<%BANNER%>

Achieving Diffraction-Limited Angular Resolutions in the Optical Through Speckle Stabilization

Permanent Link: http://ufdc.ufl.edu/UFE0042840/00001

Material Information

Title: Achieving Diffraction-Limited Angular Resolutions in the Optical Through Speckle Stabilization
Physical Description: 1 online resource (148 p.)
Language: english
Creator: KEREMEDJIEV,MARK S
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ASTRONOMY -- EMCCD -- INSTRUMENTATION -- SMBH -- SPECKLE
Astronomy -- Dissertations, Academic -- UF
Genre: Astronomy thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: I present results from a new observational technique for ground-based astronomy called speckle stabilization. This technique is similar to other speckle-based techniques and is capable of producing diffraction-limited images in the optical, but has the added advantage of being able to employ an integral field spectrograph. Performance simulations show that a speckle stabilization system on a 10-meter class telescope should be capable of achieving resolutions as fine as 15 milliarcseconds in the optical. I also show that guide stars can be as faint as 16th magnitude and be located up to 30 arcseconds away. I present the design, fabrication and assembly of a prototype instrument the Stabilized sPeckle Integral Field Spectrograph Proof of Concept (SPIFS-POC) and describe the algorithms and programming necessary to control such a system and discuss optimization efforts. I compare speckle stabilization to other methods, including lucky imaging, and find that in certain circumstances, speckle stabilization is able to match or even outperform lucky imaging by up to a factor of 3 in signal-to-noise. Finally, this dissertation also covers the scientific gains speckle stabilization should be able to achieve. In particular, I address impacts on research in the field of super-massive black holes and demonstrate that the technique will fill an important niche in detecting the kinematic influence of SMBH on their host galaxies, aiding in the detection of the highest and lowest mass SMBH.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by MARK S KEREMEDJIEV.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Eikenberry, Stephen S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042840:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042840/00001

Material Information

Title: Achieving Diffraction-Limited Angular Resolutions in the Optical Through Speckle Stabilization
Physical Description: 1 online resource (148 p.)
Language: english
Creator: KEREMEDJIEV,MARK S
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: ASTRONOMY -- EMCCD -- INSTRUMENTATION -- SMBH -- SPECKLE
Astronomy -- Dissertations, Academic -- UF
Genre: Astronomy thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: I present results from a new observational technique for ground-based astronomy called speckle stabilization. This technique is similar to other speckle-based techniques and is capable of producing diffraction-limited images in the optical, but has the added advantage of being able to employ an integral field spectrograph. Performance simulations show that a speckle stabilization system on a 10-meter class telescope should be capable of achieving resolutions as fine as 15 milliarcseconds in the optical. I also show that guide stars can be as faint as 16th magnitude and be located up to 30 arcseconds away. I present the design, fabrication and assembly of a prototype instrument the Stabilized sPeckle Integral Field Spectrograph Proof of Concept (SPIFS-POC) and describe the algorithms and programming necessary to control such a system and discuss optimization efforts. I compare speckle stabilization to other methods, including lucky imaging, and find that in certain circumstances, speckle stabilization is able to match or even outperform lucky imaging by up to a factor of 3 in signal-to-noise. Finally, this dissertation also covers the scientific gains speckle stabilization should be able to achieve. In particular, I address impacts on research in the field of super-massive black holes and demonstrate that the technique will fill an important niche in detecting the kinematic influence of SMBH on their host galaxies, aiding in the detection of the highest and lowest mass SMBH.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by MARK S KEREMEDJIEV.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Eikenberry, Stephen S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042840:00001


This item has the following downloads:


Full Text

PAGE 2

2

PAGE 3

Intheprocessofdoingtheresearchandwritingthisdissertation,Ihavehadalotofhelpalongtheway.Myresearchwouldnothavegottenveryfarwithouttheaidofmyadvisor,StephenEikenberry.OndayonehehandedmeagreatprojectideathatIamfortunatetohaveworkedon.Hisadvice,guidanceandexcitementfortheeldhavebeenaninspiration.AnthonyGonzalezdeservesquiteabitofrecognitionforallofthehelphehasgivenme.Hisexpertiseinextragalacticmattersenabledmetopursueresearchinterestsmoreattunedtomyparticulartastes.Beyondthat,hisinsightsintoacademiaandcommentsonvariousproposalsandpapersthroughouttheyearshavebeenextremelyhelpful.Theinstrumentation-sideofthingswouldnothavegottenveryfarwithouttheguidanceofNickRaines.Hehasbeenindispensableinofferingadvice,ideasandplainoldfashioncommonsensewhenputtingmyinstrumenttogether.Withouthimandtheotherengineeringfolksonthefourthoor,myprojectwouldhavelikelytakenyearslongertonish.RebaBandyopadhyayissomeoneIwouldliketothankforlongdiscussionsonthemechanicsofactuallybeingascientist.Inadditiontothat,shehasalsoprovidedvaluableeditingadvicesandresearchprojectideas.IwouldliketothankthestaffattheKittPeakNationalObservatoryandtheWilliamHerschelTelescope.Installinganewinstrumentcanalwaysbetrickyandbothfacilitiesweremorethanhelpfulwhenproblemsarose.IwouldparticularlyliketothankDickJoyce,DiHarmerandIanSkillenfortheirassistance.Inaddition,BrunoFemeniadeservescreditforhelpingmewithEMCCDissues.JoeCarson'sfundamentalquestion:howdoesSPIFScomparetoLuckyforimaging?wasessentialtotheresultsfoundinChapter7.Hewasalsoagreatobservingpartnerfortherstweek-longSPIFS-POCrunattheKPNO2.1-meter.Onthesoftwareendofthings,IwouldliketothankCraigWarnerforallowingmetousehisprototypeFATBOYdataanalysissoftware.EricForddeservescreditforhisinsightfulcommentsonGPUsandendlesspatiencehelpingmelearntoprogramCUDA.Relatedtothat,NathanDe 3

PAGE 4

4

PAGE 5

5

PAGE 6

page ACKNOWLEDGMENTS .................................... 3 LISTOFTABLES ....................................... 9 LISTOFFIGURES ....................................... 10 ABSTRACT ........................................... 13 CHAPTER 1INTRODUCTION .................................... 14 1.1CurrentSpeckle-BasedSystems .......................... 14 1.2SpeckleStabilization ................................ 18 1.3UsingSpeckleStabilizationtoProbetheMassesofSMBH ............ 21 2SIMULATIONSOFSPECKLESTABILIZATION ................... 24 2.1SimulationDesign ................................. 24 2.2SimulationResults ................................. 26 2.2.1CoreFWHM ................................ 26 2.2.2StrehlRatios ................................ 27 2.2.3GuideStarMagnitudes ........................... 28 2.2.4Off-axisGuiding .............................. 30 2.3SimulationsofSS433 ............................... 33 3DESIGN,DEVELOPMENTANDTESTINGOFTHESPIFS-POC .......... 37 3.1PrincipalComponentsandPerformance ...................... 37 3.1.1Specklesensor ............................... 37 3.1.1.1Timing .............................. 38 3.1.1.2Readnoise ............................ 38 3.1.2Faststeeringmirror ............................. 41 3.1.3Sciencedetector ............................... 46 3.2OpticalDesign ................................... 46 3.3MechanicalDesign ................................. 49 3.4TurbulenceGenerator ................................ 55 4CONTROLLOOP .................................... 57 4.1OverallLoopDesign ................................ 57 4.2SpeckleSelectionAlgorithm ............................ 58 4.3LoopSpeedandOptimization ........................... 65 4.3.12DCrossCorrelations ........................... 65 4.3.2UsingaGPUtoSpeedUpConvolutionCalculations ........... 66 4.3.2.1Methodandresults ........................ 66 6

PAGE 7

............................ 67 4.4DriverInitialization ................................. 69 4.5Windowing ..................................... 70 4.6DetectorRotationandSolution ........................... 72 4.7FinalLoopSpeedandCodeOptions ........................ 73 4.8InitialLabResults .................................. 74 5OBSERVATIONSWITHTHESPIFS-POC ....................... 77 5.12009A/BObservationsattheKPNO2.1-m .................... 77 5.22010AObservationsattheWHT .......................... 79 5.2.1SingleStarObservations .......................... 80 5.2.2ObservationsofWDS14411+1344 ..................... 83 5.3SystemAccuracy .................................. 85 6FUTUREPLANSANDDEVELOPMENTOFTHES3D ................ 87 6.1LoopSpeedandLatency .............................. 87 6.2FSMAccuracy ................................... 88 6.3HighSpeedShutter ................................. 90 6.4ScienceChannelandADC ............................. 90 6.5S3D ......................................... 92 7ACOMPARISONBETWEENLUCKYIMAGINGANDSPECKLESTABILIZATIONFORASTRONOMICALIMAGING ........................... 94 7.1Methods ....................................... 94 7.1.1SpeckleStabilizationSimulations ..................... 94 7.1.2ComparisonBetweenMethods ....................... 96 7.2Results ....................................... 99 7.2.1ReadNoiseLimit .............................. 102 7.2.2PhotonCounting .............................. 103 7.2.3OptimalLuckyImaging .......................... 106 7.2.41024x1024pixel2Detectors ........................ 108 7.3Discussion ...................................... 110 8THEFUTUREOFSMBHDETECTIONVIAKINEMATICMODELINGASENABLEDBYELTS ......................................... 111 8.1Extremely-LargeTelescopes ............................ 111 8.1.1TheoreticalImprovementsOverExistingFacilities ............ 112 8.1.1.110-meterresolutions ....................... 112 8.1.1.2ELTdiffraction-limitedresolutions ............... 113 8.1.2K-bandCOBandheads ........................... 114 8.1.3H-bandCOBandheads ........................... 117 8.1.4CalciumTripletLines ............................ 118 8.1.5NumberofGalaxies ............................. 120 7

PAGE 8

................... 122 8.2SPIFSImpactonSMBHMeasurements ...................... 125 8.3Rest-FrameJ-bandSpectralFeatures ........................ 126 8.3.1ObservationalRequirements ........................ 127 8.3.2ObservationalConrmation ........................ 130 8.3.2.1ObservationsandDataReduction ................ 131 8.3.2.2Results .............................. 132 8.4Discussion ...................................... 135 9DISCUSSIONANDCONCLUSIONS .......................... 137 9.1SimulatedandActualPerformance ......................... 137 9.2ScienticPotentialofSS .............................. 138 9.3AdvantagestoSpeckleStabilization ........................ 139 9.3.1ComparisontoAdaptiveOptics ...................... 139 9.3.2ComparisontoStandardSpeckleImaging ................. 140 9.3.3Space-basedPlatforms ........................... 141 9.4FutureDirectionsforSpeckleStabilization .................... 142 REFERENCES ......................................... 143 BIOGRAPHICALSKETCH .................................. 148 8

PAGE 9

Table page 2-1SummaryofSSPerformanceSimulations ........................ 34 4-1Listofstarsandparametersusedforspeckletesting ................... 59 4-2ComparisonbetweenBPand2DCC ........................... 60 4-3Effectsofthevariouswindowingparameters ....................... 72 4-4Componentsofloopspeed ................................ 74 7-1Summaryofsimulationresults .............................. 108 8-1ELTK-bandresolutionsanddistancesatwhichtypical(1108M)SMBHshouldbeobservable. ....................................... 114 8-2Listofgalaxies,theirSMBHmassesandcorrespondingredshifts ............ 116 8-3ELTH-bandresolutionsanddistancesatwhichSMBHshouldbeobservable. ..... 118 8-4Observedgalaxyproperties ................................ 132 9

PAGE 10

Figure page 1-1Demonstrationofsimulatedseeinglimitedimages. ................... 16 1-2Stabilized-specklesystemconceptualschematic. ..................... 19 1-3CrosssectionsofPSFsproducedbyspecklestabilizationandintheseeinglimit. .... 20 2-1SimulationsFWHMvs.wavelength ........................... 27 2-2SimulationsofStrehlratiosvs.wavelength ....................... 28 2-3Strehlratioversusguidestarmagnitudeasafunctionofr0 29 2-4Requiredguidestarmagnitudes .............................. 31 2-5EffectofguidestaroffsetonFWHM ........................... 32 2-6EffectofguidestaroffsetonStrehlratios ........................ 33 2-7OpticalspectrumofSS433usedfortheSPIFSsimulations ............... 35 2-8SpectraextractedfromasimulatedSPIFSdatacubeforSS433 ............. 35 2-9SimulatedfalsecolorimageofSPIFS-resolvedjetoutowsfromSS433 ........ 36 3-1MeasurementofthecycletimesproducedbytheAndoriXon860 ........... 39 3-2PlotofvarianceversussignalformultipleEMgainvalues ............... 40 3-3ReadnoiseasafunctionofEMGain ........................... 40 3-4PlotofFSMringingbetweenshifts ............................ 43 3-5Exampleofthetwo-stepsolution ............................. 44 3-6Dampedharmonicoscillatortstotheringingdata ................... 45 3-7Ringingsolutionappliedtoa16-pixelshift ....................... 46 3-8ZEMAXdrawingsoftheopticaldesignfortheSPIFS-POC ............... 50 3-9ThetheoreticalperformanceoftheopticalsystemasdeterminedbyZEMAX ..... 51 3-10SolidWorksdrawingoftheholderfortheEMCCDcameraoptic ............ 52 3-11SolidWorksdrawingoftheentireSPIFS-POC ...................... 53 3-12SolidWorksdrawingsoftheSPIFS-POChandingcard .................. 54 3-13Articialspecklepatternproducedbythehairsprayphasescreen ............ 55 10

PAGE 11

............................ 61 4-2ContourplotsofthecorrelationbetweenspecklelocationsselectedwiththeBPand2DCC ........................................... 62 4-3Cumulativedistributionspecklelocationsdifferences .................. 63 4-4ComparisonofGPUandCPUperformanceforcomputingrawFFTs .......... 68 4-5ComparisonofGPUandCPUperformanceforcomputingfullFFTs .......... 69 4-6Imagesofin-labspecklestabilization ........................... 75 4-7CrosssectionsofthePSFsproducedinlaboratoryspecklestabilization ......... 75 5-1ImagesoftheSPIFS-POConobservingruns ....................... 78 5-2SPIFS-POCobservationsof65tauCyg ......................... 79 5-3ObservationsoftwoPSFstarswithbothSPIFSonandoff ............... 81 5-4ObservationsofthebinarystarWDS14411+1344 .................... 84 5-5AnalysisoftheaccuracyoftheSPIFS-POC ....................... 86 6-1PlotofPSFcross-sectionsasafunctionoflatency .................... 88 6-2SimulatedSSPSFcross-sectionsusingdatafromtheEMCCD ............. 89 6-3PSFcross-sectionasafunctionoffractionofimagesused ................ 91 6-4Plotofdifferentialrefractionasafunctionofwavelength ................ 92 7-1CrosssectionsofthePSFproducedbySpeckleStabilization .............. 96 7-2ComparisonbetweenSSandSS+SwithSAAandLuckyImaging ........... 101 7-3ComparisonbetweenSpeckleStabilizationandno-read-noiseSAAandLuckyImaging 104 7-4Comparisonbetweentechniquesinthecaseofphotoncounting ............. 105 7-5ComparisonbetweenspecklestabilizationandOLI ................... 107 7-6ComparisonbetweenSSandOLIfor1024x1024pixel2detectors ............ 109 8-1PlotofSMBHMassversuslimitingobservableredshiftforKeckandthethreeELTsinK-band ......................................... 116 8-2PlotofSMBHMassversuslimitingobservableredshiftforKeckandthethreeELTsusingvariouslinediagnostics ............................... 119 8-3MassfunctionofobservableSMBHasafunctionofSMBHmassforKeckandtheELTs ............................................ 121 11

PAGE 12

. 122 8-5PlotofSMBHMassversuslimitingobservableredshiftforthefourrstgenerationinstrumentsfortheELTs ................................. 124 8-6PlotofSMBHMassversuslimitingobservableredshiftwithSPIFS .......... 126 8-7MassfunctionofobservableSMBHincludinggalaxiesobservablebySPIFS ...... 127 8-8YandJ-bandspectraofthegalaxiesinthesample .................... 133 8-9J-bandspectraofNGC7619andthebest-tmodels ................... 134 12

PAGE 13

13

PAGE 14

14

PAGE 15

1-1 (a)).Astheturbulenceprolebetweenthetelescopeanditstargetshiftsandchanges,typicallyontimescalesofafewtoafewtensofmilliseconds,thespecklepatternalsoevolvesandshifts.Thus,theaverageimageofastarpropagatedthroughtheatmospheresignicantlyblursontimescalesof1-second(Figure 1-1 (b)).Thisiswhyastronomersoftentakefocusexposuresofafewseconds,inordertoaverageovertheseeing.Whilestandardtip/tiltsystemscanimprovethisspread,theyusethecentroidoftheentireenergyenvelopeforsteering,resultingintypically10%improvementovertheseeing,butstilllosingthehighestangularresolutioninformation.Whatisofnote,however,isthateachindividualspecklewithinthepatterncontainsinformationaboutthetargetsourceatangularscalesashighasthediffractionlimit.Thishasledtothespeckle-basedtechniquesdiscussedbelowandisthebasisofspecklestabilization.Diffraction-limitedimagesintheopticalregimeremaindifculttoachieveusingconventionaladaptiveopticstechniques.Therefore,obtainingresolutionsmatchingorbeatingtheHSTrequiresclevertechniques.Oneoftheearliestattemptstoovercomeatmosphericturbulenceintheopticalwasthroughthespeckleshift-and-add(SAA)methoddevelopedby Bates&Cady ( 1980 ).Inthistechnique,thefactthattheatmosphereiscoherentfor1030ms( Kernetal. 2000 )isexploitedbytakingthousandsofimageswithexposuretimesonthesameorderasthecoherencetime.Intheresultingspecklepatterns,thehighestqualityspeckleisfoundandthe 15

