee members are:
Lee Snyder Robert D. StephensRobert Gill Jerry L. FooteCindy Foote Margaret MillerBrian D. Warner Robert K. BuchheimDale Mais
There are many others involved in a successful conference. The editors take time to note the many volunteerswho put in considerable time and resources. We also thank the staff and management of the Ontario AirportHotel, Ontario, CA, for their efforts at accommodating the Society and our activities.
Membership dues alone do not fully cover the costs of the Society and annual conference. We owe a great debtof gratitude to our corporate sponsors: Sky and Telescope, Woodland Hills Camera and Telescopes (Tele-scopes.net), PlaneWave Instruments, Apogee Instruments, Inc., and DC-3 Dreams. Thank you!
Finally, there would be no conference without our speakers and poster presenters and those sitting attentively inthe audience. We thank them for making this one of the premiere pro-am events in the world.
Brian D. WarnerRobert K. BuchheimJerry L. FooteDale Mais
2014 May 12
8/18/2019 SAS 2014 Proceedings
8/252
II
8/18/2019 SAS 2014 Proceedings
9/252
Conference Sponsors
III
Conference Sponsors
The conference organizers thank the following companies for their significant contributions and financial sup- port. Without them, this conference would not be possible.
Apogee Imaging SystemsManufacturers of astronomical and scientific imaging camerashttp://www.ccd.com
Sky Publishing CorporationPublishers of Sky and Telescope Magazinehttp://skyandtelescope.com
Woodland Hills Camera & TelescopesTelescopes.NETProviding the best prices in astronomical products for more than50 yearshttp://www.telescopes.net
PlaneWave InstrumentsMakers of the CDK line of telescopeshttp://www.planewaveinstruments.com:80/index.php
DC-3 Dreams SoftwareDevelopers of ACP Observatory Control Softwarehttp://www.dc3.com/
8/18/2019 SAS 2014 Proceedings
10/252
IV
8/18/2019 SAS 2014 Proceedings
11/252
Symposium Schedule
V
Symposium Schedule
8/18/2019 SAS 2014 Proceedings
12/252
Symposium Schedule
VI
Meeting Courtesies
Please turn of f cell phones, pagers, and alarms dur ing presentations.
Please be kind to the presenters and wait until breaks to check your email and browse the In-ternet. Yeah, we know.
Please be considerate of those in the meeting room near the wall with the vendor’s room. Ifvisiting the vendor’s room during presentations, please keep voices down and do not use thedirect entrance between the vendor’s and meeting rooms. Enter the meeting room from thehallway.
If you do not want your presentation to be recorded (video/audio) and/or not posted on theSAS web site, please let us know before your talk begins.
Thanks!SAS Program Committee
8/18/2019 SAS 2014 Proceedings
13/252
Presentation Papers
8/18/2019 SAS 2014 Proceedings
14/252
8/18/2019 SAS 2014 Proceedings
15/252
Martin: Crowd Sourced Lightcurve for SN 2014G
1
A Crowd Sourced Light Curve for SN 2014G J. C. Martin
University of Illinois Springfield
One University Plaza MS HSB 314, Springfield, IL 62704
Abstract
SN 2014G was initially classified as a Type IIn (CBET 3787) and was later revealed to be a Type II-L (ATEL5935). In addition to having an interesting classification, it was also relatively bright, nearby (peak V ~ 14.3) andeasy to observe with small to moderate sized telescope. We mounted a cooperative effort open to both profes-sional and non-professional observers with the goal of producing a light curve that could accurately measurevariations in brightness of 0.1 mag with a cadence of one every two days or better. Simply collecting measuredmagnitudes often results in a light curve with systematic offsets between independent contributors. To minimizethat effect without burdening the volunteer observers with too many additional requirements, we collected cali-brated images and processed them uniformly to produce the light curve.
1. Introduction
SN 2014G in galaxy NGC 3448 was initiallydiscovered by Koichi Itagaki in an unfiltered CCDimage taken on January 14.574 UT. Subsequent opti-cal spectra revealed a blue continuum with narrowemission lines (Denisenko et al, 2014) resulted in aType IIn classification for SN 2014G. Spectra takenseveral weeks later were consistent with a more typi-cal Type II supernova (Eenmäe et al. 2014). The tran-sition from a blue continuous spectrum with few fea-tures to a classic Type II spectrum is uncommon forType IIn but more typical of a Type II-L (see SN1979C, Branch et al. 1991). However, the lightcurve
itself looks more like a Type II-P than a Type II-L.The classification of this supernova is still uncertain.
Type IIn supernovae are associated with super-nova impostors and complex circumstellar environ-ments. Type II-L is one of the rarest classification ofsupernovae and few have been studied in detail. Thetransition in the spectrum itself was intriguing. To-gether these made SN 2014G a relatively interestingtarget. In 2012, we had observed unusual fluctuationsin the light curve of the 2012-B outburst of SN2009ip (Martin et al. 2014). That event may have
been a terminal Type IIn explosion (Prieto et al.2013; Mauerhan et al. 2013; Smith et al. 2013) or
some other non-terminal event (Pastorello et al. 2013;Fraser et al. 2013; Soker & Kashi 2013, Margutti etal 2014). SN 2014G was another bright, relativelynearby supernovae with unusual classification. Thegoal of our campaign was to observe a well-sampledlight curve of SN 2014G that would be sensitive tothe variations similar to those observed in the lasteruption of SN 2009ip (i.e. a few tenths of a magni-tude on timescales of five to ten days or more).
SN 2014G also had competition for limited ob-serving resources because in January 2014 SN 2014Gwas shortly followed by SN 2014J in M82, a nearbyType Ia with potential to help calibrate the cosmicdistance scale. As expected, SN 2014J attracted agood deal of the finite attention and resources devot-ed to following supernovae.
We did not have the resources ourselves to con-duct a successful campaign for SN 2014G. Theweather in central Illinois makes it difficult to obtainimages at a regular cadence of once every coupledays so it was unlikely the Barber Observatory 20-inch telescope could conduct a successful campaignon its own. We had some success enlisting the help of
non-professionals for the SN 2009ip 2012-B outburstcampaign in the southern sky so we considered asimilar campaign for SN 2014G. SN 2014G was bet-ter placed for northern-hemisphere observers (at dec-lination of +54 degrees) and bright enough to getimages with good signal-to-noise in a small or medi-um sized telescope. Its position in the night sky at thetime of discovery ensured we would be able to followit more than 150 days until solar conjunction. We
posted an announcement on the AAVSO “Campaigns& Observation Reports” electronic bulletin board onJanuary 19 and with the help of Matt Templeton atthe AAVSO reached out to those who had already
reported observations to the AAVSO InternationalDatabase. Within a week we had recruited a team ofseven observers plus the Barber Observatory to help
produce a light curve for SN 2014G.
8/18/2019 SAS 2014 Proceedings
16/252
Martin: Crowd Sourced Lightcurve for SN 2014
2
2. The Effort
2.1 Motivation and Philosophy
Calibrating observations from several differenttelescope and camera setups to a single photometric
system is a challenge even for stars with spectral dis-tributions similar to black bodies. Supernovae withmore complex spectra, present additional difficulty
because different system responses may include orexclude broad variable emission features. Munari etal. (2013a) outline a lightcurve merging method(LMM) for supernovae that utilize a least squaresapproach to determine the corrections for systemsrelative to each other. Their method relies on obser-vations made in multiple filers with each tele-scope/camera combination for which standard pho-tometric transformations have been previously de-termined (Munari & Morretti, 2012). They have suc-
cessfully applied their method with good results tolightcurves of both supernovae and novae (Munari etal., 2013b). Our approach is modeled on theirs butwith a focus on less stringent requirements on theobservers. In addition to getting a lightcurve for SN2014G we wanted to experiment with ways to in-crease productive non-professional participation inthis kind of work.
In the case of SN 2014G, attention had to beshared with the nearby Type Ia SN 2014J. Evenwithout competition for time and resources it would
be difficult to coordinate a SNe lightcurve for SN2014G with high temporal cadence using only pro-fessional telescopes. Non-professionals represent an
additional untapped resource well suited for crowdsourcing observations.
To encourage maximum voluntary participationin the project we considered the bare minimum wecould require from observers to achieve good results.
Non-professionals often have limited time and re-sources devoted to astronomy. Rigorous requirementscan discourage them. The professional world awardstelescope time through a comprehensive peer review
process. Non-professionals pick targets based on per-sonal interest and the positive feedback they receiveon their work. We were optimistic that we could at-tract good observers by adopting a lower threshold
for participation, offering friendly feedback andcoaching to help each make a high quality contribu-tion, and finding ways to reward them including:
praise, recognition, and credit for their contributions.We also made a conscious effort to build a coop-
erative spirit and community in the group. New vol-unteers were formally introduced to the team as they
joined and each team member had the contact infor-mation for other team members. The observers were
encouraged to discuss their observing plans with thegroup to avoid overlap in effort. Most of the commu-nication with observers that did not include individu-al advice were conducted as group emails that in-cluded everyone in the discussion. We maintained aweb page for the project at
https://edocs.uis.edu/jmart5/www/barber/
sn2014g.html
that was updated regularly, including a list of obser-vations that had been recently added to the plots sothat participants (and those who might be consider
joining our effort) could track our overall progress.
2.2 Requirements and Expectations
Our expectations for non-professionals were notlower. They were different and considerate of theirexperience and resources. To decide what we shouldadopt as standards, we critically approached our ownexpectations from their point of view. Professionalstypically have multiple standard filters with a highgrade CCD (usually back-thinned with good blueresponse) and have made an effort to determine thetransformation coefficients for their system. Non-
professionals more typically have one or two stand-ard filters, a less efficient front-illuminated CCDcamera, and are less likely to have considered thetransformation coefficients for their system. In ourexperience most non-professionals are (like manyhumans in general) eager to learn, generally willingto try new things, and curious about optimizing the
system they have. They might try something out and,like any independent operator, just decide it is not forthem. We reminded ourselves frequently that thesewere volunteers and made a point to never take it
personally when they decided to focus on somethingelse. One of the benefits of crowd sourcing is that theeffort is shared so that individuals are free to devotetime when they are available without too much im-
pact on the collective effort. Like professionals, posi-tive experiences encourage non-professionals to in-vest more time, effort, and resources in scientific
projects that have measurable benefit to the largercommunity.
For the SN 2014G project we adopted the fol-lowing requirements:
• Image quality must be good and imagesmust be properly dark subtracted, flat-fielded, and delivered in FITS format.
• The time and date and circumstances of theimages must be reported with reasonable ac-curacy.
8/18/2019 SAS 2014 Proceedings
17/252
Martin: Crowd Sourced Lightcurve for SN 2014
3
• Exposure times must be sufficient to getgood signal to noise in the target and com-
parison stars.
• Images must be taken with Johnson V andother standard filters.
• At some point during the project high quali-ty images must be obtained of M67 in eachof the standard filters that were used to im-age SN 2014G.
While supernovae lightcurves are more typicallyobserved in Johnson B, the most widely availablestandard filter is Johnson V. The throughput for theB-band can also be low when combined with the re-sponse curve of more commonly available font-illuminated CCDs. With this in mind, we required allobservers to contribute V images and made that bandour top priority. In order after that we prioritizedJohnson B, Cousins I, and Cousins R, which is con-
sistent with the advice most commonly given by theAAVSO (American Association of Variable StarObservers).
We did not collect unfiltered CCD images. Theresponse of different CCDs can be quantitativelymuch more over the full range of atmospheric trans-
parency than it is over the well-defined range of astandard broadband filter. We already accepted sev-eral unrestricted variables in our experiment and
postponed tackling the issues of unfiltered CCD im-ages for another time.
We did not require observers to add standardkeywords to their FITS images. We required that the
time and circumstance of the observations were rec-orded and delivered to us with the images. We of-fered friendly advice and feedback about what wouldmake post-processing easier. However, in our experi-ence it is possible to make good images without un-derstanding FITS headers and insisting on them couldcost us more than we would gain by requiring themfor this project.
The images of M67 were required as a standardfield and were used to calculate standard photometrictransformations for each telescope/camera system.While those transformations do not fully correct thesystem’s magnitudes for a non-blackbody supernovaspectrum, Munari et al (2013a) recommend it as anecessary first step in their LMM method. The pro-
ject fortuitously coincided with a time of year whenM67 was favorably positioned in the mid to lateevening for observations. We also learned that manynon-professionals are eager to learn more aboutstandard transformation coefficients and were grate-ful to have that done for their telescope/camera sys-tem.
Three of our observers used the iTelescope ser-vice (http://www.itelescope.net ) to obtain the imagesthey contributed. iTelescope determines the standardtransformation coefficients for each of their systemsand publishes them so we waived the requirement ofstandard exposures of M67 for images taken usingthose telescopes.
2.3 Data Analysis
Table 1 is a roster of the observers participatingin the SN 2014G project along with a brief descrip-tion of their telescope/camera system and the filtersthey used to observe SN 2014G.
Each telescope/camera combination had a differ-ent field of view and pixel scale. Usually we dealwith the personalities of only one telescope/camerasystem for a project. Including images taken with tendifferent systems complicated efforts to uniformlyextract photometry. During this project we developeda working understanding the point spread functionsand unique personality traits of each equipment set-up. We frequently relied on the operators/owners ofthe systems to share their experience optimizing theirown setup. With that knowledge we developed amethod to uniformly select optimal and consistentapertures and sky annuli across our entire image set.
The photometry used eleven comparison starsranging in V magnitude from 12.1 to 16.4 that wereselected by the AAVSO sequence team according totheir published guidelines:
https://sites.google.com/site/aavsosequenceteam/
sequence-creation-and-revision-guidelines
The field of view for some of the telescopes ex-cluded a number of comparisons so in those cases asubset was used to calculate the brightness of thetarget. As part of our data processing, we will simu-late the effect of using a subset of the comparisons onthe final results and calculate a correction for thosecases.
We processed the images of M67 and determinedthe necessary standard transformation coefficients foreach system. After applying those and the other cor-rections we will perform the LMM method of Munari
et al. (2013a). At this point we have only preliminarydata. We are hopeful that after the full analysis iscomplete we will meet or exceed our goals for preci-sion in the lightcurve of SN 2014G.
2.4 Crediting Participants and Use of Data
Another consideration was proprietary used ofthe contributed data and images. In our previous ex-
perience, many non-professionals had been unfamil-
8/18/2019 SAS 2014 Proceedings
18/252
Martin: Crowd Sourced Lightcurve for SN 2014
4
iar with the concept of proprietary data and avoided participating in projects that put restrictions on whatthey do with their own observations. Professionalsand non-professionals are rightfully concerned about
earning credit for their own work. Giving someoneelse control over how the fruits of your labor are dis-tributed and credited requires a degree of trust thatcan give one pause when they consider joining aneffort. To address this concern and build trust, weadopted a policy that clearly stated each observer still“owned” the images and data that they contributed tothe project. They were told that they could measuretheir own photometry and submit it to the Interna-tional AAVSO Database (IAD) or any other forum atany time under their own name. We promised that allcontributors would be acknowledged and/or includedas coauthors on any publication using any part oftheir contribution. We also offered at the conclusionof the project to return the fully reduced photometryin a formatted file that could be easily uploaded tothe IAD under their unique user ID.
3. Conclusions
As of the beginning of May 2014, we have over100 days of data sampled at a frequency of about twoimages every two days for the first 80 days and less
frequently since. At late times the diminished bright-ness of the target has put it beyond the practical reachof all but a few of our volunteers. The preliminarylightcurve published on the webpage for the project
looks promising. There are hints of a few interestingfeatures, but nothing like the repeated bumps thatwere observed in the 2012-B outburst of SN 2009ip.We are hopeful that when we complete our post-
processing that we will meet or exceed all the goalswe had for the lightcurve of SN 2014G.
4. Acknowledgements
This work is supported in part by work is sup- ported by the National Science Foundation grantAST1108890 and also by the University of IllinoisSpringfield Henry R. Barber Astronomy Endowment.
We are grateful to the volunteers who participated inthe SN 2014G project: Douglas Barrett, Alberto SilvaBetzler, Andy Cason, Tõnis Eenmäe, RaymondKneip, and Massimiliano Martignoni. We are alsograteful to the community volunteers and students atthe UIS Barber Observatory who assisted with this
project including: Jim O’Brien, Jennifer Hubbell-Thomas, Justin Mock, Shelby Jarrett, William “Pat-rick” Rolens, John Lord, and Kevin Cranford. Andwe thank the AAVSO sequence team for quickly
Name Location Telescope Filters
Douglas BarrettSt. Leger
Bridereix, France
Altair-Astro 8"
Ritchey-Chretien +
SBIG ST7ME
V
Alberto Silva Betzler
Universidade
Federal do
Reconcavo da Bahia,
Cruz Das Almas,
Brazil
iTelescope: iT21 B,V,R,I
Andy CasonDawsonville,
Georgia, USA
10-inch LX200 + SBIG
ST8
iTelescope: iT5,
iT11, iT21, iT24
B,V
B,V,R,I
Tõnis EenmäeTartu Observatory,
Tartu, Estonia
Planewave CDK 12.5-
inch + Apogee Alta
U42 camera
B,V,R,I
Raymond Kneip LuxembourgiTelescope: iT11,
iT18, iT24B,V
Massimiliano Martignoni Magnago, ItalyS-C 25cm-f/10.0telescope + KAF0261E
CCD camera
B,V,I
John Martin
UIS Barber
Observatory,
Springfield,
Illinois, USA
Barber 20-inch R-C
reflector + Apogee
Alta U42 Camera
B,V,
R,I
Table 1. A roster of volunteers, their telescopes and cameras, and the filters they used to take images of
8/18/2019 SAS 2014 Proceedings
19/252
Martin: Crowd Sourced Lightcurve for SN 2014
5
identifying a list of reliable comparison stars after thediscovery of SN 2014G. Their effort directly helpedgenerate interest and identify individuals for recruit-ment to our project.