PAGE 16

Demonstrationofsimulatedseeinglimitedimages.(a)representsashortexposure(10ms)while(b)isalongexposure(1second).ImageswereproducedwiththecodeexplainedinChapter2.Theysimulateobservationsona10-metertelescopewithseeingofr0=15cm.Bothimageshavethesamescale.Intheseeingdiskontheleft,thehigh-spatial-frequencyinformationhaslargelybeenblurredoutbytime-averagingeffects. imagesarestackedontopofoneanothersuchthatthebestspecklesarealwaysinthesamelocation.Theresultsofsuchworkisadiffraction-limitedcoreatopadiffusehalooflight.Luckyimaging(LI)isarecenttechniquethatborrowsthefundamentalapproachoftheSAAbutfurtherimprovestheimagequality.Itwasrstproposedby Fried ( 1978 )andhasbeenimplementedbyseveralgroupswithsuchinstrumentsasLuckyCam,FastCamandAstraLux( Law 2007 ; Oscozetal. 2008 ; Hormuthetal. 2008 ).Theideabehindthetechniqueisrelativelysimple.Sincetheturbulencepatternischaoticandobeysapower-lawdistributionbest-describedby Kolmogorov ( 1941 ),temporaluctuationsintheturbulencemeanthatoccasionallyneardiffraction-limitedimagesnaturallyoccur.Thesepocketsarenormallyshortlivedlastingonlyafewtensofmillisecondsandoccuronly10%ofthetimeon2.5-meterclasstelescopes.Therefore,toexploitthesehighqualitypocketsLItakestensofthousandsofimagesbutonlykeepsthehighestqualityimages.Thesehigh-qualityimagesarethenshiftedandaddedtooneanotherinthesamewayastheSAAmethod.TheresultisaPSFsimilartoSAAbutwithhigherStrehlratios( Baldwinetal. 2001 ; Tubbsetal. 2002 ; Lawetal. 2006 ; Law 2007 ). 16

PAGE 17

Fried ( 1978 )rstproposedtheideaofluckyimaging,detectorswereslowandtogettherequiredsignal-to-noiseratiosinvolvedobservationsspanningmanynights.TheadventofElectron-MultiplyingCCDs(EMCCDs)hasenabledLItobecomeareality.Thesedetectorsarebothhighlysensitiveandabletoreadoutquickly.A512x512pixel2EMCCDisabletoreadoutat100Hzasopposedtoatraditionaldetectorwhichreadoutsat<1Hz.Thereforeitispossibletogettensofthousandsofimagesinlessthananhour.Therstgrouptoperformluckyimagingwas Baldwinetal. ( 2001 ),whoachievedneardiffraction-limitedangularresolutionsonthe2.56-mNordicOpticalTelescopeat0:8mm.Sincethattimetherehavebeenobservationsofbinarystarsystems( Lawetal. 2005 2006 2008 ),exoplanethoststars( Daemgenetal. 2009 )andmanyotherscientictargets.WhileLuckyImaginghasbeensuccessfulatachievingdiffraction-limitedangularresolutionsintheoptical,thisfeatislimitedtotelescopeswithapertureslessthan4-metersindiameter.Thereasonforthisisthefractionofusableimagesdecreasesasafunctionoftelescopediametersuchthatat4-morlarger,thefractionofusableimagesapproacheszero.However, Lawetal. ( 2009 )developedawaytoincreasetheusefulnessofLuckyImagingtotheselargertelescopes:usingLuckyImaginginconjunctionwithanAdaptiveOpticssystem.CurrentAOsystemsareverysuccessfulatcorrectingatmosphericturbulenceintheNIRmuchofthetime.AOsystemsachievethisbyusingadeformablemirrorcorrectingforturbulenceatratesashighas1kHz.However,turbulenceismoresignicantintheopticalandrequirescorrections>1kHz.Additionally,thenumberoftermsrequiringcorrectionissignicantlyhigherintheopticalmeaningthatvisibleAOsystemswouldrequiremanymoreactuatorsthanarecurrentlyemployed.WhilenoneofthesefactsposeatheoreticalreasonopticalAOisimpossible,itisnotpracticalwithcurrenttechnology.What Lawetal. ( 2009 )notedhowever,isthatevenwiththeseshortcomings,AOstillproducesnearlydiffraction-limitedimagessomeofthetimeintheoptical.Furthermore,thefractionofthetimeAOsucceedsincorrectingopticalimagesissimilartothefractionoftime 17

PAGE 18

Lawetal. ( 2009 )atthePalomar200-inchtelescopeusingtheexistingfacilityAOsystemandachievedthehighestangularresolutioneverrecordedinopticalobservations. Eikenberryetal. 2008 ; Keremedjievetal. 2008 2010 ).SSwillenable,undercertainobservationalconditionsandconstraints,low-to-modest-Strehldiffraction-limitedimagingspectroscopyfromlargeground-basedtelescopesintheopticalbandpass(e.g.r',i',andz'bands).Whencoupledwithanintegraleldspectrograph,SSiscapableofexploringimportantscienticnicheswhicharenotcurrentlyavailableusingexistinghighangularresolutiontechniquessuchasadaptiveopticsorluckyimaging.Atthesametime,specklestabilizationisrelativelystraightforwardandfeasibleusingexisting,relatively-inexpensivetechnology.TheStabilizedsPeckleIntegralFieldSpectrograph(SPIFS)iseffectivelyareal-timeSAAsystemcoupledwithanIFS( Eikenberryetal. 2008 ).Previousspeckletechniques(SAAandLI)generallyrelyonoff-linepost-processingtorecoverthehigh-spatial-frequencyinformationbutthisinturntypicallylimitsthemtosimpleimaging,usingthesamedetectorforscienceasforspecklesensing.IntheSStechnique,weproposetouseafaststeeringmirror(FSM)tostabilizethepositionofthebrightestspeckleinreal-time.Thisthenenablesustofeedastableimagewithadiffraction-limitedcoreintoanalternatesciencechannel(suchasanintegraleldspectrograph)viapickoffand/ordichroicmirrors.Thisthendecouplesthehigh-speed(andtypicallylower-sensitivity)specklesensingfromthesciencechannel,enablinganalyses 18

PAGE 19

Stabilized-specklesystemconceptualschematic.Showstheprimarylayoutandhowthesciencechannelisdecoupledfromthespecklesensingchannel. requiringlongexposureswithmoresensitiveslowdetectors(i.e.spectroscopy,polarimetry,etc.).Ipresentanideal,cartoon-levelschematicforaspecklestabilizationsysteminFigure 1-2 .Lightfromatelescoperstencountersafast-steeringmirror(FSM)andisthenrelayedtotheremainderoftheinstrumentafterpassingthroughanatmosphericdispersioncorrector(ADC).Adichroic(or,alternately,asmallpickoffmirror)relaysthesciencebeamtothescienceinstrument,whiletheremainderofthelightcontinuesintothespecklesensor.ThespecklesensorconsistsofopticsandanEMCCDessentiallyidenticaltoaspeckle-imagingsystem.However,ratherthansimplyrecordingthespecklepatternsforeventualoff-lineanalysis,thespecklecameraimagesareanalyzedinreal-timetoidentifythelocationofthebrightestspeckleinthesystem.Thisinformationisthenusedtogenerateacommandsignaltothefast-steeringmirrorwhichkeepsthisbrightestspecklestablylocatedataconstantpositionontheoutputfocalplane. 19

PAGE 20

CrosssectionsofPSFsproducedbyspecklestabilizationandintheseeinglimit.ThecodeusedtoproducethesePSFsisdescribedinChapter2.Notethatthereisasharp,diffraction-limitedcoresittingatopaseeinglimitedhalo. TheprocessofusingaFSMtostabilizespeckleseffectivelyachievesareal-timeshift-and-addofthespecklepatterns.Whileallotherspecklesarestillmovinginanuncontrolledmanner,averagingouttoasmoothhalo,thestabilizedspeckleproducesasteadydiffraction-limitedcorewhichisthenanalyzedandrecordedbythescienceinstrument,muchasanadaptiveoptics(AO)scienceinstrumentanalyzes/recordsanAO-correctedimage.IshowacrosssectionofthePSFproducedbyspecklestabilizationinFigure 1-3 anddemonstratethatSSproducesasharp,diffraction-limitedcoreatopadiffusehalo.Whencomparedtotheseeinglimitinred,itisclearSSproduceshighStrehlratiosandhigherresolutions.Thisapproachovercomesoneofthemajorlimitationsoftraditionalspeckleimaging:thespecklesensorsystemisdecoupledfromthescienceinstrumentinsuchawaythatthescienceinstrumentcaneffectivelycarryouthigh-sensitivityobservationsoftargetsusingtechniquesrangingfromstandardimagingtointegral-eldspectroscopytopolarimetry,etc.Thelattercan 20

PAGE 21

Keremedjievetal. ( 2008 ).InChapter3IcoverthemechanicalandopticaldesignwhileChapter4isonthecontrolloop.Idiscusstheon-skyobservationsoftheinstrumentinChapter5andcommentonfuturedirectionsforthetechniqueinChapter6(muchofthecontentappearedin Keremedjievetal. ( 2010 )).InChapter7,adaptedfrom Keremedjiev&Eikenberry ( 2011 ),Icomparespecklestabilizationtootherspeckleimagingtechniquesanddiscussstrengthsandweaknesses. vanderMarel ( 1994 )proposedthatSTISontheHubbleSpaceTelescopecouldbeusedtoderiveblackholemasses,high-spatialresolutionsinconjunctionwithspectroscopicinformationhaveproventobeessentialtoworkinthiseld.DozensofthesemeasurementsusingHSTandAO+IFSobservationsongroundbasedtelescopeshavebeenmadeby Gebhardtetal. ( 2000 ); Pinkneyetal. ( 2003 ); Daviesetal. ( 2006 ); Cappellarietal. ( 2009 ); McConnelletal. ( 2011 )andothers.However,duetoinstrumental/facilityconstraintsandthe 21

PAGE 22

Ferrarese&Merritt 2000 ; Gebhardtetal. 2000 ),acorrelationbetweenthemassofthecentralblackholeinagalaxyandthevelocitydispersionofstarsinthehostgalaxy,hasbeenoneofthefundamentaldiscoveriesinourunderstandingofSMBHproperties.TherelationshipdrivesresearchintothecoevolutionofgalaxiesandtheirSMBHastheblackholesphereofinuenceR=GMBH=s( Peebles 1972 )istoosmall(ontheorderof10pc)toinuencethelargerrotationofthehostgalaxyonkpcscales.TherelationiswellcharacterizedforSMBHmassesof106108MbutatboththehighandlowmassendoftheMBHsrelationtherearelingeringquestionsthatrequirenewobservationstoaddress.OnesuchissueisthedisjointbetweenSMBHmassespredictedbytheMBHsrelationandthemasspredictedbytheblackholemass-galaxyluminosityMBHLrelation( Kormendy&Richstone 1995 ; Magorrianetal. 1998 ).Partofthedivergentmassestimatedbythetwotechniquesisdrivenbytheintrinsicscatterinbothcorrelationswhichincreaseathighermass( Law 2007 ),butthefactthatwedonotmeasuregalaxieswithsmuchgreaterthan400km/s( Shethetal. 2003 )indicatesapointatwhichtheMBHsrelationplateaus.SincetheMBHLrelationdoesnothavethesamelimitationathighluminosity,thetwotheoriespredictdifferentmassesatMBH>109M.ArecentexampleofthisdisjointcanbefoundintheinvestigationofNGC1332by Ruslietal. ( 2011 )wherethereisanorderofmagnitudedifferencebetweenblackholemassespredictedbythetworelations.AtthelowmassendoftheMBHsrelationthereisdebateoverwhetherintermediatemassblackholes(thosewith103105M)exist.EvidenceofblackholesattheboundarybetweenthetwoatMBH105Mhasbeenfoundby Greene&Ho ( 2007 ); Filippenko&Ho ( 2003 ); Petersonetal. ( 2005 ); Barthetal. ( 2004 )inAGNusingsingleepochobservationsandreverberation 22

PAGE 23

Gebhardtetal. ( 2005 )usekinematicmodelinganddetecta2104MIMBHinG1,astrippedgalaxycorearoundAndromeda.IntheglobularclusterOmegaCentauri Noyolaetal. ( 2010 )detecta5104MIMBH,but vanderMarel&Anderson ( 2010 )disputethisclaimwithpropermotionmodelingandplaceanupperlimitontheblackholemassat<7103M.AsthedebatecontinuesovertheexistenceofIMBH,theirlocationontheMBHsrelationalsoremainsandopenissue.Consequently,thesequestionsatthehighandlowmasslimitshindereffortstodevelopacomprehensiveblackholemassfunction(BHMF).OnceafullunderstandingoftheMBHsrelationisavailableatallmassranges,thenaBHMFcanbeconstructed.ItthenbecomesusefultoseeiftheBHMFchangesovercosmictimeandsuchinformationwillprovidevaluableinsightintoSMBHandgalaxyformationandevolution.ToaddressthemostmassiveSMBH,understandthenatureofIMBH(orlackthereof)anddevelopaBHMF,weneedaccesstohighangularresolutionIFSinstrumentation.WhileSTISonHSThasbeenrepairedandcurrentAO-fedIFUworkhasbeenconductedonsomelocalgalaxies,itisdifculttogetthenecessarytimetoconductthesecomplicatedandtime-intensiveobservations.Furthermore,thesamplesizeisstronglylimitedbytheresolvingpowerofthetelescopeaperture.Aneffectivemeanstoaddresstheseissuesisthroughnewtelescopes,instrumentationandtechniques.InChapter8ofthisdissertation,IwilladdresshowSPIFSandtheadventofextremelylargetelescopeswillimpactthisresearch.SincetheybothpresentthecapabilitytomeasureSMBHathigherangularresolutionsitisimportanttoseewheretheycomplimentoneanother.Finally,IwillsummarizethendingsinChapter9anddiscusstheoverallbenetsofspecklestabilizationasnotedin Eikenberryetal. ( 2008 ).Forthisinvestigation,IassumedH0=73:0km=s=Mpc( Freedman&Madore 2010 )andaatLCDMwithWM=0:3. 23

PAGE 24

Keremedjievetal. ( 2008 ).Inx2.1Icoverhowthesimulationswereconductedandwhatparameterstheyinvestigate.Inx2.2,IpresentresultsshowingFWHM,Strehlratios,limitingguidestarmagnitudesandguidestaroffsetinformation.Finallyinx2.3,IaddressthespeciccaseofthemicroquasarSS433andshowhowSSwillenableuniquemeasurementsofjetproperties. Kolmogorov 1941 )intheInteractiveDataLanguage(IDL).ThefundamentalideabehindthecodeisthatthewavefrontW(a;b)atthepupilplaneistheintegraloverallfrequencies.Thiscanbedescribedinonedimensionas: 24

PAGE 25

LeLouarn ( 2002 ).IthendeneanamplitudemaskforthepupilA(x;y)matchingthedimensionsofthephasemapf(a;b).A(a;b)issetto1foreverylocationintheunobscuredtelescopepupil,and0everywhereelse.Itisassumedforthepurposesofthesesimulationsthatthesecondarymirrorobscures1/3ofthepupil.ThenalwavefrontmapisthengivenbyW(a;b)=

PAGE 26

2.2.1CoreFWHMIsimulated100specklepatternsforarangeofwavelengths0:5mm
PAGE 27

B FWHMinmasasafunctionofwavelengthforr0=15cmforboth5-meterand10-meterclasstelescopes,alongwiththetheoreticaldiffraction-limitedFWHM.Theslightdiscrepanciesbetweentheoryandsimulationareprimarilyduetosamplingeffectsnearthediffraction-limitandthetuningparameterbinthesimulations. 2-2 .Asexpected,performanceimproveswithincreasingr0effectively,thelightisdividedamongstfewerspeckles.Furthermore,weseethatSSachieves>1%Strehlratioindecentseeingconditionsandthatingoodseeingconditions(r0=2025cm,slightlyhigherthanthemedianGTCsiter0of18cm)SSwillachieveStrehlratios>23%. 27

PAGE 28

2-2 ,Ipresenttheresultsofsimulationsfora5-meterclasstelescope.Becausethetelescopeentrancepupilisdividedintofewerr0-diameterpatchesthaninthe10-mcase,largerfractionalenergyispresentineachspeckle.Asaresult,theStrehlvaluesare2-3xhigherthaninthe10-metercaseatagivenr0.Itisworthnoting,however,thatthetotalenergyinthecentralcoreStrehlD2isgreaterforthe10-mtelescope.Thesevaluesarebroadlyconsistentwithluckyimagingobservations. Tubbsetal. ( 2002 )measuredStrehlratiosof0:06inLuckyImagingobservationswhenusing100%frameselectionona2.5-mtelescopeinthei'lter.WealsondthattheSSPSFisquitesimilartoluckyimagingPSFsproducedinrealobservationswith100%frameselectioni.e.adiffraction-limitedcoreatopadiffusehalo. B (a)SSStrehlratiosasafunctionofr0andwavelengthfora5-metertelescope.(b)SSStrehlratioasafunctionofr0andwavelengthfora10-metertelescope. 28

PAGE 29

2-3 ,IshowthatSPIFSwillbeabletouseguidestarsasfaintas15thmagnitudeintypicalseeingconditions(r015cm)ona10-mtelescopewithoutsignicantdegradationofStrehlratios.Inotherwords,guidestarshotnoiseonlybecomesanissueforstarsof1415magorfainter.Thisis1.5-2magnitudesfainterthanwouldtypicallybeusableforadaptiveopticscorrection.ThisisexpectedsinceAOsystemssplitthewavefront-sensinglightinto>200Shack-Hartmanncellsandmustmeasurethecentroidforeachofthese.SS,ontheotherhand,onlyrequiresanaccuratelocationofthebrightestspeckle,whichcontains12%ofthetotalux(severaltimesbrighter)fortypicalturbulenceconditions. SSStrehlratioversusguidestarmagnitudeasafunctionofr0ona10-mtelescope. ThesimulationsusedinFigure 2-3 assume100%frameselection.Ifahigh-speedshutterisusedsuchthatonlythebest10%ofimagesareused(akintoluckyimaging)itispossiblethatfainterguidestarsmaybeusable.Thebasicideawouldbethattheshutteropensonlywhenwe 29

PAGE 30

2-4 (a)showstheresultingperformanceasafunctionofguidestarmagnitudefora10-mtelescope.WecanseethattheshutterapproachwithfainterguidestarswillactuallyimprovestheimagequalityoftheSPIFStechniquewithfaintsourcesattheexpenseofintegrationtime.ThisimprovementinimagequalityoccursbecausetheshutterisonlyopenforspecklepatternswiththehighestStrehlratiosresultinginanimprovementintheoverallSSimage.Italsosignicantlyimprovesskycoverage(by50-100%)whilestillmaintainingthehighangularresolutionwhichistheprimarydriverforSPIFS.Wealsonotethattheshutterwillbeacriticalfeatureforobservingundervaryingseeingconditions.Thiswillbetheruleratherthantheexceptionforrealobservations(whilethesimulationshereassumeaconstantvalueofr0forthesakeofsimplicity).AdditionaladvantagestoshutteringarediscussedinChapter6. 30

PAGE 31

B (a)Pseudo-Strehlratioversusguidestarmagnitudeinthecaseofahigh-speedshutteremployingasignal-to-noisethresholdforallowinglighttoenterthesciencechannel.(b)Openshutterfractionversusguidestarmagnitudeforthesamesituation.(c)Fractionalenergyinthecoreversusguidestarmagnitude(equaltotheproduceofthecurvesinaandb).Whilethetotalenergycontinuestodropwithmagnitude,notethatthecontrastbetweenthecoreandtheassociatedhaloofthePSFremainshighforfainterguidestarsusingthisapproach(ascomparedtoFigure 2-3 ). locationsrangingfrom1to120with100specklepatternsproducedateachoffset.Simulationswereconductedforthreewavelengths(l=0:5;0:75;1:0mm)andthreeseeingconditions(r0=10;15;20cm)fora5-and10-metertelescope.TheresultsarepresentedtheresultsinFigures 2-5 and 2-6 .FromtheFigure 2-5 ,wecanseethatSSPSFsshowverylittleFWHMdegradationoveroff-axisanglesaslargeas20inradiusundernormalseeingconditionsatmosttelescopefacilities.WecanalsoseeinFigure 2-6 thattheStrehlratiosdegradewithoff-axisangle,butthat 31

PAGE 32

SPIFSFWHMasafunctionofoffsetfromguidestar.Eachpanelcorrespondstoadifferentr0.Thedottedlinedenotesthevaluegivenbytheguidestarforeachiterationandprovidesaconstantcomparisonagainstthesciencechannel.NotethatthesmalldifferencesbetweenguideandscienceStrehlsratiosatzerooffsetareduesimplytodifferentrealizationsofthesamenoisedistributionintheMonteCarlosimulations.ThedramaticdipinFWHMfor0:7mminther0=10cmcaseissimplyduetothettingroutinefortheFWHM;itassumestheFWHMis<100soitreturnsnon-physicalresultswhenthetrueFWHMis>100. atqmax20,thePSFsmaintain60%oftheon-axisStrehlvalue.Thus,Iadoptqmax=20asaworkingestimatefortheoff-axispatchsize.Basedontheseresults,IcanestimatetheapproximateskycoverageforSS.For15thmagguidestarsand20usefulradius,wecanexpectskycoverageof50%forlowGalacticlatitudes,withthefractionbeinghigher(100%)towardstheinnerGalaxy,andlower(12%)towardstheGalacticanti-center.For16thmagguidestars,theskycoverageincreasesto>30%eventowardstheGalacticanti-center.ForhighGalacticlatitudes,theskycoveragefraction 32

PAGE 33

SPIFSStrehlasafunctionofoffsetfromguidestar.Eachpanelcorrespondstoadifferentr0.Thedottedlinedenotesthevaluegivenbytheguidestarforeachiterationandprovidesaconstantcomparisonagainstthesciencechannel.NotethatthesmalldifferencesbetweenguideandscienceStrehlratiosatzerooffsetareduesimplytodifferentrealizationsofthesamenoisedistributionintheMonteCarlosimulations. naturallydrops,butremains>3%for15thmagstars(>7%for16thmag)evenattheGalacticpolarcap.Thus,weconcludethattheSSskycoveragewillbeverycompetitivewithnaturalguidestaradaptiveopticssystems,andevenwithlaserguidestarsystemsneartheGalacticPlane.AsummaryofStrehlratiosandexpectedskycoverageisgiveninTable 2-1 33

PAGE 34

SummaryofSSPerformanceSimulations 10-mTelescope 5-mTelescope Parameter r0=20cm r0=10cm r0=15cm 12.7mas 13.0mas 26.8mas 27.8mas Strehl(0:75mm) 0.009 0.013 0.013 0.024 Guidestarmag 15.0 15.5 15.5 16.0 Off-axisangle SkyCoverageGalacticPlane SkyCoverageGalacticCap microquasarastrophysics.Theopticalspectrumofthisobjectshowsanumberofstrong,broademissionlinesoftheBalmerandHeIseries,aswellasseverallinesatunusualwavelengths.Theselatterhavebeenidentiedasredshifted/blueshiftedBalmerandHeIemissionfromcollimatedjetswithintrinsicvelocitiesofn0:26c( Abell&Margon 1979 )andanexamplespectrumispresentedinFigure 2-7 .Furthermore,theDopplershiftsofthesefeatureschangewithtimeinacosinusoidalmanner,leadingtothelabelofmovinglines.ThisbehaviorisnowwidelyacceptedtobeasymptomofprecessionofthejetaxisinSS433onatimescaleof164days( Margon 1984 ).WhileradioobservationswithVLBIcanresolvetheradiojetsandtracktheirpropermotions,theopticaljetsarisemuchclosertothecompactobjectandarethusonlyobservableviatheirspectralfeatures.Spatiallyresolvingthesejetlineswouldprovidetremendousinsightsintothephysicalconditionsintherelativisticjets.BasedonthekinematicmodelforSS433(summarizedin Eikenberryetal. ( 2001 )),weexpectthejetstoappearapproximately22milli-arcsecondsfromthecompactobject.NotethatthisisnotresolvablebycurrentAOsystemson10-meter-classtelescopesoperatingattheshortendoftheirbandpass.SinceSS433hasamagnitudeofI=12mag,itisahighlysuitabletargetforSPIFSobservations.IcreatedasimulatedSS433SPIFSimageusinga10-metertelescopewithmedianr0=15cmbasedonactualopticalspectraofSS433.Iusedthesimulationsoftwaretogeneratemulti-wavelengthPSFsandsimulatedspecklestabilizationusingSS433itselfastheguidestarwithadichroicdivertingthesciencelighttotheanintegraleldspectrograph.IthenconvolvedtheresultingPSFswithaninputspectrumdissectedintoanassumedcore+2jets 34

PAGE 35

OpticalspectrumofSS433usedfortheSPIFSsimulations.Notethatboththeapproachingandrecedingjetsareredshiftedcomparedtorest-frameHduetorelativistictimedilationeffects.ThefeaturebetweenthestationarylineandthebluejetisastationaryHeIline. SpectraextractedfromasimulatedSPIFSdatacubeforSS433atthespatiallocationsofthecentralcompactobjectandthetwojets.Notethattheblue/redjetcomponentsarecleanlyseparatedattheresolutionofSPIFS. 35

PAGE 36

2-8 .Notethesignicantspectraldiversitybetweenthethreespectra,allowingtheclearidenticationofthespatiallocationsfromwhichparticularfeaturespredominantlyarise-thecriticalfactorfortheseobservations.Finally,weselected3spectralchannelsfromthedatacubeandcodedthemred(forthewavelengthchanneloftheredjet),blue(forthewavelengthchannelofthebluejet),andgreen(foracontinuumwavelengthbetweenthetwo)tocreateafalse-colorsimulatedimageofthespatially-resolvedjetsinSS433(Figure 2-9 ).ThesesimulationsrevealthatitshouldbepossibletospatiallyresolvetheopticaljetsofSS433usingthetechniqueofspecklestabilization.Sincenoothertechniquecurrentlyexiststhatiscapableofthisfeat,itpresentsaninterestingsciencecasewhichSSiswell-suitedtoaddress. B (a)FalsecolorimageofSPIFS-resolvedjetoutowsfromSS433,withred(forthewavelengthchanneloftheredjet),blue(forthewavelengthchannelofthebluejet),andgreen(foracontinuumwavelengthbetweenthetwo).(Right)Continuumsubtracteddifferenceimagesmadefromthe3imagesintheleftcomposite.Notethecleanseparationofthejetcomponent,whichareseparatedby45-milliarcsecondsontheskyinthissimulation. 36

PAGE 37

37

PAGE 38

3-1 .Therethecycletimemeasuredcloselymatchesexposuretimeplusreadtime.Infact,themeasureddifferencebetweenthetwois<10msec.Therefore,thedetectormeetsspecicationsintermsofexposuretimesandreadouttimes. Howell 2000 ).Togetsignalversusvariance,ItookatexposureswithvariableintegrationtimessuchthatIhadexposuresrangingfromunder-sampledtosaturated.Ithencalculatedthesignal(medianbias-subtractedintensity)andvarianceforeachimage. 38

PAGE 39

MeasurementofthecycletimesproducedbytheAndoriXon860.Theredlinedenotesthecaseofnoread-outtime,hencetheonlycontributiontothecycleistheexposuretime.Theblacklineisthemeasuredcycletimeandthedottedlineisthereadtimeofthedetector.TheresidualsbetweentheIdeal+ReadTimeandMeasuredarelessthan10msec. IgiveaplotofvarianceversussignalforthedifferentEMgainvaluesinFigure 3-2 andpresentthelineartstothelinearregimeoftheplotsinFigure 3-2 .Fromthegure,IshowthatwherethelinearregimeactuallyoccursdependsontheEMgainvalueandthatthereisatrendbetweendecreasingslopeasafunctionofEMGain.WhatisinterestingisthatthesizeofthelinearregimealsoincreaseswithEMgainandatEMgain=250nearlytheentiredynamicrangeofthedetectorislinear.Ipresenttheslopesofthets(readnoise)versusEMgaininFigure 3-3 .ThetwocurvesdenotedifferentbiasmeasurementsfortheEMCCDandtheyareconsistentwithoneanother.ThesedataconrmAndor'sstatementthattheeffectivegainoftheDU-860convergesto0:2e/pixelathighEMgains.ThisextremelylowreadnoiseindicatesthatthedetectorshouldbesensitivetofainttargetsathighEMgains. 39

PAGE 40

PlotofvarianceversussignalformultipleEMgainvalues(EMGain=0isthecasewhereonlythenormaldetectorgainispresent).Thetsarealltodatawhichappeartobeinthelinearregimeofthedata.Slopeofthetsprovidereadnoiseinformation. ReadnoiseasafunctionofEMGain.Thedataappeartoconvergetoward0:2e=pixathighEMgain.Thetwodifferentcurvesarefromdifferentbiasmeasurements. 40

PAGE 41

3-4 .RingingoccursforallmotionsandappearstobeinherenttotheFSM.Whatisinterestingtonoteisthatformotionsbetween2and64pixels,theshapeoftheringingappearstobeconstant.Theonlydifferenceappearstobetheamplitudeoftheringing.Whileringingthatquicklydampsoutwouldnotnormallyhinderspeckle 41

PAGE 42

Bifano&Stewart ( 2005 ).Thebasicideaisthatiftheringingcanberepresentedbyadampedharmonicoscillator,itshouldbepossibletousedestructiveinterferencetoremovetheringing.TodetermineifthissolutionisappropriatefortheFSM,theringingmustobeythefollowingform: 3-5 andshowhowthesumofthetwocommandsdestructivelyinterferetoproduceadiscretemotion.Thedownsidetothissolution,however,isthatitimposesaminimumcycletimeequaltoone-halftheresonantfrequency.Todetermineifthereisindeedasinglewforallmotions,Iconductedmeasurementsoftheringing.ThiswasdonebyshiningalaserofftheFSMandsteppingthemirrorbetweentwolocationsseveraltimes.IrecordedthesemotionswiththeEMCCD,givinghightimesamplingofthedata.TheresultingmeasurementsandtsarepresentedinFigure 3-6 .IfoundthattherewasindeedonewthatcouldbeusedtodescribetheringingoftheFSManditsvaluewas9.51ms.Todeterminetheoptimalamplitudefortherststep,Iranasmalloptimizationroutinewhere 42

PAGE 43

PlotofFSMringingbetweenshifts.Datafrom9differentcyclesisfoldedtoaccentuatetheoverallshapeoftheringing.Themedianofthe9cyclesisrepresentedbythesolidlines.Thedifferentlinesrepresentdifferentrequestedpixelshifts. 43

PAGE 44

Exampleofthetwo-stepsolution.TheredcurveistheinitialcommandgiventotheFSMandthebluecurveisthesecondcommand.Thetwodestructivelyinterferetoproducetheblackcurve. variousamplitudesweresimulatedandtheonethatproducedthesmallestresidualringingwasadopted.FortheSPIFS-POCFSM,theoptimalamplitudefortherststepis0.80.Withatheoreticalsolutioninplace,Iappliedittothedata.Thesolution'seffectonthedataisevidentandIpresenttheresultofapplyingthetwo-stepsolutiontoa16-pixelshiftinFigure 3-7 .Heretheringingisgreatlydiminishedanddampsoutquicklyresultinginmotionsthatrequire<6mstostabilizefromtheinitialcommand.Thissolutiontotheringingissuegreatlyspeduptheroutine,butitbroughttolightthefactthatthefastesttheFSMcouldoperatewasat200Hz(5mscycletime).ThisissignicantlyslowerthanadvertisedandwillbediscussedmoreinChapter4andFigure 3-7 alsoshowsthatthesolutionisnotperfect.Thereisstillabitofanovershootwhenthemirrorrstmovestoitsnallocation.Whilethisovershootislessthan1/2inthecaseofnoringingsolution,itstillaffectsthedataandwillbeaddressedinChapters5and6. 44

PAGE 45

MediandatafromFigure 3-4 withdampedharmonicoscillatortstothedata.WhatisinterestingaboutthetsisthatIndwisindeedindependentoftheamplitudeofmotion.Thisisimportantbecauseitmeansagenericsolutioncanbeappliedtosolveforallmotions. 45

PAGE 46

Ringingsolutionappliedtoa16-pixelshift.Notethatthereisstillaresidualovershootof2pixelsbutthattheringinghasbeendampedoutresultinginmotionsthatrequire<6mstostabilizefromtheinitialcommand. 1. ThefocalplanesofthedetectorsNyquistsamplethediffraction-limitini'band 2. Theinstrumentshouldacceptanf/15beamfromthetelescope 46

PAGE 47

ThedistancebetweentheFSMandtheEMCCDbemaximizedsothatminimalmotionsarerequiredtomovetheFSM 4. TheoreticalimagequalityfromtheopticshadtohaveStrehlratios>90%for0:70mm
PAGE 48