5. References
Branch, D., et al . (1991). Comm. Astrophys. 15, 221-237.
Denisenko, D., et al. (2014). CBET 3787.
Eenmäe, T., et al . (2014). ATEL 5935.
Fraser, M., et al . (2013). MNRAS 433, 1312.
Margutti, R., et al . (2014). ApJ 780, 21.
Martin, J.C., et al . (2014) astro-ph/1308.3682
Mauerhan, J.C., et al . (2013). MNRAS 430, 1801.
Munari, U., et al . (2013a). New Astronomy 20, 30.
Munari ,U., et al . (2013b). IBVS 6080.
Munari, U., Moretti, S. (2012). Baltic Astronomy 21,22.
Preito, J.L., et al . (2013). ApJ 763, L27.
Pastorello, A., et al . (2013). ApJ 767, 1.
Smith, et al . (2013). MNRAS 438, 1191.
Soker, N., Kashi, A. (2013). ApJ 764, L6.
8/18/2019 SAS 2014 Proceedings
20/252
Martin: Crowd Sourced Lightcurve for SN 2014
6
8/18/2019 SAS 2014 Proceedings
21/252
Hambsch: Observations of Novae from ROAD
7
Observations of Novae from ROAD Franz-Josef Hambsch
VVS, AAVSO, BAV, GEOS
Oude Bleken 12, B- 2400 Mol, Belgium
Abstract
The author discusses observations of galactic novae and some extragalactic supernovae from his remote obser-vatory ROAD (Remote Observatory Atacama Desert) he commenced in August 2011 with Nova Lupi 2011 (PRLup). The observed novae are mainly chosen according to alert notices and special notices of the AAVSO aspublished on their website. Examples of dense observations of different novae are presented. The focus goes tothe different behaviors of their light curves. It also demonstrates the capability of the remote observatory ROAD.
1. Introduction
Galactic novae and extragalactic supernovae
seem to be a very intensive field of research in pro-fessional astronomy. Just recently (2013) a confer-ence on Stella Novae: Past and Future Decades washeld in Cape Town
http://www.ast.uct.ac.za/stellanovae2013/
It was mentioned in one of the presentations,which are available in internet, that more than 700refereed papers on the subject of novae were pub-lished in the past decade.
The author started to observe novae based on re-quests from the AAVSO and different mailing list
(mainly VSNET and CVNET).Due to the very good observing possibilities atthe remote observatory (300+ clear nights) intenseobservations of those stars were possible. Those ob-servations are mainly done as snapshots every clearnight and sometimes also as time series if of interest.
During the many clear nights at the remote site alot of data is being gathered on many stars. In the
present paper a selection of novae is presented withtheir respective light curves based on data from theAAVSO including the ones of the author.
2. Observatory
The remote observatory in Chile houses a 40-cm f /6.8 Optimized Dall-Kirkham (ODK) from OrionOptics, England. The CCD camera is from FLI andcontains a Kodak 16803 CCD chip with 4Kx4K pix-els of 9-µm size. The filter wheel is also from FLIand contains photometric BVI filters from Astrodon.
In Belgium, where the author lives, a roll-off-roof observatory is used in addition for variable starobservations from the Northern hemisphere, if it ever
gets clear. Belgium is neither famous about itsweather nor its light pollution, nevertheless there isstill some room for interesting observations. The tele-
scope is a Celestron C11 working at f /6.3. The tele-scope is equipped with an SBIG ST8XME CCDcamera using BVRI photometric filters. Soon aStaranalyser SA200 grid for low resolution spectro-scopic investigations will be installed, too.
Fig. 1 shows an image of the remote telescope inChile in operation during night. It is housed in aclamshell dome, making easy movement of the tele-scope possible without the need to follow with a shut-ter of a normal dome.
Images of a night’s session are acquired withCCDCommander automation software. Further anal-ysis in terms of determination of the brightness of thestars is done using a program developed by P. dePonthierre (2010). The data are then finally submittedto the AAVSO.
Fig. 1. Photo of the remote telescope installation in Chilein operation at night.
3. Novae
The observations at the remote site started onAugust, 1, 2011. The first nova which was observedwas Nova Lupus 2011 (PR Lup). But before we start
8/18/2019 SAS 2014 Proceedings
22/252
Hambsch: Observations of Novae from ROAD
8
to discuss the observations we should look at what isknown about novae in literature. Since the author isnot a specialist in this field, nor a professional as-tronomer, he has to rely on the internet and e-mailinformation on relevant literature. The most compre-hensive paper is probably the one of Strope et al .(2010). It contains a catalog of 93 nova light curveswith a classification of the light curves (see Fig. 2)and a list of properties of the novae, like peak magni-tudes, decline speeds etc.
Fig. 2. Different prototypes of novae based on the formof their light curve.
One of the problems mentioned in this paper isthe fact that only very few of the brightest novaewere followed more extensively down to quiescencemainly by amateurs and only two of them by profes-sionals. Hence, there is still room for amateur obser-vations if they are done in a more systematic way.The covered light curves of novae in this paper aremainly based on observations from the AAVSO in-ternational database. The database contains more thandouble the number of novae, but only those coveredin the article have a decent number of observations.In the paper the light curves could be classified ac-cording to 7 different types, which are called classes.The prototype light curves of the different classes aregiven in Fig. 2.
In the following a few examples of novae obser-vations from the remote observatory are given. Notall classes could be observed. All images shown are based on own observations (blue squares) and addi-
tional observations found in the AAVSO internation-al database. BVI photometric filters were used.
4. Nova Cen 2012-2 (V1368 Cen)
A new star in Centarus was discovered by J.
Seach, Chatsworth Island, NSW, Australia on April4.765, 2012, using a DSLR with a 50-mm f /1.0 lens.It was the second nova discovered in the constellationCentaurus in 2012. It was assigned the identifier TCPJ14250600-5845360 and was published on the Tran-sient Object Confirmation page of the IAU CentralBureau for Astronomical Telegrams
http://www.cbat.eps.harvard.edu/unconf/tocp.html
Due to the very short focal length of only 50-mmthe position of the nova was not very well known. A better astrometry position was determined with the 2-m Faulkes Telescope South by E. Guido, G. Sosteroand N. Howes. A spectroscopic confirmation obser-vation of the novae was performed by T. Bohlsen.
As one can see from the light curve in Fig. 3, thenova was discovered before maximum light. Obser-vations at ROAD were started already as early asApril 5.902, hence just a bit more than one day afterthe initial discovery. Observations were continuedover 46 nights in V and I band filters (see Fig. 3). Asmooth behavior of the light curve was observed (S-type light curve).
9. Nova Mon 2012 (V0959 Mon)
Nova Monoceros 2012 (V0959 Mon) was dis-covered on August, 9.8048 UT, 2012 by S. Fujikawa(Kagawa, Japan) using a 105 mm FL camera lens anda CCD camera at mag. 9.4 with a clear filter. It wasassigned the identifier PNV J06393874+0553520.Remote observations at ROAD started on Aug.12.918 using V and I band filters. The nova was fol-lowed both from Chile and New Mexico. Both snap-shot and long time series observations were done.From New Mexico using two different telescopes andobserving the stars over its full observing period dur-ing several nights BVRI time series could be ac-quired. An analysis of those time series confirmed the
7.1h period of this stars (Hambsch et al ., 2013), dis-covered in UV/X-ray observations of this nova (Os- borne et al., 2013). As one can see the observationshave continued from the discovery until now, onlyinterrupted by the fact that the star’s observing sea-son ended.
8/18/2019 SAS 2014 Proceedings
23/252
Hambsch: Observations of Novae from ROAD
9
Fig. 3. Light curve of the Nova Cen 2012-2 (V1358 Cen) based on observations found inthe AAVSO database. The blue squares are observations taken remotely at ROAD.
Fig. 4. Light curve of the Nova Mon 2012 (V0959 Mon) based on observations found in theAAVSO database. The blue squares are observations taken remotely at ROAD.