3-8 .In(a)IpresenttheopticalpathforlightheadedtotheEMCCDwhile(b)hasthepathforlightgoingintotheSBIG.Thevariousopticalcomponentsarelabeledandshowhowtheinstrumentshouldbelaidout.IgivetheperformanceoftheopticalsysteminFigure 3-9 .Alongtheleftcolumn,thespotdiagram(a),encircledenergy(c)andStrehlratios(e)oftheEMCCDopticsaregiven.Indthatdiffraction-limitedperformanceisexpectedfortheSPIFS-POCsamplingandthatStrehlratiosgreaterthan90%canbeexpectedatallrelevantwavelengths.OntherightcolumnofFigure 3-9 Ishowthespotdiagram(b),encircledenergy(d)andStrehlratios(f)oftheSBIGoptics.Heretheperformanceisnotasgood.Whilethespotdiagramistight,theencircledenergydoesnotmatchthediffractionlimitandStrehlratiosdropbelow90%atwavelengthslongerthan0:8mm.Whilethisperformanceissub-optimalforhighprecisionphotometricwork,forproof-of-conceptdemonstrationswheretheprimaryconcernishigh 48

PAGE 49

3-10 .Thesecustomopticsholdersweredesignedsothatthelensescouldbeboltedtotheinstrumentandreplacedifneeded.Asinglealuminumplatewithcustomdrilledholesservedastheopticalbenchuponwhichallthecomponentsrest.Nearlyalltheboltswereofthesamesizetoreducethecomplexityofon-siteassemblyandshipping.Two1gussetsrunalonghalftheinstrumentandprovidearigidstructuralsupport.Additionalsupportisgivenbystandardopticalmountingpoststhatareboltedintothebase-plateandthetopplate.Furthermore,Idesignedthehousingssothatallofthenecessaryelectronicsandcomputerscouldbeboltedtotheinstrumentaswell.ThenetresultisthatonlytwoethernetcablesandonepowercablearerequiredtooperatetheSPIFS-POC.IpresentdesigndrawingsofthenalinstrumentinFigure 3-11 .Theinstrumentisrelativelycompactgiventhelongfocallengthsofthemultipleopticsandhasnaldimensionsof48x 49

PAGE 50

ZEMAXdrawingsoftheopticaldesignfortheSPIFS-POC.In(a)theopticaldesignforlightpassingtotheEMCCDispresentedwhereas(b)showstheopticalpathoflightpassingtotheSBIG.(0)isthetelescopefocalplane(1)isthecollimator(2)isthelocationoftheFSM(3)iswherethebeamsplitteris(4)isthecamerafortheEMCCD(5)and(6)arefoldmirrors(7)isthedetectorplaneoftheEMCCD(8)isthecamerafortheSBIGand(9)isthefocalplaneoftheSBIG. 50

PAGE 51

B D F ThetheoreticalperformanceoftheopticalsystemasdeterminedbyZEMAX.Alongthetoprow,thespotdiagram(a),encircledenergy(b)andStrehlratios(c)oftheEMCCDopticsaregiven.Onthebottomrow(d),encircledenergy(e)andStrehlratios(f)oftheSBIGopticsaregiven. 51

PAGE 52

SolidWorksdrawingoftheholderfortheEMCCDcameraoptic.Thedrawingdenotesthefourcomponentsthatmakeuptheholderasawhole. 42x27.Tomovetheinstrumentaround,IalsodesignedahandlingcartthatwasmadeoutofUnistrutandIshowthedesignforthecartinFigure 3-12 .Alltoldtherewere72differentcustompartsfortheSPIFS-POC,butgiventhesimpledesignandsmallnumberofbolttypes,thesystemiseasytoassembleon-site.IaddedtwoadditionalcomponentsforthesubsequentobservingrunattheWilliamHerschelTelescope.Therstwasamotorizedlterwheel(59-769)fromEdmundOptics.Itiscapableofstoringve50mmltersinitsmagazine.Thelterwheelcaneitherbemanuallymovedorbecommandedbycomputerwithincludeddrivers.ThesecondmodicationforobservationsattheWHTwastheinterfaceforthetelescope.ObservationswiththeSPIFS-POCin2010AattheWHTwereperformedintheGround-BasedHigh-ResolutionImagingLaboratory(GHRIL)lab.Thislaboratoryactuallysimpliedthedesignrequirementsfortheobservingrun.TheGHRILisalablocatedontheNaysmithplatformofthetelescopeandhasanopticalbenchsituatedafteraderotator.Therefore,tomakeSPIFS-POCcompatiblewiththetelescopeonlyrequireddesigningafewblocksforthemainplatetositatoptheopticalbenchsuchthattheopticalaxeswerealigned. 52

PAGE 53

TheentireSPIFS-POC.(a)istheinstrumentasitwouldappeartotheuserand(b)isaninsideview.Thecircularplatein(a)boltsontothemountingplateatCassegrainfocusattheKPNO2.1-m.In(b),allofthemaincomponentsarelabeledtoshowtheirlocation.NotethesimilaritytotheopticaldesigngiveninFigure 3-8 53

PAGE 54

SolidWorksdrawingsoftheSPIFS-POChandingcard.In(a)Ishowthecartdesignandin(b)Ishowhowtheinstrumenttsintothecartsothatitcanbemountedonthetelescope. 54

PAGE 55

Articialspecklepatternproducedbythehairsprayphasescreen.Notethattherearemultiplespecklesofvaryingintensity.Whenthephasescreenisrotated,thespecklepatternchangesandmovesabouttheentiredetectorwithbothlowandhighorderchanges. AsecondoptionthatIdecidedtopursuewasthatofusingasheetofglassandhairspray. Thomas ( 2005 )noticedthatproperlydepositedhairsprayonglasshassimilarpropertiestoaKolmogorovturbulencespectrum.Specically,theyfoundthatputtingdownoneevencoatthen 55

PAGE 56

Thomas ( 2005 )speculatedthattheamphomercompoundinthehairsprayiswhatmakestheturbulenceappearKolmogorov.Giventhatthisisapotentiallyeasysolutiontoadifcultproblem,Idecidedtotestformyself.TheGregoryMarshalBlondSaloninGainesville,FloridadonatedhairsprayforthepurposeofscienticinquiryandaPlexiglasswheelthatpreviouslyservedasanentrancewindowwasdonatedbytheFLAMINGOS-2team.Iexperimentedwiththehairsprayandfoundthatitdoesindeedproduceadecentturbulencepattern.IpresentanexampleofthespecklesproducedbyshiningalaserthroughthephasescreeninFigure 3-13 .TimesampledturbulencewasproducedwhenIhooked-upthephasescreenuptoasteppermotorandgraduallyrotatedit.Adjustingtherotationspeedenablemetotunethecoherencelengthofthespecklepatternstomatchatmosphericpredictions. Rampyetal. ( 2010 )recentlyconrmedtheKolmogorovnatureoftheturbulenceproducedbythesemethods.Theyalsofoundthatwhilehairsprayworkswell,thereareadditionalpracticaladvantagesthatcanbegainedbyusingclearspraypaint.Notablythephasescreenisnolongerwatersolubleandthatbettercontrolinproducingthephasescreencanbeexercised. 56

PAGE 57

Law ( 2007 ).In Law ( 2007 ),theiralgorithmanalyzedspecklepatternstakenfromanEMCCDtondthelocationbestspeckleandshiftedandaddedontooneanotherviaadrizzleroutine.InmanywaysthisisexactlywhattheSPIFS-POCrequiresexceptinsteadofshiftingandaddingtheimages,shiftinformationissenttotheFSMforrealtimestabilization.Theoveralllooparchitectureisasfollows: 1. AcquireaspecklepatternwiththeEMCCD 2. Identifythelocationofthebestspeckle 3. Convertthelocationfrompixelunitstovoltages 4. OutputvoltagestotheFSMandstabilizethespecklepattern 5. Takeanewexposure 6. RepeattheloopThesystemoperatesinaclosedloopfashionwhereeachmotionisdependentonthepreviousonesuchthatsolutionsareiteratedonwitheachstep.Aclosedloopapproachis 57

PAGE 58

Kernetal. 2000 ))andisontheorderof30milliseconds.Sincethereshouldbetwomotionsperspecklelocation,theloopneedstooperateat>70Hz.Tooptimizeroutinespeedsandensurethatthecomputationisminimallyimpactingthelooptime,allofthecodewaswritteninC++.Sinceatthetimeofpurchase,driversforallofthecomponentswereonlyavailableinWindowsVista,thatistheoperatingsystemofchoice.ThecomputerrunningtheloopisaDellT7400WorkstationwithdualIntelQuadCoreprocessorsat2.33GHz.Thereare8GbofRAMandthree15,000RPMSAS150GbharddrivesinaRAID0congurationi.e.dataaresplitamongstthedifferentharddrivestominimizewritetimes.TheactualC++codewaswritteninMicrosoftVisualC++2008Edition. Law ( 2007 )fortheluckyimaginingsystemLuckyCam.Withthismethod,Iconvolveanassumed,idealPSF(generatedeitherthroughsimulationorpreviousdata)withthespecklepattern.TheinverseFouriertransformofthisconvolutionresultsina2Dcross-correlationthatprovidesthelocationofthebestmatchbetweenspecklesandidealPSF. Law ( 2007 )usedthismethodbecauseitanalyzesthespecklepatternattheresolutionofthediffractionlimitand,intheory,shouldndthebestoverallmatchinqualitynotjustbrightness.TheBrightestPixelmethodisasstraight-forwardasitsounds.Withthistechnique,Ideterminethebrightestpixelinthespeckleimageandassumethatlocationisalsothelocationofthebrightestspeckle. 58

PAGE 59

Law ( 2007 )hasdone,drizzlingroutinesareslowandhavelittleeffectontheanalysisdonehere.Tocomparethemethods,IuseddatacollectedattheKittPeak2.1metertelescopeinJuneof2009(describedinChapter5.1).Duringthisobservingrun,Iobtainedthousandsofspeckleimagesofstarsatvaryingmagnitudes.Thecurrentanalysisislimitedtoi'observationswheretherewasarangeof7magnitudesinbrightness.Thelistofstarsusedforthisanalysis,includingtheirV-bandmagnitudes,exposurelengths,andnumberofexposuresaregiveninTable 4-1 .TherangeinexposurelengthsandtargetmagnitudescoverstheparameterspaceofspeckleexposurestheSPIFS-POCwillemployandprovidesandexcellentopportunitytotestoptimalalgorithms.TheprimarydiscriminantsIusetodeterminewhichalgorithmissuperioraretheFWHMandStrehlratiosoftheofthestackedPSFcores.Otherfactorstoconsiderforrealoperationarethelengthoftimethealgorithmtakestondthebestspeckleandtherequisitenumberofphotonstogetenoughinformationforspeckleselection. Table4-1. Listofstars,vmagnitudesandexposureparametersusedforspeckletesting. Star ExpLength(ms) 7.0 1ms 100,000 hd122574 7.0 5ms 90,000 hd173416 6.057 10ms 70,000 18Del 5.522 10ms 40,000 14and 5.22 10ms 40,000 hr4905 1.77 1ms 200,000 hr4905 1.77 5ms 100,000 Icalculatedthelocationsofthebestspecklesforeveryimageusingboththe2DCCandBPmethods.For2DCCanarticial,diffraction-limitedi'imagewasproducedusingthesimulationcodeoutlinedinChapter2.1.Iusedthebestspecklelocationstoshift-and-adddatainIDLusingaprogramIwroteforthetask.InsimulatingSSIassumedtherewasnoinstrumentaldelay,i.e. 59

PAGE 60

4-2 .Fromtheseresults,therearesomeinterestingndings.TherstndingisthatPSFsproducedby2DCCtendtohavehigherStrehlratiosthanBPindicatingbetterimagequality.HowevertheBPPSFs,onaverage,tendtohavetightercores(smallerFWHM)than2DCC.WhatisinterestingaboutthisresultisthattheFWHMforBPhavesignicantscatter,varyingbetween1.3pixelsto4.4pixelswithameanof2.9080.997pixels.For2DCC,ontheotherhand,themeanis3.1130.529pixels.Sowhilethemeanvaluesaresimilar,thescatterinthe2DCCmethodismuchless.Infact,dataforthefaintestdataset(HD122574at1ms)areparticularlytroubling.TheFWHMfortheBPmethodareafactorof2smallerthanthediffractionlimit(whichis2.1pixels).Sincethedatawereallreducedinthesameway,thisindicatessomethingunusualisgoingonwiththeBPmethodinthelimitingcaseoffaintguiding. Table4-2. ComparisonbetweenBPand2DCCinbothFWHMandStrehlratiosforthesamedatasets.FWHMareinpixelsandxandydenotethedirectionoft. BrightestPixel 2DCross-Correlation Star FWHMx FWHMx HD122574(1ms) 1.310.06 1.530.07 0.01930.0008 2.310.11 2.460.08 0.02190.0010 HD122574(5ms) 2.340.18 2.530.14 0.01650.0015 3.100.12 3.110.11 0.01750.0016 HD173416 3.230.16 2.960.03 0.01410.0002 3.580.19 3.280.05 0.01460.0002 18Del 3.530.11 3.290.23 0.01480.0010 3.800.11 3.580.21 0.01510.0010 14and 4.410.19 3.110.10 0.01370.0002 4.660.17 3.490.11 0.01400.0002 HR4905(1ms) 2.370.06 2.930.09 0.02070.0010 2.770.05 3.310.08 0.02160.0010 HR4905(5ms) 3.160.04 3.560.05 0.01450.0002 3.450.06 3.880.05 0.01460.0002 4-1 IpresentadifferenceimageoftheBPresultandthe2DCCresultforthesamesetofdata.Whiletherearesignicantdifferences,oneofthemoststrikingisthatthecenter-mostpixelisnegativewhiletherestofthediffraction-limitedcoreispositive.Thisindicatesthatonlytheverycenterpixel 60

PAGE 61

Differenceimageof2DCCandBP.Thebrighterregionsdenoteandexcessinthe2DCCPSF.Notetheonedarkpixelatthecenter.ThisshowsthepeakednatureoftheBPmethodandindicatesthatitispronetoselectingspuriouspixelartifactsproducedbyclockinducedcharge. isbrighterintheBPmethodwhiletheentirecorefromthe2DCCisoverallmoreluminous.Therefore,insomeinstances,theBPmethodislikelytargetingspuriousdetectionseventswhereonlyonepixelisactivatedfromstrayphotonsornoise.Withenoughofthesesingle-pixeleventsbeingusedforimagestacking,itcausesatighterbutnon-physicalFWHMaswellasacorrespondinglylowerStrehlratio.Iproducedcontourplotsofthedegreeofcorrelationbetweenspecklelocationsdeterminedbyeachofthetwomethodsforallofthespeckleimagestoanalyzethediscrepancybetweenthesetsfurther.TheresultisgiveninFigure 4-2 .Ifthetwotechniquesfoundthebestspeckletobeinthesamelocation,therewouldonlybeastraightlinewithaslopeofone.Whileitisclearthatmuchofthetimethetwotechniquesareingeneralagreement,thereissignicantscatterindicatingthatBPand2DCCarendingdifferentspecklesforstabilization. 61

PAGE 62

Contourplotsofthecorrelationbetweenspecklelocationsselectedwiththetwomethods.Thegureontheleftisforspecklelocationsinthexdirectionwhilethegureontherightisfortheydirection.Iftherewerecompleteagreementbetweenthetwomethods,theresultwouldbeastraightlineofslopeequaltoone. Todeterminehowoftentheresultsdiverge,IplotthefractionofimageswithaparticularpixeloffsetbetweenthetwoselectionmethodsinFigure 4-3 .Sincethediffraction-limitedFWMHfortheseobservationswas2.1pixels,anythingbeyond4pixelsdifferentisover2l=Dawayi.e.notthesamespeckle.Throughthisanalysis,Indthatthetwomethodsonlyagreeonexactlythesamelocationlessthanhalfofthetimewithasignicantspreadforthedifferentstars.Inthebestcasescenario,thetwomethodsagreetowithin2l=D80%ofthetime,butintheworstcasethatnumberfallsto<50%.AlsofromtheFigure 4-3 ,Indthatthereisacleartrendbetweenguidestarbrightnessandagreementbetweenmethods.Inthefaintestcases(HD122574at5msandHD122574at1ms)thereismuchlessagreementthaninthebrightestcases(HR4905at5msandHR4905at1ms).Assuch,thisconrmstheearlieranalysiswhereBPwasfoundtobeheavilyinuencedbysinglepixelevents.GiventhatBPseemstobeaffectedbysinglepixelevents,itisimportanttotrytodeterminethecauseoftheseartifacts.Asdetailedinx1ofChapter5,cosmicraysareunlikelytobetheprimarydriver.Thisisbecauseofthelownumberofcosmicrayeventsandthefactthatcosmic 62

PAGE 63

Cumulativedistributionofimageswherethedifferencebetweenspecklelocationsdeterminedbythetwomethodsarelessthanorequaltoagivenpixeloffset.Notethatgreaterthan4pixeldifferentisover2l=Daway.Thetotalnumberofimagesusedfortheanalysisis640,000. raystrikescoverseveraladjacentpixels.ThenextpossibilitythatIexaminedisifthisisapropertyresultingfromnoise.EMCCDsproduceanextrap 63

PAGE 64

1. Therearesignicantdifferencesbetweenthespecklesselectedbythetwomethods 2. 2DCCissensitivetotheentirePSF,notjustthebrightestpixel 3. BPproducesoverlynarrow,unrealisticFWHMduetoclockinducedcharge 4. 2DCCproduceshigherStrehlratiosatallmagnitudesoverBP 64

PAGE 65

Kernetal. 2000 )andbecauseofthisfact,theSPIFS-POCneedstobeabletooperateasfastaspossible.SinceIhaveshownthatcorrectingtheFSMforitsringingalreadyproducesnearly5msoflag(Chapter2.1.2)itisimportantthateveryothercomponentofthesystembeoptimizedforspeed. Frigo&Johnson 2005 )routinewasusedfortheactualFouriertransformcalculations.FFTWisapubliclyavailablesubroutinewritteninC.TheperformanceofFFTWissuperiortootherfreeFouriertransformcodeandisevencompetitivewithplatform-speciccodes.Anadditionaltrickwastocutoutquadrantswapping.Thenaloutputofthecross-correlationisanarraywherethequadrantshavebeenswapped.Reconstructingthearrayistimeconsumingandinefcient.Ratherthanperformthisreconstruction,Igeneratedalook-uptablesuchthatwhenthelocationofthebest-speckleisdeterminedintherawoutputIplugthatvalueintothetableandndwhattheactuallocationofthebestspeckleisindetectorspace. 65

PAGE 66

Ford ( 2009 )investigatedtheutilityofCUDA-enabledGPUsandfoundthattherewasuptoafactorof200xincreaseinspeedforalgorithmssolvingplanetaryorbits.ThiskindofpowerforKepleriansimulationsisvitaltothestudyofexoplanetdynamicsandgreatlyreducessolvingtimes.TheseimpressiveresultsmightbeapplicabletotheSPIFS-POCalgorithms.Thecurrentcross-correlationtimeof1.33msisfastbutsincethealgorithmisdependentonaseriesof2-dimensionalFFTs,itisplausiblethataGPUwouldbeabletoreducethecalculationtime. 66

PAGE 67

Frigo&Johnson 2005 ).IexecutedtwoteststodeterminetheutilityoftheGPUfor2DFFTprocessing.First,IgeneratedasingleFFTplan,allocatedmemoryonceandputaforlooparoundtheactualexecutioncodetogetasenseoftheabsolutespeedoftheFFTs.IchosetouseNpixelsonasiderangingfrom4to2048,whereNisalwaysafactoroftwoasthisistheoptimalcaseforbothroutines.IshowtheresultinFigure 4-4 .TheCPUoutperformstheGPUforoperationssmallerthan104elements,butbeyondthat,theGPUisalwaysfasterbyafactorofnearly20x.TheGPUappearstoat-lineat0.15ms/FFTforthesmalltransforms.ItislikelythatthisoccursbecausenotallcoresareutilizedinthesmallFFTcasesandthealgorithmscannotrunanyfaster.TheCPU,ontheotherhand,risesatasteadyrate.Thesecondsetoftestsinvolvesdatatransfer.WhiletheGPUcanperformveryfastexecutions,itrequiresthatthedatabetransferredontothecard'sinternalmemory.ThisobviouslytakestimeandIoptedtoinvestigatejusthowmuchofanedgetheGPUlosesinaccountingforthislimitation.Inthesetests,IcreatedarraysinaseparateC++program,thenpassedtoCUDAforexecution.Forcomparison,IranFFTWinasimilarmanner.TheresultsofthesetestsaregiveninFigure 4-5 .Whenaccountingfordatatransfer,itisclearthattheGPUunder-performstheCPUwhenthearraysaresmallerthan105elements.Greaterthanthisthreshold,theGPUattains10xincreaseinperformancespeed. 67

PAGE 68

ComparisonofGPUandCPUperformanceforcomputingrawFFTs.Memoryisonlyallocatedbeforetheexecutionsbegan.Thedottedbluelinedenotesthesizeofa128x128array. ThereforethesmallnatureoftheEMCCDimagewouldnotbenetfromthehighlyparallelnatureoftheGPUarchitectureandSPIFS-POCwouldactuallyrunslowerthroughtheGPU.Furthermore,itappearsthatingeneral,GPUsdonotcurrentlyprovideasignicantbenetforFFTsonanysizelessthan104elements.Theyprovidethebestincreasesinspeedforlargearrays.Thesendingsarelargelyconsistentwith Merz ( 2007 )whoperformedsimilar,albeitmoredetailed,investigationsofthisquestionwithanNVIDIAQuadroFX4600.TheFX4600inanindustrial-gradeGPUwith112processorsresultinginabouthalfthecomputingpoweroftheGTX-280.Thisisreectedbythefactthat Merz ( 2007 )comestothesameconclusionsasthisinvestigation,butfoundthatthresholdtobeaboutafactoroftwotimesmorepixelsfortheFX4600.GPUsareextremelypowerfultools,andinthecasesoflargeimageconvolutions,theyprovideatremendousreductionincomputationtime.However,onsmallerscalestheyprovide 68

PAGE 69

ComparisonofGPUandCPUperformanceforcomputingfullFFTs.MemoryisallocatedbeforeeachexecutionanddataistransferredfromahostprogramtotheFFTprogram.Thedottedbluelinedenotesthesizeofa128x128array. littleutility.Itislikelythatthisnewtechnologywilldevelopfurtheranditisworthfollowingtheprogressforuseinastronomicalinstrumentation. 69

PAGE 70

70

PAGE 71

1. Window=64x64pixels2,N=10 Window=64x64pixels2,N=100 Window=32x32pixels2,N=10 Window=32x32pixels2,N=100 Window=16x16pixels2,N=10 Window=16x16pixels2,N=100Totestthewindowing,IuseddatacollectedduringtheKPNO2.1-mobservingrundescribedinChapter5.1.Thehigh-timeresolutionspeckledatatakenfromtheEMCCDwereusedtosimulatetheperformanceoftheSPIFS-POCalgorithm.Isimulatedthiseffectbyassumingaparticularimage(dependingontheNparameter)istherstimageinaparticularsequence.IthenplacedawindowonthelocationofthebestspeckleandfurtherbestspecklesinNsubsequentimageshadtobelocatedinthisregion.Theimageswerethenshiftedandaddedbasedontheconnedinformation.RatherthanattempttocalculateStrehlratios,Iemployedasimpleuxratiowhichwasreferencedtoaseeinglimitedimagecomposedbystackingallthespeckleimageswithnoshifting.Iplacedadiffraction-limitedaperturenthepeakoftheseeinglimitedimageandtheuxwithinthataperturewasdenedasthebaseline.Ialsoappliedthesameaperturetoimagesproducedbystackingspecklepatternsthroughshiftandadd.Idenetheratiooftheuxesfluxspeckles=fluxseeingastheuxratioanduseitasaproxyforStrehlratios.Intheoptimalcase,IwantthewindowandNvaluestohavehighuxratiosaswellaspreservingtheresolutiongains.IgivetheresultsofthesetestinTable 4-3 .Myanalysisrevealsthatwindowinghasalmostnoeffectonresolutionwhatsoever.Thismeanstheonlyrealparameterofconcernistheuxratio.InalloftheN=100cases,theuxisreducedby>10%,whichistoolargeahitforanalreadylowStrehlratiosystem.N>10suffersfromtoohigha 71

PAGE 72

Table4-3. Allsimulationsassumea2mslagbetweenactualspecklepatternandcommandssenttotheFSManearlyidealSSsystem. Window Iterations Star FWHM FWHM ResGain ResGain Fluxratio (x) (y) (x) (y) 64x64 10 hr4905 3.44 3.96 14.2 11.4 3.98 64x64 10 hd122574 3.08 3.13 15.2 15.2 4.07 64x64 10 bd2512 2.73 2.99 14.2 15.9 4.36 64x64 100 hr4905 3.44 3.94 14.2 11.5 3.93 64x64 100 hd122574 3.07 3.17 15.3 15.2 4.07 64x64 100 bd2512 2.72 2.98 14.2 15.9 4.32 32x32 10 hr4905 3.40 3.91 14.3 11.6 3.71 32x32 10 hd122574 3.04 3.09 15.4 15.4 3.85 32x32 10 bd2512 2.69 2.96 14.3 16.0 4.28 32x32 100 hr4905 3.37 3.89 14.5 11.6 3.38 32x32 100 hd122574 3.05 3.06 15.4 15.5 3.47 32x32 100 bd2512 2.70 2.97 14.3 15.9 3.72 16x16 10 hr4905 3.35 3.87 14.5 11.8 3.11 16x16 10 hd122574 2.97 3.09 15.8 15.6 3.16 16x16 10 bd2512 2.66 2.98 14.5 16.0 3.77 16x16 100 hr4905 3.36 3.85 14.5 11.8 2.58 16x16 100 hd122574 2.91 2.97 16.5 16.0 2.39 16x16 100 bd2512 2.67 3.03 14.4 15.7 2.83 72

PAGE 73

4-4 .Inotethatthisdoesnotincludeanytimeforspeckleexposure.Whenspeckleexposuretimesareincluded,theloopspeedis<125Hz.Whilethismeetstheminimumrequirementof>70Hz,itonlydoessobyafewms.Thisistroublingbecause30msisagenerousestimateof 73

PAGE 74

Thetimingsofthevariouscomponentsthatmakeuponecycletime. Component Time(ms) FSMringingcorrection 4.50 SpeckleSelection 1.33 Driverinitializations 1.70 Othersystemoverheads 0.50 thecoherencetimeofspecklepatterns.Ifthewindspeedincreasesorr0decreases,thecoherencetimecandropto10msorless.AnadditionalfeaturethatisincludedintheSPIFS-POCcontrolcodeisspooling.WhiletheEMCCDisacquiringspeckledata,itisabletospoolthedatatotheharddiskinrealtimewithnonoticeabledropinperformance.TheuserinterfacealsoallowstheusertoimplementEMCCDcoolingaswellasadjustmentoftheEMGain. 4-6 .AvarietyofspeedsfortheloopweretestedbutfortheimagesinFigure 4-6 theloopspeedwas63Hz(thisisslowerthantheoptimal125Hzbecauseofexposuretimesaswellasthefacttheversionofthecodeusedwasnotcompletelyoptimized).Intheleftpanel,aseeinglimitedhaloisproducedwithSPIFS-POCoff.Intherightimageasimilarhaloispresentbutwithasharp,diffraction-limitedspikeinthecenter.Totalintegrationtimeforbothimageswas100secondstoaverageoverturbulenceproducedbythephasescreen.AmorequantitativeapproachisgiveninFigure 4-7 wherea1Dslicethroughthebrightestpartoftheimageispresented.TheleftpanelcontainsaslicewhenSPIFS-POCisoff.AroughlyGaussian-shapedPSFispresentandhasameasuredFWHMof9.4pixels.Intherightpanel,withspecklestabilizationon,asimilarbroadhaloispresent,butasharp,diffraction-limitedspikeisontop.FittingaGaussiantothespikerevealsithasaFWHMof2.6pixelsnearlyafactorof 74

PAGE 75

B Imagesofin-labspecklestabilization.In(a)SPIFS-POCisoffandthespecklepatternissmearedoutoverthecourseof100secondsofintegrationproducingaseeinglimitedimage.(b)SPIFS-POConwiththesameexposuretime.Theresultisasharpspikeinthecentersurroundedbyadiffusehalosimilarinsizetotheleftimage. B CrosssectionsofthePSFsproducedinlaboratoryspecklestabilization.In(a)SPIFS-POCisoffandin(b)SPIFS-POCisonwiththesameexposuretime.Theresultisasharpspikedenotedinred. 75

PAGE 76

76

PAGE 77

5-1 ,IpresentaphotographofthenishedinstrumentasmountedontheKPNO2.1-mtelescopeandprovideanannotatedschematicoftheinstrumentintherightofFigure 5-1 wherethelocationsofthevariouscomponentsareshownWeconductedtherstsetof2.1-mobservationswiththeSPIFS-POCin2009A(June8th-June14th)andthesecondin2009B(August6th-August10th)foratotalof11.5nights.Theseobservationswereprimarilycharacterizationrunsdesignedtotesttheoverallfunctionalityofthesystemandassessthevalidityofthedesign.In2009AIwasaccompaniedbyDr.JosephCarsonandacquiredspeckledatafromstarsofvariousbrightnesses.Table 4-1 summarizesthetargetsobserved.TheinformationfromthesedatawereusedtotunethesystemforfuturespecklestabilizationandmuchoftheanalysisgiveninChapter4.2andChapter4.5isbasedontheseobservations.The2009Brunallowedmetotestloop-closure.Unfortunately,duetoinsufcientlyoptimizedcode,thefunctioningloopspeedwas50Hzforthisrun.Furthermore,thecombinationofthisslowloopspeedandpoorseeingconditionsattheKPNO2.1-m(averageseeingof1.5)resultedinlow-qualitycorrections. 77

PAGE 78

B Ontheleft:photooftheSPIFS-POCmountedontheKPNO2.1-mtelescope.Ontheright:photoofthekeycomponentsoftheSPIFS-POCwithlabels. Despitepoorobservingconditions,observationsofthebinarystarsystem65tauCygprovedencouraging.WeusedtheSPIFS-POCtoobservethissysteminaclosed-loopmodeat50Hz(thelowspeedwaslargelyduetodriverinitializationissuesdiscussedinChapter4.4).Whilethisloopspeedissignicantlyslowerthanthecoherencetimeoftheatmospherewewereabletoimprovetheimagequalityenoughoverseeingtodiscernthelocationofaknowncompanion.Ipresentanimageofthisbinarystarbothintheseeing-limitedandspecklestabilizedcaseinthez'lterinFigure 5-2 .Thesystemhasaknownseparationof0.77andapositionangleof236.55deg.Bothobservationswere100secondslongandwetooktheminseeingof1.6.Theleftimage(a)isaseeinglimitedimagewiththeSPIFSsystemturnedoff.Ontheright(b),wepresenttheimagewithspecklestabilizationenabled.Thetwostars(whichhaveathreemagnitudedifferenceinbrightness)arecleanlyresolvedataseparationandpositionanglematchingpreviouslypublishedresults( Masonetal. 2001 ).Toproducetheimagein(b),aseeing-limitedPSFwassubtractedfromtheprimarystartoemphasizethesecondaryintheresiduals.Thephysicalseparationandorientationmatchknownvaluestowithin2sandhintat 78

PAGE 79

B SPIFS-POCobservationsof65tauCyg.Theleftimage(a)isaseeinglimitedimagewiththeSPIFSsystemturnedoff.Ontheright(b),wepresenttheimagewithspecklestabilizationenabled. thehighangularresolutionpotentialofSS.Theseresultsprovedessentialtobeinggiventimein2010AfortheWHT.Besidespartiallysuccessfulloopclosure,oneofthekeytestsIwasabletoconductwastoproduceanestimateoftheeffectofcosmicraystrikesontheobservations.BetweenobservationsofBD2512andHD122574,Ihad300,000speckleimagestoanalyze.Inalltheseimages,thetypicalbrightestpixelinabest-specklewasbetween1,000and2,000counts.Therefore,saturationwouldbeindicativeofacosmicray.Isearchedthedataandfoundonlyonesaturationeventinthewholesample.Thiswouldindicatethatcosmicrayeventsarenotamajorissueforspecklestabilization.Thisisnotunreasonablegiventhattheexposuretimesaremuchlessthan1secondandthattheareaofthedetectoris0.09cm2. 79

PAGE 80

5-3 ,Ipresentcross-sectionsofstellarPSFsproducedwithspecklestabilizationonandoff.IacquiredthedatawiththeSBIGsciencecamerathrougha3nmltercenteron846nm.Itookthreeten-secondimagesandpresentthemedianaveragesoftheresults.ThespeckledataacquiredwiththeEMCCDforstabilizationlledabroaderwavelengthrangespanningSloanz'andloopspeedwas100Hz.ThetoprowdepictsobservationsofPSFStar1whilethebottomrowisofPSFStar2.Intheleft-mostimagesofFigure 5-3 ,cross-sectionsofseeing-limited(specklestabilizationoff)observationsarepresented.Ontopofthesecross-section,Gaussiantstothedataarealso 80

PAGE 81

ObservationsoftwoPSFstarswithbothSPIFSonandoff.ThetoprowpresentsobservationsofPSFStar1andthebottomrowpresentsobservationsofPSFStar2.ThedottedlinesinallimagescorrespondtotheGaussiants.TheFWHMvaluequotedineachplotisinunitsofpixels. 81