Fig. 5. Phase diagram of nova Mon 2012 with P = 0.29575 ± 0.0005 d = 7.098 h.
8/18/2019 SAS 2014 Proceedings
24/252
Hambsch: Observations of Novae from ROAD
10
Fig. 6. Light curve of the Nova Oph 2012 (V2676 Oph) based on observations found in theAAVSO database. The blue squares are observations taken remotely at ROAD.
Fig. 7. Light curve of Nova Sgr 2014 (PNV J18250860-2236024) based on observationsfound in the AAVSO database. The blue squares are observations taken remotely atROAD.
Fig. 8. Light curve of the Supernova SN2013aa in NGC5643 based on observations foundin the AAVSO database. The blue squares are observations taken remotely at ROAD.
8/18/2019 SAS 2014 Proceedings
25/252
Hambsch: Observations of Novae from ROAD
11
There is a clear behavior visible that I band mag-nitudes do not behave in a parallel way as the V bandmagnitudes. Maybe this nova could be regarded as amember of either the smooth or plateau class.
Fig. 5 shows the phase diagram of the time seriesobservations from New Mexico. A period of P =7.098 h was used to generate the image.
10. Nova Oph 2012 (V2676 Oph)
This nova in Ophiuchus has been discovered byH. Nishimura, Mar. 25.789, 2012 using a Canon 200-mm f /3.2 lens with a Canon EOS 5D digital camera.It was assigned the identifier PNV J17260708-2551454. On Mar. 27.74 low resolution spectra weretaken at the 1.3m Araki telescope at the Koyama As-tronomical Observatory (KAO) to confirm the newobject as a classical nova.
Observations started at ROAD on Mar. 28.865UT using V and I band filters. Observations wereconducted as snapshots during 42 nights. Fig. 6shows the light curve based on all observations foundin the AAVSO international database. Clearly theshape of the light curve is very different to the previ-ous ones and shows a kind of jitter or oscillationsyielding a light curve showing minima and maximauntil after about 85 days an abrupt decline started.Unfortunately no further observations have been per-formed after this drop.
11. Nova Sgr 2014(PNV J18250860-2236024)
Another example of a recently newly discoverednova in Sagittarius is Nova Sgr 2014 (PNVJ18250860-2236024). It has been discovered by S.Furuyama, Ibaraki-ken, Japan on Jan. 26.857, 2014using a 200-mm f /2.8 lens and CCD camera. A low-dispersion spectrum (R about 980 at 650 nm) of theobject was obtained using the 2-m Nayuta telescopeat the Nishi-Harima Astronomical observatory, Ja-
pan, which confirmed the nature of the object as anova.
Observations started on Jan. 28.393, using V andI band filters. Observations were conducted as snap-shots one and a half day after the discovery. Observa-tions are still ongoing as the nova is still bright.
Fig. 7 shows the light curve based on all obser-vations sent to the AAVSO. Looking at the bluesquares as the observation from the author it is clearthat the data from ROAD are the majority for thisstar. From the light curve it is clear that oscillationsin the brightness of the nova exist and it belongs tothe oscillation class. Since about 20 days the lightscurve has change its behavior and a drop in magni-
tude in both I and V band is visible. It will be inter-ested to see the further behavior of this star in themonths to come.
12. Supernova SN2013aa in NGC5643
This supernova in NGC5643, a galaxy in theconstellation Lupus was observed after an e-mailalert. It was discovered on Feb. 13.621, 2013. Itsidentifier is PSN J14323388-4413278. In Fig. 6 ob-servations of the AAVSO database are shown ofwhich observations from ROAD form the majority.Observations started as snapshots on Feb. 15.895 UTin V and I band filters. From Fig. 6 it is obvious thatin I band the light curve showed a second brighteningabout 40 days after the maximum brightness. Alsothe V maximum brightness came after the I bandmaximum by about 10 days. Then the supernova was
brighter in V band for about 20 days compared to theI band brightness.
Observations extended over a period of 171nights only with a few interruptions due to badweather.
13. Nova Trianguli Australis 2008(NR TrA)
Fig. 9. Phase diagram of NR TrA with P = 5.259 ± 0.003 h.
This nova was observed in collaboration with theCentre of Backyard Astrophysics (CBA). The novawas shining at about mag 8.5 in April 2008. Timeseries observations were started Mar. 27, 2013 and
lasted for 33 nights. The present brightness of thisnova is around mag. 15.5. Fig. 9 shows the phasediagram of the observations. This figure was generat-ed using a period of 5.259 ± 0.003 h.
This period is very close to the one observed atthe CTIO 1.3-m telescope in Chile. This CTIO resulthas been presented at the conference mentioned inthe introduction in 2013 in Cape Town (SA). The
presentations given can be found on the website of
8/18/2019 SAS 2014 Proceedings
26/252
Hambsch: Observations of Novae from ROAD
12
the conference. It shows that with amateur equipmentand many clear nights results can be achieved whichare comparable to professional investigations.
14. Conclusion
The remote observatory under pristine skies inthe Atacama Desert opens up many possibilities toobserve variable stars. Intensive follow-up observa-tions over many days, weeks or even months are pos-sible due to the stable weather conditions. The givenexamples show impressively what is possible. Col-laborations are searched for in order to contribute toscientific research of common interest.
For this research the information given inAAVSO Alert and Special notices has been used.Also the information given in the VSX database isused.
15. Referencesde Ponthierre, P. (2010) LesvePhotometry software.http://www.dppobservatory.net/AstroPrograms/Softw
are4VSObservers.php
Strope, R.J. , Schaefer, B.E, Henden, A.A. (2010).http://arxiv.org/abs/1004.3698
Hambsch, F.-J., Krajci, T., Banerjee, D.P.K. (2013).The Astronomer's Telegram, 4803.http://www.astronomerstelegram.org
Osborne, J.P., Beardmore, A., Page, K. (2013). The Astronomer's Telegram, 4727.http://www.astronomerstelegram.org
8/18/2019 SAS 2014 Proceedings
27/252
Green: Nova Del 2013 Observations
13
Impact of Observing Parameters on17 Nights with Nova Del 2013
Wayne L. Green
Boulder Astronomy and Space Society
7131 Oriole Lane, Longmont CO, 80503
Abstract
Nova Del 2013 was reported in Astronomer’s Telegram #5279 (PNV 20233073+2046041) on 2013 Aug. 14. Onthe following day, the University of Colorado, Boulder granted our observing request for use of the R3000-5000spectrograph attached to the their 60-cm telescope. The planning, operational approach, analysis techniques,results, issues, and operational conclusions for data taken from 15 August through 2 September are reportedhere.
1. Introduction
The Boulder Astronomy and Space Society(BASS) caught wind of PNV J20233073+2046041on the 15 Aug. 2013 and made an immediate requestfor observing time with the R3000-5000 Boller andChivens spectrograph on the University of Colorado's60-cm (24-inch) research telescope. While the novawas bright and well positioned for observing, the sitewas under a seasonal Arizona Monsoonal weather pattern. The hour-angle of the object supported ob-servations into the early pre-dawn time. Data werecollected during intermittent clear periods in therange of 3,800-7,800 angstroms. Reports by theFrench “spectro-l message group” facilitated nightly
adjustments to the plan since they were east of us byabout 8 hours. A total of 51 spectra, 37 of the novaand 14 of reference stars were made yielding 8.7 Gi-ga-bytes of intermediate results.
Figure 1: AAVSO light curve spanning dates of programobservations.
These data were taken with a Boller-ChivensR5000 spectrograph and a Peltier-cooled ApogeeU1109 camera, nominally cooled to –20C° underambient. The grating is a 4800 l/cm blazed nominallyfor 4,500 angstroms. The dispersion is 1.04 ang-
stroms per 11 micron pixel. Slit widths of 25 and 54microns (1.15 and 2.5 arc-seconds) were used.
The camera was controlled by RandomFactory software and ds9, both running on a Raspberry Pisingle-board computer attached to the telescope. Anolder Macbook Pro provided the X-Server connectionvia ssh. This solved the problem of a long-run of theUSB cable.
Abnormal weather conditions existed across theentire observational period, including high relativehumidity and mixed clouds characteristic of the Ari-zona Moonsonal flow. Initial observations were madead-hoc to gain a re-familiarization with the instru-mental setup. The spectrograph requires a lot of man-ual intervention, especially with focus as it varies
with temperature and grating angle. No attempt tomaximize parallatic angle was made. The ds9 pro-gram was used for focus by saving a region for agood line-sequence and reloading for each focus im-age.
A cell-phone snapshot was mailed in the middleof the night to a fellow researcher, remotely observ-ing from the CU campus with the 3.5 meter ApachePoint telescope and the Triple-Spec spectrograph.The telescope was closed due to their own localweather issues. His response was to hammer away.Observations began in earnest after that.
2. Observations
Darks, zeros, and flats were obtained during ear-ly evening. The scope would be parked and the domeopened when weather was deemed safe. The targetwould be acquired and a 'run' would commence. Firstthe auto-guider would be initialized. The guider dis- play provided seeing condition information.
The university observatory is very well placedfor students and poorly placed for astronomy. The
8/18/2019 SAS 2014 Proceedings
28/252
Green: Nova Del 2013 Observations
14
facility is within walking distance of the front-rangefoothills. The observatory is generally leeward of thefoothills causing turbulent air flow resulting in seeingon the order of 3-4 arc-seconds for a nominal night.
A run consisted of focusing the spectrograph, asit pointed to the target, by setting to the smallest slitand using ds9 to choose between values from thefocuser's dial indicator. The slit matching seeing con-ditions would be reset and a 3-3-3 sequence started.The strategy called for 3 pre-comparison images, 3science images, 3 post-comparison images. The com-
parison lamps use separate low voltage HeNe and Ar bulbs. The telescope would be moved to a nearbyspectrophotometric reference star and 3-3-3 referencesequence obtained. Weather and time permitting, thespectrograph's grating angle would be changed to the'other' end of the spectrum and new run of 3-3-3 tar-get sequence followed by a 3-3-3 reference star se-quences would be taken.
It was often the case that only partial runs at one
spectrograph setting could be completed in a night.The run would usually commence using the spec-trograph's last setting to capture night-to-nightchanges of the target. Exposures would be aborted ifthe 6-inch guide telescope lost tracking ability due to
passing clouds. Weather interruptions would result invarying shifts in the spectra due to tube flexure, etc.over the 3-3-3 extended sequence.
Early one evening, during especially bad weath-er, data for a handbook with comparison images foreach setting on the spectrograph at each of its markedsettings from deep blue to past 1-micron was collect-ed. Line intensity varied as the lamps warmed. The
handbook of these annotated comp images facilitatedcomparison line identification, during the reduction
process.This exercise is recommended for all spectro-
graphs. Reproducing the document periodically willdocument lamp changes. This makes an excellentfirst exercise for a spectroscopy course.
ADU counts for the Hydrogen alpha line on theorder of 51,000 at a gain of 1.0 were noted, and pre-sumed to be well within saturation. This was an error.Later in the program, while visually aligning the tar-get in the guide scope, the star's remarkably red ap-
pearance was noticed. A quick series of very short
exposures proved that the camera was saturating atexposures levels previously used. This demonstratedwe had compromised the majority of Hydrogen ob-servations. A few serendipitous observations due tocloud obscuration helped to confine Hydrogen alphalines but with no trusted science value.
The program terminated on 3 September to allowa CU course to start their semester's programs. Theexposure times were becoming very long.
3. Data Reduction
The reductions were performed with IRAF andthe twodspec/apextract package. The steps beganwith building the master zero image. The master zerois subtracted from darks to build the master dark
frame. The images were zero subtracted, then darksubtracted to obtain the raw science image. The apalltask on the science image extracted and characterizedthe dispersion axis for the main image.
The apall task was first applied to the scienceimage to obtain the exact dispersion axis of the nova'simage. It was then applied to the comparison imageusing the science image as reference. This assured thecomp-lines come from the same patch of pixels alongthe dispersion axis. The identify task was used to-gether with the comp-line booklet produced duringthe run. The dispcor task was used to produce finalscience images. The dispcor task was applied a sec-ond time to produce a multi-spectrum file with acomplete WCS (spectral world coordinate) solution.The flat images were not trusted as the flex due totelescope repositioning was noticeable and weatherimpacts were severe.
The comparison lamp spectra from the start tillend of several sequences were subtracted with theimarith task. Flexure was noted across each run, es-
pecially between science and reference locations.Preparation of the master zero and dark images were
performed with custom photometry pipeline routines.
4. Operational Results
The net effect of this work was to refine the ob-servational approach to be used with the spectrographfor future work. The observational cadence strategyis to shoot a comparison lamp image after each spec-tral image to reduce error from flexure over long ex-
posures.The amount of flexure between a target position
and the flat screen meant some loss of accuracy wasintroduced while making the flats. Sky flats may bemore appropriate.
Analysis was performed with IRAF using the NOAO twodspec and apextract packages. Masterzeros, flats and darks were prepared but not em-
ployed for this paper. The noise from the atmospherewas made apparent by the noise frames produced bythe IRAF splot package.
A special program was developed in Python to produce publication quality graphs from IRAF multi-spec files. The data are extracted and converted to.dat files, common to other utilities. This allows over-
plotting several nights of data, including astrophysi-cal lines germane to the objects.
8/18/2019 SAS 2014 Proceedings
29/252
Green: Nova Del 2013 Observations
15
Figure 2: Flexure obtained by subtracting a comparisonimage from the beginning of a sequence from one at the
end.
5. Discussion
The reduced images clearly show the fireballstage early in the nova's life. The prominent P-Cygni profile is very evident. The nova quickly progressedinto the period where the expanding shell is opticallythick. The observations concluded just as the novamoved into later evolutionary stages.
Figure 3. Hydrogen alpha line profile changes across 17Aug to 28 Aug, with related lines shown in LFR.
Figures show the changes in the Hydrogen lines,displayed from the bottom for earlier dates. Thechanges in the P-Cygni profile were fast. The opaqueshell shows changes in other features.
The IRAF splot program allows zooming into aline area and displaying features against a velocity
profile. The usual lambda over d-lambda is takenfrom the dispersion data. The splot velocity imagewas hacked with emacs to make it more visible here.
Figure 4. IRAF splot image of Hydrogen-beta line dis-played with velocity scale.
Figure 5. Hydrogen beta line profile changes across 15Aug to 30 Aug, with related lines shown in LFR.
The asymmetry in P Cygni profile provides windvelocity and shell density information. As the shellexpands, the opacity opens up and blue light filtersout more easily. This happens quickly with novae ofthis type and was the case for Nova Del 2013.
The effective widths of features are easily deter-mined using the IRAF splot task. In this case, theHydrogen beta line is examined. The effictive widthswere with uncalibrated flux values.
The main window is zoomed into the feature us-ing a w-e-e command 'riff'. The colon command
8/18/2019 SAS 2014 Proceedings
30/252
Green: Nova Del 2013 Observations
16
“:log” is used to start logging. An e-e riff is usedacross the whole P-Cygni feature, results in a refer-ence line. A second e-e riff is made from the red sideto the crossing point of the reference line with therising middle edge of the feature. A third e-e riff isused from the crossing point back to the blue-start ofthe feature. This yields effective width's for the twosides of the feature. In figure: e1 to e2; e3 to e4; e5 toe6 creates EW values recorded into the file splot.log.
These values, if trusted are redeemed from thelog for proper analysis. Comments may be added tothe log with the colon-command “:# … comment “.The log is local to the analysis directory.
Figure 6. Results from IRAF task splot, reduced effectivewidths for Hydrogen beta P-Cygni profile.
6. Conclusions
Science happens regardless of the weather. Thelate summer conditions presented usable seeing butwith high water content. The relative humidity withinthe dome and surrounding air was high. The ds9quick manual focus technique helped maximize timeon target. The use of a quick and dirty analysis ofeach spectrum was not used during these runs - andshould have been. This technique has now been add-ed to observing strategy. The comparison imagehandbook, even in pixel coordinates, would havehelped assure the grating angle was uniformly set.These images have been rendered with notes in thedocumentation. The Hydrogen alpha and beta lines
were saturated reducing our ability to infer reddeningalong the LOS for the target.
A new run sequence strategy of setting the grat-ing; seek to target; focus; seek/take flat; seek to tar-get and use the strategy of a a comp - science -comp move comp-reference-comp is recommended tomaximize throughput with two grating angles underroughly same airmass and weather. Additional comp-science segments should be added to allow co-addingspectra for dimmer targets. The comp-science-comp
is preferred to the 3-3-3 method to reduce flexureover long exposures. The reference exposure of bright comparison stars are short enough not to needa following comp exposure.
Full characterization of the science camera isneeded. Setting the gain such that the ADU countsmaximize as the chip enters its upper non-linear re-gion prevents clipping issues for visiting observers.
7. Acknowledgements
The author wishes to thank the University ofColorado (CU) for their support of the Boulder As-tronomy and Space Society and providing the use oftheir facility for this experiment. Many members ofBASS attended the early runs while learning aboutspectroscopy. They include Evan Anderson, DavidBender, Steve Hartung, and David Ito. CU studentsBrandon Bell, Sara Grandone, Mike Potter, and JacobWilson attended the initial runs. The Observatorymanager, Fabio Mezzalira provided operational sup- port and encouragement.