PAGE 82

Eikenberryetal. 2008 ; Keremedjievetal. 2008 ),thereisasharppeaksittingatopahalo.Thehaloisofsizecomparabletotheseeing-limit.ThemiddlepaneldemonstratesthefactthatasingleGaussiantisnotsufcienttocharacterizethePSFandresultsinanRMSnearlythreetimesgreaterthanthettotheseeing-limitedcase.ThelargestresidualstothetareseenintheexcessuxesinthecentralpartsofthePSF.Theseexcessesaresignaturesofspecklestabilization.TherightmostpanelsofFigure 5-3 presentatoftwoGaussianfunctionsatoponeanother.ThetwasmadebysimultaneouslysolvingforbothfunctionsusingmptfunfromtheIDLAstronomyUser'sLibrary1.ThistismuchbetterandyieldsadditionalinformationabouttheSSPSF.WhilethereisstillsomeexcessinthecenterofthePSF,thetproducesanRMSnearlyidenticaltothesingleGaussiantoftheseeing-limiteddata.Inparticular,theFWHMofthestabilizedspecklesis133masforStar1and165masforStar2.Thismeansspecklestabilizationproducedafactorof3.25improvementinresolutionoverseeingforStar1and3.16timesbetterresolutioninStar2.Infact,thecoreshaveFWHMonly3-4timesgreaterthanthediffraction-limitatthiswavelength.However,giventhatthereisexcessuxatthecenterofthestabilizedspecklePSFsinFigure 5-3 nottbyaGaussianshape,ImeasuredtheFWHMmakingnoassumptionsabouttheoverallshapeofthecore.Tomakethismeasurement,Irstsubtractoffthehalosothatonlythesharpcoreremains.Ithenfoundthehalfmaximumofthepeakintensityandperformedalinearttothedataaroundthatpoint.ThistgavemethepixellocationhalfmaximumandperformingthetfortheothersideofthePSFgivesmeadirectmeasureoftheFWHM.ForStar1,theFWHMis85masandforStar2itis99mas.Therefore,thedirectmeasurementoftheFWHMproduces 82

PAGE 83

5-3 donotaccountforthepeakintensity.Usingthesevalues,IndthattheSPIFS-POCwasableto5timesbetterresolutionthantheseeinglimitforbothstars.ItalsomeansthatStar1wasapproaching2l=D.AnotherinterestingnotefromFigure 5-3 isthefactthattheFWHMoftheunderlyingPSFintheSPIFS-oncaseislargerthantheseeinglimit.Inbothcases,thehalois1.3timeswiderthantheseeinglimit.Thiseffectisexpectedintheeventofspecklestabilizationandwaspredictedin Keremedjievetal. ( 2008 ). 5-4 ,IpresentanimageofthestarwithSSoffandSSon.ThedatapresentedwereacquiredwiththeSBIGcamerainthenarrow-band846nmlter.Exposureswere5secondslongandtheimagesusedforanalysiswerethemedianaveragesofthreeexposures.Thesystemhasbeendisplayedwiththesameintensityscalesothatdifferencesaremorereadilyapparent.In(a)ofFigure 5-4 ,Ishowthatintheseeinglimit,thetwostarsareclearlyunresolved.Whilethereiselongationinthedirectionofthesemi-majoraxis,thereisnowaytodistinguishbetweenthetwocomponents.Ontherightside,specklestabilizationclearlyresolvesthesystem.Analysisoftheimagerevealsthatthemeasuredseparationis0.577andapositionangleof294.62degrees.Thereforethereisagoodagreementbetweentheobservedsystemparametersandthetheoreticalexpectations. 83

PAGE 84

ObservationsofthebinarystarWDS14411+1344.(a)comprisedoftwoimages.Theleftimageistheseeing-limitedobservationwhiletherightimageistheSPIFS-onobservation.Thescaleofthegureisthesameforbothimages.ToheightentheresolvingpowerofSPIFS,wepresentacontourplotofthesameimagein(b).Fromthesegures,wecanseethatSPIFSwasclearlyabletoresolvetotwocomponentsofthebinarysystemwhileintheseeing-limit,theyareblurred.Themeasuredseparationwas0.577. 84

PAGE 85

5-5 .Ontheleft,Ishowahistogramoftheoffsetbetweenthenominalspecklestabilizationandtheactuallocationofthebestspeckle.FortheSPIFS-oncase(thesolidline)thereisaclearspikeat1pixeloffsetthatdominatesthehistogram.TherightpartofFigure 5-5 isthecumulativedistributionoftheoffsetsandfromthesedataweclearlyseethattheoffsetis<2pixelsmorethanhalfthetime.Ialsoplottheseeing-limitedcase(dottedline)inFigure 5-5 .Wetooktheseeing-limiteddatafromthesamestar,justbeforetheweacquiredtheSPIFSondata.Assuch,weexpectatmosphericconditionstobenearlyidentical.Withtheseeing-limit,thedataaremarkedlydifferent.Thereisawiderdistributioninoffsetsandthepeakoccursaround5pixels.Lookingatthecumulativedistributioninthelowerplot,Ishowthatintheseeing-limittheoffsetis<7pixelsonlyhalfthetime.ThedifferencebetweenthesetwodatasetshighlightsthefactthatourspecklestabilizationsystemdoesbringthebestspecklestothecenteroftheFOVmuchmoreoftenthanintheseeinglimit.ThismeanstheSPIFS-POCwasfunctioninginasuccessfulclosed-loopmode.However,thehistogramsinFigure 5-5 alsoshowspecklestabilizationwasnotfunctioningatoptimallevels.Havingoffsetsof1pixelormoreonlyservestodegradetheimagequalityandislikely 85

PAGE 86

AnalysisoftheaccuracyoftheSPIFS-POC.Ontheleftisahistogramofoffsetvaluesbetweenthebest-speckleandtheoptimalSSlocation.ThesolidlineisforSPIFSonandthedottedlineisforSPIFSoff.Ontherightweshowthecumulativedistributionofthedata. oneofthekeyreasonswhythe2010Aobservationsonlyachieve3l=Dperformance(althoughthiswillbediscussedinmoredetailinChapter6).Asaninterestingcheck,Inotethatthehalf-pointforthecumulativedistributioninbothcases(2pixelsforSPIFS-onand7pixelsofSPIFSoff)roughlymatchesthehalf-widthhalfmaximumsmeasuredinFigure 5-3 (forSPIFSon,theHWHMwas3pixelsandfortheseeinglimititwas10pixels).Thereforethehalfpointvaluesappeartocorrelatewiththesharpnessofthecoreinthenalimage. 86

PAGE 87

Kernetal. ( 2000 ).Asaresult,theidealloopshouldworkatleasttwiceasfast,meaningasystemcapableoffunctioningat>200Hz.Asawaytodemonstratethispoint,IshowtheeffectoflatencyinthePSFforSSin(a)ofFigure 6-1 .Here,usingon-skyspeckledatatakenwiththeEMCCD,Ihavesimulatedlaginthesystem.Theeffectoflagwasreproducedbyndingthebestspeckleeverynimagesandusethatlocationtoshiftandaddallsubsequentspeckleimagesuntilnoccursagain.Bychangingn,Icansimulatelagsfrom2msuptoseveralminutes. 87

PAGE 88

Ontheleft,plotofPSFcross-sectionsasafunctionoflatency.WEnotethatthepeakuxdecreaseswhiletheFWHMincreasesasafunctionoflag.Tohighlightthis,wepresentattoFWMHasafunctionoflagontheright.Thetrendappearsroughlylinearwithaslopeof1.422mas/ms. InFigure 6-1 Ipresentarangebetweennolagand20msofdelay(50Hz).ItisclearthattheStrehlratiochangesdramaticallyasafunctionoflag,butsodoestheFWHM.In(b)ofFigure 6-1 IshowthedependenceofFWHMonlatency.ThereappearstobealineartrendbetweenthetwoandareducedsquarestyieldsFWMH(mas)=1:4lag+55(herelagismeasuredinmilliseconds).Thisimpliesthatincreasingthespeedbyafactorof2overthecurrentoperationallooptimeof10msforSPIFS-POCincreasesthesharpnessby10%.Howeverslowingthesystemdownbyafactoroftworesultsina20%degradationinFWHM. 88

PAGE 89

PSFcross-sectionofourspecklestabilizationsimulationusingdatafromtheEMCCD.Weusedalagof10msinadditiontoarandomoffsettoFSMpositions.ThedatamatchourresultsfromFigure 5-3 fairlycloselyanddemonstratethecurrentlimitationsoftheSPIFS-POC. differencesbetweenthetwomethods,Ifoundoneparameterthatshiftingandaddingspeckledatainsoftwaredoesnottakeintoaccount:uncertaintyintheFSMsteering.IftheFSMdoesnotmovetoexactlythedesiredlocation,theerrorwouldmanifestitselfintheformofrandomoffsetsintheSAAlocationandthuswidentheFWHMofthePSF.AsrevealedinFigure 5-5 (andnotedintheringingsolutionofChapter3.1.2),whenspecklestabilizationisturnedonthereisstilla2pixelinaccuracyinpointinguptohalfthetime.IsimulatedthisinaccuracybyusingaGaussianweightedrandomnessintheFSMsteeringwiths=2pixels.ToforcethespeckledatatomatchtheSSdataascloselyaspossible,Ialsofactoredinalatencyof10ms.TheresultingPSFispresentedinFigure 6-2 anditappearsthatthesimulationreproducesthequalityofobservationsacquiredwithactivespecklestabilization.Therefore,mytestsrevealthatspeedingupthesystemandincreasingtheaccuracyoftheFSMcangreatlyimprovesystemperformance.Whatisinterestingisthatthesetwofactorsare 89

PAGE 90

6-3 andseethereisadramaticdifferenceintheStrehlratioasafunctionofimagequality.Infact,thepeakvalueofthetop1%ofimagesissomefourtimesgreaterthanwith100%ofimagesused.This,ofcourse,isnotaltogethersurprisinggiventhatluckyimagingteamshavebeenexploitingthisfactasthebasisoftheirinstruments.However,itdoesmeanthatahigh-speedshuttercouldgreatlyimprovetheimagequalityinaSSsystem. 90

PAGE 91

PSFcross-sectionasafunctionoffractionofimagesused.ThePSFshaveallbenormalizedsuchthattheyallsimulatethesametotalnumberofimagesused.Itisclearthatusingonly1%oftheimagesresultsinthehighest-Strehlimageswhileusing100%(asintheSPIFS-POC)resultsinlowerStrehlratios.Wenote,however,thatFWHMisnotstronglydependentonthefractionofimagesused,so100%ofimagesusedstillhastheresolutionalgainsof1%used. anditdispersesthelight.Undernormal,seeing-limitedconditionsatlowairmassatmosphericdispersionisnotusuallyalargeeffect.However,athighangularresolutionsitmustbetakenintoaccount.TodemonstratetheseverityoftheissueIhavecalculatedthedifferentialrefractionasafunctionofwavelengthforthreedifferentairmasses.Thecalculationsarebasedonequationsin Filippenko ( 1982 )andIpluginvaluesforLaPalmaspecically.Temperatureandpressurevaluesweretakenfrom Lombardietal. ( 2006 )and Lombardietal. ( 2007 )respectively.TheresultingdifferentialrefractionispresentedinFigure 6-4 .ThisFigureclearlydemonstratestheneedforanADC.Evenatlowairmass,theoffsetbetween0:5mmand1:0mmcanbenearly0.5.Sincethepixelscalewillbe<<0.5,thisissueneedstobeaddressed. 91

PAGE 92

PlotofdifferentialrefractionasafunctionofwavelengthforthreedifferentairmassesatLaPalma.Notethatevenatlowairmass,theoffsetcanbeasgreatas0.5. 92

PAGE 93

93

PAGE 94

7.1.1SpeckleStabilizationSimulationsTocompareSpeckleStabilizationtoSAAandLuckyImaging,weneedtoensureourestimationsofthevariousparametersareaccurate.Wechosetoconductourcomparisonwithasimulated2.5-metertelescope.Thisparticularsizewaschosenbecauseatlargerapertures,whileLuckyImagingandSSbothproducethesamespatialresolutionsinthecore,theStrehlratiosandusablefractionofspeckleimagesforLuckyImagingdecreasesdramatically.Asaresult,a 94

PAGE 95

Kolmogorov ( 1941 ).TheadaptiveopticsgroupattheJetPropulsionLaboratoryhasusedasimilaralgorithmextensivelyandhasverieditsaccuracyincomparisonwithactualperformanceresultswiththePalomarAdaptiveOpticssystem(PALAO).ThecodeusedtoproducetheSSPSFisthesamecodedetailedinChapter2.Fivehundreddistinctspecklepatterns(frames)wereproducedtoformanSDSSi'=0magstarasproducedbya2.5-metertelescope.Eachspecklepatternwassampledatvedifferentwavelengths(0.70mm,0.75mm,0.80mm,0.85mm,and0.90mm)toaccountforthebroadbandnatureoftheimaging.TosimulateSpeckleStabilization,wefoundthebestspeckleineachframeusinga2Dcross-correlationbetweenthespecklepatternandanidealPSF.ThisidealPSFwasproducedusingthesamecodebutwithnoturbulenceapplied.Wethenshiftedtheimagesaccordingtothelocationofthebestspeckleandstackedthem.Tocontrasttoseeing-limitedobservations,wesimplyaddedtheframesontopofoneanotherwithnoshiftingwhatsoever.WepresenttheresultsofthesimulationsinFigure 7-1 .TherewedemonstratethatthestabilizedimagecorehasaFWHMsimilartothediffraction-limitedcase.AnalysisrevealsthattheSSPSFisonly6%widerthanthediffraction-limitedcase.StrehlwasmeasuredbycomparingtoanidealPSFalsoproducedbythecode.WeusedtoFWHMoftheidealPSFtodeneanaperturewherepixelsintensitywouldbemeasured.TheStrehlwasmeasuredastheintensitywithintheaperturemeasuredfortheSSimagedividedbytheintensitymeasuredwithintheapertureoftheidealimage.ThisgivesaStrehlratioof0.085whichissimilartothe Tubbsetal. ( 2002 )measuredStrehlratiosof0:06inLuckyImaging 95

PAGE 96

CrosssectionsofthePSFproducedbySpeckleStabilization,theseeinglimit,andthediffractionlimit.Thediffraction-limitedimagehasbeenarbitrarilyscaledtodemonstratethattheSSPSFhasasimilarFWHM.Notethattheseeing-limithasasimilarFWHMtotheextendedhaloofthespeckle-stabilizationimage. observationswhenusing100%frameselection(SAA).ThereforewendthesesimulationsconrmthattheSSPSFisquitesimilartotheSAAPSFsproducedinrealobservations. Law 2007 ; Hormuthetal. 2008 ; Oscozetal. 2008 )anduseacomparablestandardCCDforSpeckleStabilization.CurrentLuckyImagersuseEMCCDsofsizesaround512x512pixels2,sowewillmodelastandardCCDofthissizefortheSSas 96

PAGE 97

Baldwinetal. 2001 ; Tubbsetal. 2002 ; Law 2007 ; Baldwinetal. 2008 ).ForSAA(100%frameselection)weassumeStrehlof0.085theStrehlcalculatedfromoursimulationsinx2.1asSAAhasasimilarPSFtoSS.ThetotaltimeonsourceisTExpandisdeterminedbythenumberofexposures,nExptimesexposurelengthtExp.WeassumenExp=50000andtExp=0:030secondsmeaningthetotaltimespentonsourceis1500sec.Anexposuretimeof30millisecondsisusedbecausethecoherencelengthoftheatmosphere,generallygivenbyt0r0=v(wherer0istheFriedparameterandvisthebulkwindvelocityofthedominantturbulentlayer( Kernetal. 2000 )),isontheorderof30milliseconds.ThereisnotimeassumedforthereadoutasallcurrentLuckyImagingsystemsuseframe-transferEMCCDsanda30msisgreaterthanorequaltothereadouttimefora512x512pixel2EMCCD( Law 2007 ; Hormuthetal. 2008 ; Oscozetal. 2008 ).Wealsousethetermw,whichisthefractionofneardiffraction-limitedimageskept,whichweassumetobe0.01and0.10forLuckyImagingcasesand1.0forSAA.Thischoice 97

PAGE 99

Keremedjievetal. 2010 ).Withashutterfastenoughtoopenandcloseat>100Hz,aspecklestabilizationsystemwouldbeableensureonlyhigh-Strehldatandtheirwayontothedetector.Inthisway,aSSwouldactlikeareal-timeLuckyImagingselectionalgorithm.Therefore,wealsoincludemodelswhichincludeahigh-speedshuttertocomparetothelowerframe-selection,higher-StrehlLuckyImagingdata.WerefertothistechniqueasSpeckleStabilization+Shutter(SS+S).TheonlymodicationstoEqn.(2)neededtocharacterizeashutter-basedsystemarethatwechangew=0:01and0:10toreectthefractionofLuckyImagingimagestypicallyselected,butconverselywewouldalsogetthehigherStrehls,b=0:30and0:20.Althoughfordarkcurrent,wewouldstillusethefullobservingintervalasthedetectorcontinuestoaccumulatedarkchargewiththeshutterclosed.Onefurthernotewewishtoaddressisthatofguidestars.Forthesemodels,weassumethatthereisabrightguidestarnearenoughthatthereislittleornodegradationintheimagequalityoftheanalysisstar.Wemakethisassumptionbecauseweareonlyinterestedinthetheoreticallimitsofthesetwotechniquesandtoprobethefaintmagnitudesofthesetestsrequiresabright,nearbyguidestarforbothtechniques. 7-2 .Therearesixcurvesinthetoppartofthegure.TheredlinesrepresentLuckyImagingandSAAwhiletheblacklinesdenoteSSandSS+S.Abluedashedline 99

PAGE 100

Keremedjievetal. 2010 ).Therefore,wewilluse100Hzasabaselineandrequire10cycles(withtheshutteropen)tohaveelapsedtogettheminimumnecessarycorrections.ThismeanswedenetheminimumexposuretimeaswTExp;SS=0:1.Usingthisconvention,weseethatSAAcanobservetargetsbrighterthani'=6.87,LuckyImagingat10%frameselectioncanobservetargetsbrighterthani'=7.80andLuckyImagingat1%frameselectioncanobservetargetsbrighterthani'=8.24whereasthecorrespondingSSandSS+Stechniquescannot.Atthefainterend,advantagestoSSandSS+Sbecomeheightened.WeseethatspecklestabilizationtechniqueshavehigherefcienciesandsensitivitiesthanLuckyImaging(againwhereefciencyisdenedasS/Nperobservinginterval).Sensitivityisthedifferenceinmagnitudesatthe3sdetectionlevel.Specically,SSis3.35timesmoreefcientand1.42magnitudesmoresensitivethanSAAattheSAAdetectionlimit.SS+Sat10%selectionis2.40timesmoreefcientand1.10magnitudesmoresensitivethancorrespondingLuckyImagingandSS+Sat1%selectionis1.28timesmoreefcientand0.31magnitudesmoresensitivethanthecorrespondingLuckyImages.Whilethesevaluesarenotexcessivelylarge,theyareinteresting. 100

PAGE 101

ComparisonbetweenSSandSS+SwithSAAandLuckyImaging.AlsoplottedisabluelinedenotingS=N=3,acommondetectionlimit.OnthelowerpartoftheplotaretheratiobetweenS/NSSandS/NLucky.Theseratiosaretheefciencyratios.Alsoplottedonthelowerplotinblueisalinedenotinganefciencyratioof1wherebothsystemsperformthesame. 101

PAGE 102

7-3 .Forthe1%frameselectionLuckyImagingdata,theredoesnotappeartobeanytruncationallthewaydowntothedetectionlimit.IntheSS+S1%data,atruncationaroundi'=20stilloccurswhenskybackgroundbeginstoaffectthedata.ThesemodelsshowthatnoreadnoiseLuckyImagingisactuallyafactorof1.28timesmoreefcientthanSS+SattheSS+Sdetectionlimitandis0.41magnitudesmoresensitive. 102

PAGE 103

7-4 .ThelocationwherephotoncountingswitchesonisimmediatelyapparentinthegureathisisthelocationwheretheS/Nofthespeckletechniquesjumpsinadiscontinuousfashion.Whenthisoccurs,LuckyImagingdataexactlymatchtheSSandSS+Sdataforafewmagnitudes.Inthe1%selectioncase,LuckyImagingactuallytracksSS+Sallthewaydowntothedetectionlimitandevenslightlyout-performsitwithafactorof1.12boostinefciencyandanincreaseof0.14magnitudesinsensitivity. 103

PAGE 104

ComparisonbetweenSpeckleStabilizationwithandwithoutshutterandno-read-noiseSAAandLuckyImaging.InthecaseofLuckyImaging,thereisnotruncationinthefunctionpresented.AlsoplottedisalinedenotingS=N=3,acommondetectionlimit.ThelowerpartoftheplotisthesameasFigure 7-2 ,butinthiscase,thedottedblacklineistheratiooftheS/NSS+ShuttertoLuckyImaging. 104

PAGE 105

Comparisonbetweentechniquesinthecaseofphotoncounting.Thediscontinuousjumpsinthespeckleimagingdataareduetothepointwherephotoncountingbecomesactive.AlsoplottedisalinedenotingS=N=3,acommondetectionlimit.ThelowerpartoftheplotisthesameasFigure 7-2 105

PAGE 106

7-5 wherereadnoiseiszeroandphotoncountingenabled.OurmodelsrevealthatanOLIisbothmoreefcientandsensitiveforLuckyImagingthanSS+SandhasnearlyequivalentperformancewhencomparingSAAtoSS.InparticularwendOLIis1.80timesmoreefcientatthedetectionlimitand0.96magnitudesmoresensitivethanSS+Swith1%frameselectionand1.29timesmoreefcientand0.32magnitudesmoresensitivewith10%frameselection.However,onaverage,SS+Sisstillmarginallymoreefcientatbrightermagnitudes.SSandSAAhavenearlythesamepropertiesintermsofsensitivityandefciency.Overall,thismeansthatwithfairlyfewmodications,itispossibletooptimizeaLuckyImagingsystemsothatitismaximallyefcientandsensitive.Furthermore,oneissuewehavenotaddressedinoursimulationsisaninherentadvantagetoLuckyImaging:becausealltheanalysisisdoneoff-line,inpostprocessingimageselectionalgorithmscanbetunedtomaximizeS/NandStrehlratios.Theobservercanalsodecideatwhatpointtoenablephotoncountinginpostprocessingaswell.ThisgivesLuckyImagingmuchmoreexibilitywithrespecttodataproductsandispartoftheoptimizationprocess. 106

PAGE 107

ComparisonbetweenspecklestabilizationandOLI.Thediscontinuousjumpsinthespeckleimagingdataareduetothepointwherephotoncountingbecomesactive.AlsoplottedisalinedenotingS=N=3,acommondetectionlimit.ThelowerpartoftheplotisthesameasFigure 7-2 107

PAGE 108

Summaryofsimulationresults.EisdenedastheefciencyratioatthelimitingmagnitudeoftheLuckyImagingsystem.DSistheincreaseinsensitivityofSSoverLuckyImaging. SS100% SS+S10% SS+S1% Simulation Eff. 3.35 1.42 1.54 2.40 1.10 1.26 1.28 0.31 1.39 NoReadNoise 1.36 0.36 1.28 1.11 0.12 1.34 0.787 -0.41 1.33 PhotonCounting 2.44 1.04 1.33 1.76 0.70 1.33 0.896 -0.14 1.28 OLI 0.964 -0.04 1.10 0.778 -0.32 1.20 0.556 -0.96 1.24 1024x1024 1.81 0.66 1.86 1.63 0.57 2.11 1.15 0.17 2.22 7-6 wedemonstratethecomparisonbetweenanOLIsystemandaSSsystembothwith1024x1024pixels2detectors.FortheSSdetector,weassumeareadouttimeof4seconds.Inthiscase,becausethereadouttimesaresolongforanEMCCDofthissize,SpeckleStabilizationhasanadvantageinbothsensitivityandefciencyforboththeshutterandnon-shuttercases.WhencomparingSS+StoLuckyImaging,wendthatSS+Sis1.15timesmoreefcientatthefaintestmagnitudeand0.17magnitudesmoresensitivewithanaverageefciencyratioof2.22with1%frameselection.For10%frameselection,thesevaluesare1.63timesmoreefcientattheLuckyImagingdetectionthresholdand0.57moresensitive.InthecaseofSSandSAA,wendSSis1.81timesmoreefcientand0.66magnitudesmoresensitive.Wenotethatbeingabletouseadetectorofthissizehasgreatscienticpotential.At30maspixelsampling,a1024x1024pixels2detectorwouldhaveaFOVof30arcsec.WhilethiskindofFOVisabitlargerthantheisoplanaticpatch, Keremedjievetal. ( 2008 )haveshownthatspecklestabilizationiseffectiveouttooffsetsaslargeas30arcsecondsandwouldbeofhighscienticvalue. 108

PAGE 109

ComparisonbetweenspecklestabilizationandOLIfor1024x1024pixel2detectors.Thediscontinuousjumpsinthespeckleimagingdataareduetothepointwherephotoncountingbecomesactive.AlsoplottedisalinedenotingS=N=3,acommondetectionlimit.ThelowerpartoftheplotisthesameasFigure 7-2 109

PAGE 110

7-1 .BothSSandLuckyImaginghavetheirownstrengths.Ingeneral,SSismoresensitiveandefcientthanspeckleimagingatthefaintestmagnitudeswithnormalframeselectionmeaningitcouldbewell-employedforfaintobjectwork.Additionally,formostmid-magnitudetargetsSSandSS+Sisafactorofp 110

PAGE 111

Batcheldor&Koekemoer ( 2009 )wholookedatvarietyofspectralfeaturesandtelescopeaperturestodeterminegeneralrequirementsoffuturetelescopesandfacilities.Inparticular,theyfounda16-mspace-basedtelescopetobetheidealtoolformeasuringSMBHmassesover90%ofcosmichistory.Inthispaper,werestrictourstudytotheplannedground-basedELTsandtheirinstrumentstodetermineredshiftlimitsandestimatethenumberofaccessibletargets.Wealsorestrictourstudytostellarkinematicsastheseobservationsarepossibletoconductinnearlyeverynon-activegalaxy. 111

PAGE 112

112

PAGE 113

Daviesetal. 2006 ; Nowaketal. 2007 ; Cappellarietal. 2009 ; Krajnovicetal. 2009 ),willtake40minutes,27minutesand14minutesontheGMT,TMTandE-ELTrespectivelytoreachthesameS/Nforthesameangularscales.Infact,theS/Nwillevenbemarginallyhigherowingtothefactthatshorterexposuresmeanlessnoisefromdarkcurrent. 113

PAGE 114

ELTK-bandresolutionsanddistancesatwhichtypical(1108M)SMBHshouldbeobservable.Notealsothatthedistanceisthecomovingdistance.Notethatthesedistancevaluesassumethattherearespectralfeaturesthatcanbeobservedtomeasurethemass.IftheredshiftlimitsoftheCObandheadsareemployed,thenalloftheredshiftlimitsforalltelescopesatboth1108Mand2109Mare0.0331. Dist(Mpc) 19.4 137 0.0335 642 0.163 TMT 15.8 168 0.0414 818 0.209 E-ELT 11.3 239 0.0591 Gultekinetal. ( 2009 ).Foracharacteristicblackholemassof1108Mwithavelocitydispersionof183km/s,theblackhole'ssphereofinuenceis12.3pc.Withknowledgeofthephysicalscalethatneedstoberesolved,weusethediffraction-limitedangularscaleofthetelescopetondthecomovingdistanceforthegivenangularsizedistance.Assumingthesphereofinuencellsatleastoneresolutionelement,wendthedistancesgiveninTable 8-1 .Fromthesevalues,wecanseethattheELTswillbeabletoperformmassmeasurementsof1108MSMBHwellpasttheVirgoclusterandevenpasttheComacluster. 114

PAGE 115

Shethetal. 2003 ; Law 2007 ; Gultekinetal. 2009 ).WhilelargerSMBHhavebeendetected,mostareinactivegalaxieswhichusedifferentmassestimatesthanarewithinthescopeofthispaper.Atthislimit,wegetas=400km/sandasphereofinuenceequalto53.7pc.Assumingtheblackholesphereofinuenceisresolvedbyoneresolutionelementyieldsthevaluesofcolumn5inTable 8-1 .Whenweconvertthesedistancestoredshiftvalues(columns4and6,Table 8-1 )wendthatusingtheE-ELTwewillbeabletoobservecharacteristicgalaxiesouttoz=0:0591(lookbacktimeof0.759Gyr)andinthemostmassivecases,wewillbeabletoobservez=0:334(lookbacktimeof3.57Gyr).Itshouldbenotedhowever,thattheK-bandCObandheadsusedtomeasureSMBHkinematicinuencerangefrom2:294
PAGE 116

PlotofSMBHMassversuslimitingobservableredshiftforKeckandthethreeELTs.Allthreetelescopesreachthesameredshiftlimitofz=0:0331forhighmasseswhentheCObandheadsarenolongerintheK-band.ThediamondsareSMBHmassesthathavebeenmeasuredusingK-bandCObandheadsonexistingfacilities. Table8-2. SMBHmassmeasurementsforavarietyofsystemsdeterminedbyK-bandintegraleldspectroscopy.Theredshiftofthehostgalaxyandreferencesarealsoprovided. Galaxy BHMass(M) z Reference NGC3227 Daviesetal. ( 2006 ) NGC4486a Nowaketal. ( 2007 ) NGC4151 Onkenetal. ( 2007 ) NGC1316 Nowaketal. ( 2008 ) NGC5128 Cappellarietal. ( 2009 ) NGC524 Krajnovicetal. ( 2009 ) NGC2549 Krajnovicetal. ( 2009 ) NGC3368 Nowaketal. ( 2010 ) NGC3489 Nowaketal. ( 2010 ) wedemonstratethatresolutionwillnotbethelimitingfactorforthemostmassiveSMBHmeasurementsandthatcurrentobservationsofSMBHon8and10-meterclasstelescopesareapproachingtheexpectedredshiftlimits. 116

PAGE 117

McConnelletal. ( 2011 )observationsofNGC6086.BecausetheirK-banddatahadhighthermalbackgroundnoise,theyusedCObandheadsinH-bandtomeasurethemassoftheSMBHatthecenterofthisBrightestClusterGalaxy.TheirworkrepresentstherstattempttousefeaturesintheH-bandandtheyreportaSMBHmassof3:6109M.GiventhattheMKO-NIRH-bandrangesfrom1:49mm
PAGE 118

ELTHandK-bandresolutionsanddistancesatwhichtypical(1108M)SMBHshouldbeobservableusingthe1:619mmCOfeature.N/Aisusedwhenthemaximumredshiftisstilllessthan0.254,therequisiteredshiftforthefeaturetomoveintotheK-band.Notethatthedistanceisthecomovingdistance. Dist(Mpc) Dist(Mpc) 15.0 185 0.0455 399 0.0994 19.4 N/A N/A N/A N/A TMT 12.2 229 0.0564 399 0.0994 15.8 N/A N/A N/A N/A E-ELT 8.74 328 0.0813 399 0.0994 11.3 N/A N/A 8-2 .InTable 8-3 wesummarizethelimitsforSMBHwithmassesof1108Mand2109M.WenotethatonlytheE-ELTwillbeabletobenetfromthe1:619mmfeaturebeingredshiftedintotheK-bandforSMBHwithmassesof2109M.However,fromFigure 8-2 weshowthatforSMBHwithmassesgreaterthanthisvalue,GMTandTMTwillbeabletoobservesometargets.Therefore,whileSMBHouttoz=0:337shouldbemeasurable,thereisagapbetween0:09940:377allofthecalciumtripletfeatureswillbelocatedintheJ-band.Ifonlythelongestwavelengthfeatureisused,therequisiteminimumredshiftbecomesz=0:351.SincecurrentAOsystemsarecapableofdiffraction-limitedangularresolutionsatStrehlratiosof10%intheJ-band( vanDametal. 2006 ),theJ-bandcouldbeusedtomeasurethekinematicinuenceofSMBHathigherangular 118

PAGE 119

PlotofSMBHMassversuslimitingobservableredshiftforKeckandthethreeELTs.Fourdifferentlinediagnosticsarepresentedtoshowthetotalrangeofredshiftsaccessiblebythevarioustelescopes.Useofthe1.619mmCOfeatureintheH-bandisdenotedbytheCO(H)inH-bandcurveandwhenthefeatureisredshiftedintotheK-band,thecurveisdenotedbyCO(H)inK-band.UseoftheopticalcalciumtriplelinesintheJ-bandisgivenbytheCaTinJ-bandcurves.Thegapbetween0:0994109M).However,inobservingthesemostmassivegalaxieswewillbeabletomeasureSMBHmassesouttoz=0:565wherethecalciumtripletfeaturegetsshiftedoutoftheJ-band. 119

PAGE 120

1. SMBHmassesshouldbemeasurableouttoz=0:565 Thereisagapbetween0:0994
PAGE 121

MassfunctionofobservableSMBHasafunctionofSMBHmassforKeckandtheELTs.ThenotchesaretheresultofwherethebreakinredshiftduetoalackofusablespectralfeaturesforSMBHmassmeasurementoccurs. 8-4 .Whiletheguredoesindicatethatsmallertelescopesshouldbeabletocarryoutthisresearch,thesensitivitylimitsofthesmalleraperturesislikelytobemoreofalimitingfactor.Wealsonotethattheseresultsareindependentofsight-lines.Thenumberofactuallyobservabletargetscouldbemuchlessduetothephysicallocationofthetelescopes,obscurationfromtheGalacticdiskandotherGalacticphenomena. 121

PAGE 122

PlotoftotalnumberofSMBHobservableasafunctionoftelescopeaperturediameter.Theverticaldashedlinesspecicallydenote8-m,10-m,24.5-m,30-mand42-mapertures. Daviesetal. 2006 ; Nowaketal. 2007 ; Larkinetal. 2006 ; Krajnovicetal. 2009 ).TheIFSsontheseinstrumentseachhaveangularresolutionsof100masandspectralresolutions 122

PAGE 123

Larkinetal. 2010 ),anear-IRIFSwithR4000designedtobeamongtherstgenerationofinstrumentsforthetelescope.Itisplannedtohave4,10,25and50masresolutionsettingsforexibilityinobserving.TheentireFOVisexpectedtobe3.ThesecondinstrumentcapableofmeasuringSMBHonTMTwillbeIRMOS( Eikenberryetal. 2006b ),anear-IRmulti-objectIFUplannedforlaterdeploymentinthesecondorthirdgeneration.Itwillbeabletosimultaneouslyobservealeast10different,2eldswith10000:3meaningthisinstrumentwillcertainlybeusefulforSMBHkinematicmeasurements. 13.5 NIRMOS .pdf/download 123