Special thanks to Dr. Guy Stringfellow of CU forhis early input about approach and his encourage-ments. Data reduction was done with the NOAO/KPNO IRAF program and addons. Thanks tothe AAVSO for their Light Curve Generator and theL'Spectro Yahoo Group for fast reporting of results inadvance of our observing sessions.
IRAF is distributed by the National Optical As-tronomy Observatories, which are operated by theAssociation of Universities for Research in Astrono-my, Inc., under cooperative agreement with the US
National Science Foundation.
8. References
Astronomical ring for access to spectroscopy (aras)spectro-l. 2012.http://groups.yahoo.com/ group/spectro-l/
Darnley, M.J., Bode, M.F., Smith, R.J., Evans, A.(2013). “Spectroscopic Observation of PNVJ20233073+2046041 with the Liverpool Telescope.”The Astronomer’s Telegram 5279.
8/18/2019 SAS 2014 Proceedings
31/252
Patterson: Recovering from the Classical-Nova Disaster
17
Recovering From the Classical-Nova Disaster Joseph Patterson
Center for Backyard Astrophysics
Abstract
Classical novae rise from obscurity to shine among the brightest stars in the Galaxy. The story of how they returnto quiescence is still only dimly known. Vast amounts of energy are loosed upon the white dwarf and its compan-ion, and the light curves of post-novae suggest that they take not a few years, but a few thousand years, to re-turn to quiescence. In the meantime, the secondary may experience a lot of heating from the white dwarf's radia-tion - enough to overwhelm its intrinsic nuclear luminosity. I'll discuss the stellar physics behind this suggestion,and propose how it might be tested by time-series photometry in the months and years (and if possible, centu-ries) after outburst.
1. Introduction: The TextbookClassical Nova
It's dangerous to place hydrogen on a whitedwarf. For temperatures above ~20,000 K, most ofthe hydrogen is ionized, and if the conditions areright, the protons are then vulnerable to H->He fu-sion.
There's no problem at low density, because the protons don't find each other; and no problem at lowtemperature, because Coulomb repulsion keeps the protons apart. At high temperature (~10 million K)and high density (100 g/cm3), fusion can occur, andthis is the prescription for nuclear reactions in theinteriors of stars like the Sun. The fuel consumption
sounds prodigious ("six hundred million tons of hy-drogen per second"), but it's a steady and benign pre-scription, because the heat release slightly expandsthe burning region, and that expansion cools the burning region slightly – just enough to keep thetemperature from running away. Presto, you have amain sequence star.
White dwarfs can't do this, because prior stellarevolution has stripped them of hydrogen. The greatmajority are pure carbon, or pure helium. You mighthave heard of "DA white dwarfs", showing rather pure hydrogen in the spectrum – but this is the resultof diffusion. Heavier elements have sunk, and the H
spectrum is formed by a tiny layer of H right at thesurface (where the temperature and density are mod-est). Figure 1 shows a garden-variety white dwarf(WD).
Now start piling H on the surface, which hap- pens in a close binary system. After 10-4 MSun hasaccreted, conditions at the base of the envelope (i.e.,at the outer edge of the core) are prime for nuclearignition. The temperature and density here are nowquite high – high enough to begin nuclear reactions.
Because the core is supported by degeneracy pres-
sure, not thermal pressure, it does not expand and
cool; it heats up, so the reaction rate increases, so itheats up more, so the reaction rate increases stillmore, etc. Thus occurs a thermal runaway. All thatheat violently ejects the overlying 10-4
MSun. This is a
classical nova, as seen by theorists.
Figure 1. The core and envelope (dotted) of a whitedwarf. Danger lurks at the core-envelope boundary,
where T and are both high – and the igniter (hydrogen)is present.
The ejected shell is initially very opaque; 10-4
MSun is a lot of matter! Increasing rapidly in area, theshell brightens to MV~ -8, and the spectrum lookslike an A supergiant. Then the shell thins out andgoes transparent. Now we can see down to the inner binary; and lo and behold, a very hot dwarf(~500,000 K) becomes visible to soft X-ray tele-scopes. Eventually that soft X-ray source fades, andthe star returns to quiescence as a normal CV pow-
8/18/2019 SAS 2014 Proceedings
32/252
Patterson: Recovering from the Classical-Nova Disaster
18
ered by mass transfer. This is a classical nova, asseen by observers.
2. Filling in the Details
These views comprise the basic working hypoth-
eses of essentially everyone studying classical novaenowadays. That doesn't make them “correct,” butthey’re not currently under challenge. However,many remaining details are up for grabs, e.g., thetimescale of these events. The basic picture hypothe-sizes:
1. One very sudden, faster than any observedtimescales, and spherically symmetric ejection.
2. A return to quiescence after a few years.
Recent studies of the ejected shells prove that the
shells are not spherically symmetric – not even close – and also suggest the existence of multiple ejections,stretching over weeks to months. So item (1) willlikely soon be replaced by something new and im-
proved. As for (2), we think that’s overdue for re-tirement as well. That's the subject of this paper.
3. The Post-Nova
Simple inspection of the light curves suggeststhat “a few years” is the correct timescale for returnto quiescence. Robinson (1975) studied the pre-eruption magnitudes, as recorded on archival photo-
graphic plates, and concluded that essentially all no-vae are equally bright before and after outburst. Hisdata had a post-eruption time baseline ~10-70 years,so this was taken to mean a return to quiescence in <20 years. This much-cited result has graduated intothe category of “conventional wisdom.”
None of the forgoing discussion refers to the or- bital period of the binary. Does POrb matter at all?The powerful events of the nova are seated in theWD, the secondary star is just slowly feeding the WDwith matter, and return to quiescence is presumablysome kind of WD cooling process. Seemingly, POrb shouldn't matter.
However, all the novae studied by Robinsonwere of long POrb, and those stars have powerfulmechanisms (called “magnetic braking”, though thetrue mechanism is unknown) for stimulating masstransfer at 10-8 MSun/yr. So they probably can’t fadeany farther, and a more accurate statement of Robin-son’s result could be merely “a few years to fade to10-8 MSun/yr (corresponding to MV~ +4).” To studythe fading of novae, it would be better to look at starsof short POrb, which transfer matter at much lower
rates (10-10 MSun/yr), are intrinsically much fainter(MV ~ +10), and thus can more accurately track thetrue evolution of the nova.
4. BK Lyn Points the Way
I started thinking about these matters again twoyears ago, when we identified a classical nova thatwas observed by Chinese astronomers on December30 in the year 101. Actually, it was identified by anamateur astronomer, Keith Hertzog, in a 1986 paper(Hertzog, 1986). For whatever reason, this was un-known or ignored by everyone, including me, alt-hough he had written a letter to me about it! Any-way, in the 1990s I became extremely interested insuperhumps, the large-amplitude waves sprouted bydwarf novae during their brighter outbursts (“su-
peroutbursts”). Every dwarf nova with POrb < 2 hrs showed these waves, and we found many old
novae and novalike variables did as well. During oursurvey we found superhumps in BK Lyn, which was-n't surprising since its POrb (103 min) was in the re-gime where every star accreting at a high rate, eventemporarily, showed these waves. Eventually the
phenomenon came to be understood as the result ofan eccentric instability which develops at the 3:1 or-
bital resonance of an accretion disk (Whitehurst,1988, and about 100 subsequent papers).
The BK Lyn mystery was: why is it accreting atsuch a high rate? Below the period gap, there are~500 known or very probable dwarf novae, 5 classi-cal novae, and just 1 novalike variable (BK Lyn).
One out of 500, that's not many, but theory says thereshould be zero because short-period CVs should beentirely powered by gravitational radiation, which
proceeds at a known rate and can only power an ac-cretion rate of 10-10 MSun/yr, come hell or high water.But the brightness of BK Lyn appears to demand anaccretion rate of 10-8 MSun /yr. What powers it at sucha high rate?
I should have remembered Hertzog's paper atthat point, but I didn't. The star provided another bighint in 2012, when the mean brightness faded slightlyand the star turned into an indisputable dwarf nova!Theory predicts that should occur to most classicalnovae as they fade in light, but doesn't specify the
timescale for its occurrence. Only one other nova(GK Per = Nova Per 1901) has ever clearly done that,
but with POrb = 1.99 days, it resides far outside the bounds of CV normalcy. Anyway, then I rememberedHertzog's paper; I re-read it and saw that his argu-ment rests on a very accurate (within 0.25 squaredegrees) positional agreement between the modernCV and the ancient Chinese nova. This extreme accu-racy was possible because the Chinese described the
8/18/2019 SAS 2014 Proceedings
33/252
Patterson: Recovering from the Classical-Nova Disaster
19
guest star as situated right next to the “fourth star ofthe constellation Hsien-Yuan” – a star that we todayknow as Alpha Lyncis. Other Chinese scholars con-firmed this. So, very accurate positional agreement, avery unusual star (1 in 500), and performing a trickexpected for centuries-old postnovae. It all added upto a very high probability that BK Lyn = Nova Lyn101.
So that was nice – “the oldest old nova” – andspurred us to write a paper about the star (Patterson etal., 2013). Yet the really important question remains:why is it still blazing away at MV ~ 5.7, about 100x brighter than nearly all other CVs of similar period?Although we could not find any photographic recordof its brightness prior to the year 101, this very likelycontradicts the Robinson rule (“same brightness, be-fore and after outburst”).
It's mighty tempting to suppose that this is an af-ter-effect of the nova eruption, still very obvious 2000
years later .
5. Follow The Brightness: Short- andLong-Period Novae
Could this possibly be a general effect in novae,contrary to Robinson’s rule? We studied the long-term visual brightness history of one nova of longPOrb (V603 Aql = Nova Aql 1918) plus one of shortPOrb (V1974 Cyg = Nova Cyg 1992). The former wasthe brightest and nearest nova of the 20th century,and the latter is sometimes called “the nova of thecentury” because of the great wealth of observational
detail obtained during and after eruption. (Now thatthe century is over, I’d have to say that that supremelabel is deserved). We also like these stars becausewe have been studying them with intensive time-series photometry over the past 20 years.
We used the AAVSO database for this historicalstudy. That's almost always a good choice, becausevisual and V-band photometry tends to minimize problems with calibration (always a potential threat,and especially over long time baselines).
The result is shown in Figure 2. Both stars reachmaximum light near MV = –8 and then decline, rapid-ly at first and then slowly. V603 Aql is now at MV =+3.5 and still declining (see also Johnson et al .,2014). V1974 Cyg appears to be in somewhat morerapid decline, at least in these logarithmic units. Ac-tually, that “more rapid” decline must go on for avery long time, because just before eruption, sky sur-veys showed V > 22, and hence MV > +9, at the posi-tion of the nova (Collazzi et al ., 2009). So the behav-ior of V603 Aql slightly chips away at Robinson'srule, and that of V1974 Cyg strongly violates it. Wealso put BK Lyn on this plot; at an age of 2000 years,
it’s on a plausible interpolation of the V1974 Cygline (assumed to be representative of short-periodnovae).
Figure 2. Road map for the decline of classical novae.Stars of long POrb asymptote around MV > +5, because ofmagnetic braking keeps them bright. Stars of short POrb,
ineligible for magnetic braking, follow the "natural" (ac-cording to us) decline of a cooling nova.
What about other novae, not shown here? Well,scores of long-period novae are available for com- parison; at the scale of Figure 2, they all substantiallyagree with V603 Aql – asymptoting near MV = +4.Essentially all the novae in Robinson's study were oflong POrb, and the V603 Aql line suggested in Figure2 is entirely consistent with his result (10-30 yearsafter outburst basically the same as 10-30 years be-fore). The only other short-period nova available forcomparison is CP Puppis (= Nova Puppis 1942). The
AAVSO light curve for that star, on the scale of Fig-ure 2, is basically identical to that of V1974 Cyg.Even 72 years after outburst, it remains at least 3.5magnitudes above the pre-outburst brightness level(measured by Collazzi et al ., 2009).
So, to our eye, Robinson’s rule is slightly violat-ed by the long-period novae, and flatly contradicted by those of short period.
Finally, although BK Lyn is the only actual no-valike among the short-period CVs, there are a few(5-10) dwarf novae that are significantly brighter thanthey are entitled to be under the rule of GR. These arethe ER UMa stars. They are 1-2% of all dwarf novae,so we estimate their age to be 1-2% of the entire 10 6 year interval between eruptions. They provide a stagefor BK Lyn to evolve into, on its path to short-POrb mediocrity (500 objects at ~300,000 yrs).
6. The Inner Binary, Up Close, andPersonal During Eruption
The nova eruption's origin and powerhouse islikely to rest solely in the WD; after gas arrives at the
8/18/2019 SAS 2014 Proceedings
34/252
Patterson: Recovering from the Classical-Nova Disaster
20
WD surface, it's unlikely to preserve much memoryof where it came from (or equivalently, the POrb of the
binary). So, guided by Ockham’s Razor, we are in-clined to seek an understanding of novae which as-sumes that all nova outbursts are basically the same(mutatis mutandis; there is a known dependence onWD mass, because WD density depends sharply onmass). When we do the arithmetic on the total radiat-ed energy of the eruption, we always find somethinglike 2x1046 ergs – roughly independent of POrb andthe nova's “speed class.”
The secondary is very near to this brilliant light bulb, and must absorb some – about 1% – of its lu-minosity. We should ask: how will all this extra un-wanted energy affect the beleaguered secondary?Well, binaries of short POrb have secondaries of ~0.1MSun, and the nuclear luminosity of such a star is just1030 erg/s – which implies 1038 ergs over the ~ 3year-long eruption. Thus the nova shines on the sec-ondary for 3 years with an energy 2 million times
greater than provided by the secondary's intrinsicnuclear luminosity.
Such brilliant illumination must give the second-ary a terrible case of sunburn. There should be a
bright spot on the secondary on the side facing theWD, and as it wheels around, we should see a largeand roughly sinusoidal signal at the orbital period.
Novae should all be monitored for this effect – it'smandatory for the nova, with three qualifiers:
1. It shouldn't exist near maximum light, when theejected shell is opaque and thus obscuring theview down to the inner binary. The effect
should begin approximately when the soft X-ray source first appears (our first view of thefreshly erupted WD); that usually occurs a fewmonths after maximum light. It should thengradually decline as the WD cools, over a peri-od of years.
2. It's much less likely to exist in novae with red-giant secondaries – sometimes called “symbi-otic novae”. These secondaries generate plentyof their own nuclear luminosity, and are lesslikely to be fazed by the new light bulb shiningin their sky.
3. The observable effect should be zero for thosefew binaries orbiting in the plane of the sky(“face-on”).
How much of this is solidly supported by evi-dence? We don’t really know, because most profes-sional studies focus on the emission lines from theouter gas bag (the ejecta). The inner binary has onlyweak emission lines, and they're usually over-whelmed by the many lines from the ejecta. So these
ideas will be mainly tested by time-series photome-trists, who can discover the orbital period of novaewhen they are still bright and accessible, and evenmeasure the cooling rate of the WD from the declin-ing amplitude of the heating effect. The latter would
be great to know, and is poorly determined by soft X-ray telescopes, which have practically no sensitivityfor WD temperatures below ~300,000 K.
7. And for The Many ThousandYears Thereafter
We expect that time-series photometry will revealthis heating effect in nearly every nova with POrb
8/18/2019 SAS 2014 Proceedings
35/252
Patterson: Recovering from the Classical-Nova Disaster
21
cratic. Is it really? We've suggested here that they'reall pretty much the same, depending only on POrb.Finally, our total supply of “old novae” is strangely
biased – more than 100 in the 20th century, 8 in the19th, 1 in the 18th, and 1 in the 2nd. In the 20th century,6 were magnitude 1 or brighter. Who is going to findall those bright novae that went off between 101 and1783? BK Lyn was found by an amateur, a college
professor of anthropology. He’s no longer available.Who will do his job now?
8. References
Collazzi, A., et al . (2009). Astron. J. 138, 1846.
Hetzog, K.P. (1986). Observatory 106, 38.
Johnson, C.B., et al . (2014). ApJ 780, L25.
Patterson, J., et al . (2013). MNRAS 434, 1902.
Robinson, E.L. (1975). Astron. J. 80, 515.
Whitehurst, R. (1988). MNRAS 232, 35.
8/18/2019 SAS 2014 Proceedings
36/252
Patterson: Recovering from the Classical-Nova Disaster
22
8/18/2019 SAS 2014 Proceedings
37/252
Vail: CCD Flat Fielding
23
Pushing the Envelope: CCD Flat Fielding James Vail
SAS/IFAS
2050 Belmont Ave
Idaho Falls ID 83404
Abstract
In this paper the author discusses the design aspects and considerations of flat field systems. Illumination calcu-lations and construction techniques and materials are investigated. Several actual systems along with testingmethods to determine quality are presented.
1. Introduction
A flat field image is a representation of an opti-cal path that only shows the optical and instrument
flaws of the path. It is produced by imaging a blankfeatureless field that in the absence of optical andCCD flaws would be an image of an even field ofgrey. Alas, the absence of optical and CCD flaws isseldom found in the real world. Vignetting, reflec-tions, lens flare, scattered light, dirt, CCD gain varia-tions and flaws typically produce an image of gradi-ents, uneven pixels, out-of-focus dust rings and deadzones. If these artifacts can be imaged in exactly thesame manner as a data image is recorded, they can beused to remove these flaws from the data image.
The process of CCD image flaw correction or“flat fielding” is an electronic version of the dark-
room “masking” technique used to correct lightfalloff in wide-angle photographs. In the case ofCCDs, an image of an evenly lit blank field is takenunder the same optical conditions as the desired dataimage (Wodaski 2002). The exposures must be ad-
justed for about half saturation or “half well” so ahigh signal to noise ratio is maintained without riskof saturation or nonlinearity. All of the flat field pix-els are then normalized based on the average pixelvalue over the full flat field image and this normal-ized image is divided into the data image pixel by
pixel. The trick to this whole process is to acquire aflat field image that does not have artifacts that arenot also found on the data images. Otherwise, the
process would add errors to the final product that didnot appear on the original data image.
2. Present State of the Art
2.1 Box Flats
Small telescopes have an advantage over largeobservatory telescopes in this area. The flat illumina-tion can be produced right at the primary opening ofthe telescope (Wodaski 2002), but even this has prob-lems. With fabric, opal glass, or a plastic diffuserover the end of the telescope, it still has to be illumi-nated evenly. The illumination is typically provided
by small lamps around the inside corners of a flatwhite box with shields on the bulbs to prevent directlamp light from impinging on the diffusion screen.An additional diffusion sheet is used to even it outfurther. As the telescope gets larger, much more baf-fling is required and the box must be deeper.
2.2 Sky Flats
Another popular flat field technique is the skyflat (Wodaski, 2002). They come in two varieties, thetwilight flat and the blank sky flat.
2.3 Twilight Flats
The twilight flat is taken at twilight facing oppo-site the Sun and looking at empty, cloudless sky be-fore the stars are visible. Unfortunately, the sky dark-ens quickly and a full set of 5 flats per filter must be
taken during this fleeting period of about 20 minutesor so. For all to go well, the sky must be free of cloudor haze that could produce a gradient. The observermust be nimble enough to make each exposure andcheck each exposure for filter effects and changinglight level in less than a minute. However, with prac-tice, this can be done. (Henden)
8/18/2019 SAS 2014 Proceedings
38/252
Vail: CCD Flat Fielding
24
2.4 Blank Sky Flats
The blank sky flat is exposed in the dark of nightusing the light from the background sky. In order forit to work, an area of the sky with very few brightstars is selected. Several long exposures are then tak-
en after moving the telescope a little bit each timesuch that what stars are recorded can be averaged out by median combining the sets of images. When doneright this can consume a large period of valuable ob-serving time. The light being dominated by spectrallines and atmospheric emissions rather than a smooth
black body spectra further complicates the process.(Magnier 2001).
2.5 Dome Flats
Almost every major observatory and many ama-teurs have an evenly illuminated screen or wall thatcan be used as an out of focus flat field. These domeflats as they are called are convenient and repeatable.One would think it was easy to create an evenly litwhite target to image for a flat field, just project dif-fuse light on a screen painted with a color neutral
paint (NOAO 1991). However, even better qualitydiffusion darkroom enlargers have light gradientsacross their projections. Philip Massey (Massey1997) states in his Users Guide to CCD Reductionswith IRAF : “It is not unusual to find 5-10% gradientsin the illumination response between a dome flat anda sky exposure, and this difference will translate di-rectly into a 5-10% error in your photometry.” Theyshould be combined with sky flats because “...dome
flats alone will leave a low-frequency non-flat scalingin the images. (Gates, 2002).
2.6 Drift Scanning Flats
Several professional and amateur observation programs (SDSS 2004, TASS 2004, and COAA2003) involve a fixed telescope with a CCD camerathat tracks images electronically. This is accom-
plished by marching the pixel charges across theCCD chip at the same rate as the image drifts acrossthe chip. Thus, each star image can accumulate overthe total time of a chip transit. However, unlike the
two-dimensional drifting sky, a fixed flat image can be represented as a single line of pixels such thateach pixel represents an average of an entire column.The flat field becomes a single dimensional field thatmatches the column-by-column variation in the skyimage field (Gunn et. al. 2004).
3. Creating an Ideal Flat Field
3.1 Box Flats
Taking each method separately, the light box isthe easiest to improve. First, a look at how light is
distributed in a reflecting light box.If a point source of light illuminates a flat sur-face, the light intensity at any point is proportional tothe cosine of the angle from the surface zenith to thelight source. If the flat surface is an ideal scatteringsurface, it will scatter light in a Lambertian distribu-tion (Tulloch 1996) that acts as a point source at each
point on the surface.In a light box with a tungsten bulb pencil beam
at each corner, a truly flat light distribution over adiffuser surface is only possible with a box aspectratio of one in all dimensions (see Figure 1). Similar-ly, if bare bulbs are used, the middle portion of the
reflecting surface will have flat illumination. In turn,the illumination of the diffusion screen at the bottomof the box will be relatively even in the center butwill still have a small increase around the edges.
Figure 1. The Illumination profile from an array of fourLambertian scatterers radiating downwards onto a plane(Tulloch 1996).
8/18/2019 SAS 2014 Proceedings
39/252
Vail: CCD Flat Fielding
25
3.2 Dome Flats
A similar design could be applied with domeflats. Instead of a single point projection illuminationof a dome flat, it could be illuminated with a set offour lights attached to the nose of the telescope
spaced in a cubic array with the flat screen. For addedversatility, multiple lamps of different wavelength bands could have their own dedicated array.
3.3 Twilight Flats
Improving sky flat methodology is a bit more in-direct. Not much can be done to change the sky butthere are changes to technique that are possible. The
present method of collecting twilight flat (Henden) isto:
• Expose for a few seconds• Read the well depth from the CCD
• Adjust the exposure time• Expose again for the new exposure time• Repeat the same exposure several more
times• Change filter and start from the top again.• All of the above are repeated until all the fil-
ters are used.
With a rapidly changing twilight sky it is a con-tinuous cut and try process.
What if a small CCD was placed adjacent to themain CCD that could act as a photometer and termi-nate the exposure when there was sufficient light
collected to fill the main CCD to half well. DigitalSLR cameras have just this sort of feature. Manyobservatory and amateur CCD units also have such adevice for guiding and tracking that could be used forthis purpose if combined with custom software. Thenall the cut and try will end and dozens of perfect flatscan be exposed in rapid sequence.
3.4 Blank Sky Flats
Blank sky flats are even harder to improve.Hopefully the observer has a list of low star popula-tion regions on hand and collects flats in those areas.The standard practice of suppressing star and cosmicray pixels by median averaging of several slightlyshifted images works best when many images areinvolved. Some additional improvement can beachieved with selectively clipping of high and lowvalue pixels with software features such as IRAF’sFLATCOMBINE command (Massey 1997). Alt-hough automated clipping is a valuable tool for highimage volume photometry projects with well-established equipment performance and highest
quality CCDs, it may not be for everyone. Typically,clipping is set at about 2.5 to 3 sigma from the medi-an average. Most of the features that a flat must haveare within this range so the flat is effortlessly cleanedup. What of the unique optical setup that is not wellengineered or maintained and the poorer qualityCCDs? Their desirable range of pixel values is muchwider than 3 sigma especially on the lower end of therange. An additional problem is the lack of this clip-
ping feature outside of IRAF (IRAF, 2004).I would propose a more selective approach. A
histogram display of a bias and/or dark subtractedsky image shows a distribution with an upper edgefalling off at the sky glow level. Above that curve is a
background of dim stars within a magnitude or so ofsky glow and a significant peak at the upper end rep-resenting saturated stars. Many photometry and im-age correction software will allow histogram manipu-lation such that the pixels above a certain threshold(e.g. 3 x square root of average sky glow) can be re-
set to the value of that threshold. If this process is performed and the flat images are arithmetically ormedian averaged, the stars and cosmic rays are sup-
pressed without suppressing the pixel-to-pixel gaindifferences or large image flaws. Clipping at the low-er end of the histogram should be approached cau-tiously if at all. Especially with amateur and studentequipment, many of the low pixel values are real de-fects and should not be touched.
3.5 Drift Scanning Flats
The SDSS creates flats by binning 32 columns to
a bin and calibrating standard stars within each binand applies a smooth correction to each column toremove low spatial frequency structure (Gunn et. al.2004). High spatial frequency structure is ignored asrandom variations. They will be examined in the fu-ture with oblique scan data.
4. Example systems
Several efforts to build stand-alone flats were made by the author.
4.1 Paint Materials for Reflective Flats
The initial approach to creating dome flats wasto produce a practical color neutral paint that wouldhave full range