PAGE 124

PlotofSMBHMassversuslimitingobservableredshiftforthefourrstgenerationinstrumentsfortheELTs.Thegapbetween0:0994
PAGE 125

8-6 ,IpresentthelimitingredshiftsspecklestabilizationwillbeabletoachieveinmeasuringSMBHmasses.IalsoincludewhatSPIFSshouldbeabletoachieveifusedonELTsaswell.WhileSPIFSonanELTwillhavelowerStrehlratiosthanon10-metertelescopes,itshouldstillprovidediffraction-limitedangularresolutionsandcanbeusedtomeasureSMBHmasses.Fromthesecurves,IproduceanewSMBHmassfunctionandpresentitinFigure 8-7 .TherstnoticeabledifferencebetweenthisgureandFigure 8-3 isthatthecurvesareshiftedupwardtoreecttheincreasednumberofblackholesmeasurable.Thesecondnotabledifferenceisthatthenotcheshaveadifferent,lesspronouncedshape.Thisisbecausethegapinredshifthasbeenreduced.WhenIintegrateunderthesecurves,Indthata10-meterclasstelescopeshouldbeabletomeasure1105SMBHandthatGMT,TMTandE-ELTshouldbeabletomeasure2106,3106and1107SMBHrespectivelyusingalllinediagnosticsonallthetelescopes.ThereforeSPIFSwouldenablenearlyafactorofveincreaseinthetotalnumberofSMBHmeasurableoverthenear-IRinstrumentsalone.Whatisinterestingtonote,howeveristhattheremaininggapinaccessibleredshiftsissomewhatarticial.Thisisfortworeasons,therstisthatitispossibleforSPIFStoworkat 125

PAGE 126

PlotofSMBHMassversuslimitingobservableredshiftforKeckandthethreeELTs.ThelinesarethesameisinFigure 8-2 exceptfortheadditionoftheSPIFScurve.TheredshiftgapwherenolinediagnosticsexisttomeasureSMBHmassviathekinematicmodelingtechniquehasbeenreducedto0:177
PAGE 127

MassfunctionofobservableSMBHasafunctionofSMBHmassforKeckandtheELTsincludinggalaxiesobservablebySPIFS.ThenotchesaretheresultofwherethebreakinredshiftduetoalackofusablespectralfeaturesforSMBHmassmeasurementoccurs. potentiallyenableSMBHtobeobservedathigherredhshiftsand/orbridgethecurrentgapbetween0:177
PAGE 128

( 2009 )alsoobservedmanynearbystarsandfoundthatfeaturescouldbedetectedatR2000usingthe3.0-mIRTF. Daviesetal. ( 2010 )tooktheobservationsonestepfurtherandfoundthatmetallicitycouldbederivedfromJ-bandobservationsofredsuper-giants(RSG)atR2000.TheysurmisedthatJ-bandobservationsofRSGcouldbeusedtocharacterizemetalabundancesindistantgalaxiesandestimatedthatintwonightsofobservationona10-mclasstelescope,onecouldgetS=N100forFeI,MgI,SiIandTiIonsfromaRSGat40MpcwhereitwouldhaveanapparentJmagnitudeof19.Todeterminethenecessaryintegrationtimes,weuseobservationsofSMBHkinematicsfromthecalciumtripletline(CaT)asastartingpoint.ComparisonsbetweenCaTlinesandatomicJ-bandfeaturesmakelogicalsensebecauseanybroadeningobservedinthefeaturesisindicativeofkinematicinuence.AnimportantparameterwewillusetoestimatethenecessaryintegrationtimesistheS/Noftheline(S=Nline).TheeasiestwaytocalculatethisvalueisfromtheequivalentwidthsWlandisgivenby: Dlp Pinkneyetal. ( 2003 )wereabletosuccessfullymeasureSMBHingalaxiesusingS=Ncont=30forSTISonHST.TheirspectralresolutionwasR=3800meaningDl=2:2A.WeuseWl=7:7AforCaTinnormalgalaxiesfrom Garcia-Rissmannetal. ( 2005 ).WeestimatetheFWHMfromthevelocitydispersionofagalaxywithaSMBHmassof1108Musingparametersfromthe Gultekinetal. ( 2009 )Msrelation.Thisgivess=183km/sandthusalinewidthof5.2A.PuttingthesevaluesintoEquation 8 ,wendthat Pinkneyetal. ( 2003 )shouldhaveachievedaS=Nlineforthecalciumtripletof165. 128

PAGE 129

Wallaceetal. ( 2000 ); Daviesetal. ( 2010 ).Thelinewidthfora1108Mblackholeat1.25mmis7.6A.WealsoincludeadilutionfactorbecausewhileRSGcontributesignicantlytotheoverallspectrumofagalaxyintheJ-band,thereareothersourcesofcontinuum.Therefore,asanupperlimitweassumethathalfthelightofthegalaxyisfromRSGmeaningthatweapplyadilutionfactoroftwotothelinestrength.Whenweincorporateallthesevalues,wendthatwewillneedtoreachaS=Ncont:=340formeasuringthekinematicinuenceonSMBH.NextweneedtoturntheS=Ncont:intoanestimationofintegrationtimes.Weassumeobservationsarecarriedoutona10-mtelescopewithaspatialresolutionof100mas/pixel(thesameangularresolutioncurrentlyusedinK-bandSMBHmassmeasurements).Weuseduxvaluesfrom Cox ( 1999 )foranassumedsurfacebrightnessneartheblackholesphereofinuenceof13.0mags/arcsec2.Totalsystemefciency(telescope+instrument)isbasedontheOSIRISinstrumentatKeckwhichhasameasuredefciencyof0:10.Withouttakingintoaccountreadnoise,darkcurrentorskybackgroundweemployPoissonstatisticsfornoiseandndthatanintegrationtimeof2:8106seconds(780hours)isrequiredtoreachtheminimumS=Ncont:necessary.Fora30-mand42-mtelescope,thisdropsto3:1105seconds(86hrs)and1:6105seconds(44hrs)respectively.Sincethesenumbersonlyreecton-sourcetime,theydonotincludetherequisiteskyobservationswhichwilldoubletherequiredtelescopetimeforastandardsource-skypattern.Therefore,whileitistheoreticallypossibletomeasureSMBHintheJ-band,itisnotefcient.Weemphasizeagainthatthesecalculationsassume100masspatialresolutionaresolutionalreadyemployedbyK-bandobservationsofSMBH.Becauseofthisfact,the200xincreaseinintegrationtimerequiredtoachievetheexactsamescienceasafourhoursofK-bandobservation( Daviesetal. 2006 ; Cappellarietal. 2009 )makestheJ-bandinefcientfor 129

PAGE 130

Daviesetal. ( 2010 )foundthereisarelationshipbetweenWlandmetallicityinmanyJ-bandlines.Whilethisisdesirablefromanabundancecharacterizationperspective,itposesanadditionalproblemforSMBHdetections.Toavoidintroducingadditionalsystematicerrors,thisfactrequiresthemetallicityofstarsinthecoreofthetargetgalaxybeknownbeforehandandthatstarsofsimilarmetallicitytothetargetgalaxybeobservedforcalibration.Whenoneconsidersthelongexposuresnecessaryandtheadditionalcalibrationdifculties,theJ-bandbecomesunappealingforSMBHdetections.SowhileitistheoreticallypossibletoconductthisworkintheJ-band,itisnotpractical. Mannuccietal. ( 2001 )performedastudyofrest-frameJ-bandspectraofgalaxiesbuttheirobservationsweremetwithlimitedsuccess.Theyreportedspectraofnon-active,nearbygalaxiesintheJ-band.Theyconductedobservationsof25galaxies(vegalaxiesofvedifferentmorphologicaltypesacrosstheHubblediagram)toproducecompositespectraofE,S0,Sa,Sb,andScgalaxies.TheirnalmedianaveragedJ-bandspectrahadaS=N30andaspectralresolutionofR300.TheydidnotidentifyanyfeaturesintheJ-banddatacitingthefactthattheirJ-banddatawereoflowerqualitythantheirHandK-bandspectraandthatJ-bandstellarspectralanalysisisinitsinfancy.Therefore,ourobservationsseektoexaminelocalgalaxiesathigherspectralresolutionandS=Nthanpreviouslyobtainedandseeifanyfeaturesareclearlydiscernible.Thepresenceorlackoffeaturescanhelpconstraintheanalysisinx3.1.Itshouldbenoted,however,thattheseobservationsarenotmeanttoprovideacomprehensivesurveyoflocalgalaxiesintheJ-band.Weconducttheseobservationsasaprobetoseeifanythinginterestinghasbeenmissedbypreviousworkanddetermineifmorefollow-upobservationsarewell-motivatedatthisspectralresolution. 130

PAGE 131

8-4 .Ourscienceexposureswere300secondslongandwetooktheminanABBAABscience-skypatterntoaccountforthetimevariablenatureoftheinfraredskybackground.BecauseoftheexureintheFLAMINGOSsystem,weacquiredatandHeNeArcalibrationlampdataaftereachsetofthreescienceexposures.Thenweobtainedspectraofatelluricabsorptionstarofsimilarairmassforabsorptionlinecalibration. 131

PAGE 132

Thesemi-majoraxislengthisdenotedbya.Jmagnitudesaretakenfromthe2MASSsurvey(Skrutskieetal.2006).Redshifts,z,wereacquiredfromtheNASA/IPACExtragalacticDatabase(NED). Object Type (Vega) (minutes) NGC7619 E 2.9 8.992 0.012549 45 70 NGC6824 Sb 2.1 9.228 0.011294 35 34 NGC83 E 1.6 10.279 0.020771 45 25 NGC13 Sc 2.7 10.667 0.016038 45 24 WeperformedinitialreductionofthedatawiththeFloridaAnalysisToolBornofYearningforhighqualityspectroscopydata(FATBOY)reductionpackage(Warneretal.2008).UsingFATBOY,werstateldcorrectedwithlampats.Next,wematchedeachscienceexposuretoitscorrespondingoffsourceskyframe(taken10'away)andperformedskysubtraction.Afterskysubtraction,werectiedthespectralimagesandalignedthewavelengthsusingatandlampdata.Fromtheresulting,calibratedslices,weextractedrawspectra.WecalibratedfortelluricabsorptionbydividinggalaxyspectrabystandardG0-5Vstars.Thenwemultipliedthegalaxyspectrabyablackbodycorrespondingtoeachcalibrationstar'sintrinsictemperature.Toproducethecompositespectra,weco-addedspectrafromtheimageslices.However,notallslicesproducedusabledataandblindlysummingthemwouldintroducesubstantialerrors.Toselectthebestspectraforsumming,theS/Nwascalculatedforeachsliceinthe1:22mm10wereadded.ForNGC6824weusedacutoffofS/N>9andforNGC83andNGC13wechoseS/N>7. 8-8 .Databetween1:1mm25isconsistentwithpreviousndingsfromMannuccietal.(2001)atR300

PAGE 133

andS/N>30.Therefore,atalmostvetimesgreaterspectralresolution,wendnoclearlydiscerniblefeatures.InFigure 8-9 ,wepresentcomparisonsbetweenourbest-qualityspectrumofNGC7619andcurrentgalaxystellarpopulationmodelsproducedby Maraston ( 2005 )and Bruzual&Charlot ( 2003 ).Tousethemodels,weneedboththemetallicityandageofeachgalaxy;however,thesevalueshaveonlybeenpublishedforNGC7619. Bregmanetal. ( 2006 )foundNGC7619tobe14:42:2GyroldandhaveZ=0:210:03.Usingtheseparameters,andassumingane-foldingtimeof1GyrforaSaltpeterIMF,wecanproducemodelgalaxyspectra.Bothmodelshaveverylowresolution,R250,sowereducedtheresolutionofourdatacorrespondingly.Becauseourdatahavenotbeenuxcalibratedwealsoscaledthemodeltomatchtheuxlevelsofourdata.Thetwasfoundbyminimizingthec2differencebetweenourdataandaconstanttimesthemodelux.For Maraston ( 2005 )c2min=1:24andfor Bruzual&Charlot ( 2003 )c2min=1:09.Boththedegradedandfullresolution,togetherwiththemodelsarepresentedinFigure 8-9 133

PAGE 134

Maraston ( 2005 )modelthatbesttsitspropertiesinredandthe Bruzual&Charlot ( 2003 )modelinblue.TheupperspectrumisthefullresolutionresultandthelowerspectrumisadegradedversionoftheNGC7619spectrumtomatchtheresolutionofthemodels. Athighresolution,bothmodelscloselymatchthecontinuumandatthedegradedresolutionthereisnearlyaperfectmatch.Therefore,ourobservationsconrmtheoverallshapeofbothmodelsanddemonstratethateithermodelissufcienttocharacterizetheJ-band.Ourdatacanbeusedtoconstrainfuturehigherresolutionmodelsofgalaxyspectra,butshowthatsuchmodelswilllikelycontinuetobefeaturelessatR1400.Becausenoobviousfeaturesarepresentinourreducedspectra,wedecidedtocalculateourdetectionlimits.Todothis,weelectedtoaddarticialfeaturestoNGC7619'sspectrumandndthelineequivalentwidthsnecessarytomeeta3or5sdetectionthreshold.Weassumedthefeatureswereinabsorption,Gaussianinshape,andhadaFWHMequaltotheresolutionlimitofourdata(R1400).ExaminingourspectrainFigure 8-8 weseethattherearetwonoiseregimestoourdata.Thereisanoisiersectionblue-wardof1:215mmandacleanersectionred-ward.Therefore,wechosetondourlimitingEWinbothregimes.The3and5serrorsin 134

PAGE 135

Wallaceetal. ( 2000 )aswellastheanalysisofourprevioussection.ManyofthestrongeststellarfeaturesdetectedinobservationsofstellarJ-bandfeatureshaveanEW2A.Sincesomedilutionofthesestellarabsorptionfeaturesisexpectedonthegalacticscale,weconcludethattheyarerightbelowourdetectionlimits.Ourobservationsdemonstratethatinthelocaluniverse,emissionandabsorptionlinesremainelusiveforcurrentmedium-resolution,near-IRinstrumentsatS/N<70.Therefore,itispossiblethatinstrumentationmatchingorexceedingtheWallaceetal.(2000)R3000observations,suchasFLAMINGOSIIatGeminiSouth( Eikenberryetal. 2006a ),NIRSPEConKeck( Magorrianetal. 1998 ),orLUCIFERattheLBT( Mandeletal. 2007 )willbeabledetectspectralfeaturesinlocalgalaxies.SinceourS=N>50J-bandspectracontainnodiscerniblefeaturesat2700secondsofon-sourceintegrationwithanR=1300spectrographona4-mtelescopewitha528arcsec2aperture,wecanconrmthendingsfromx8.3.1thattheJ-bandrequiresextremelylongintegrationtimestodetectanyusablefeatures. 135

PAGE 136

Batcheldor&Koekemoer ( 2009 )aswendsimilarredshiftlimitsfortheELTs.Ourwork,however,alsoestimatesthetotalnumberofgalaxiesobservablebythesetelescopesandtakesintoaccountwhatredshiftrangesshouldbeaccessible.Unfortunately,wendthatrest-frameJ-bandfeaturesaretooobservationallydemandingtobeeffectiveforSMBHresearch.Betweentheoreticalcalculationsandobservationalconstraints,wendthefeaturesareexceptionallydifculttodetect.Iftargetsintheredshiftrangeof0:177
PAGE 137

5-4 )whereintheseeing-limitedcasethetwostarswereblended,butwhenusingSS,theywereclearlyseparated.IalsofoundthatwhiletheSPIFS-POCwasausefuldemonstratorofSSpotential,itwasnoteffectiveforscienceoperations.Themainlimitingfactoronthisinstrumentwasthefaststeeringmirrorwhichhadsignicantringingrequiringalongtimetocompensatefor.Thenextgenerationofspecklestabilization,theS3DwillremedythisissuebyhavingabetterFSM.TheS3Dwillalsolikelybeabletoacquireinterestingscienceresultsasitwillhaveahigh-qualitysciencecameraandahigh-speedshuttertoincreaseStrehlratios.Alltold,thepotentialofspecklestabilizationisjustbeginningtoberealized. 137

PAGE 138

138

PAGE 139

139

PAGE 140

140

PAGE 141

141

PAGE 142

Lawetal. ( 2009 )wereabletorealizeaddedbenetstoluckyimagingfromthis,SSmightalsobeabletogethigherStrehl-ratiosandincreasedefciency.ItwillalsobeinterestingtoseewhatSSwillbeabletodoonELTsandthescienceitwillbeabletoaccess.Specklestabilizationhasthepotentialtoanswermanyoutstandingquestionsinastronomy.Inthisdissertation,Ihaveshownthatitisboththeoreticallyandexperimentallyvalidandthatinthenextfewyearsthetechniquesshouldbegettingscienticreturns. 142

PAGE 143

Abell,G.O.,&Margon,B.1979,Nature,279,701 Baldwin,J.E.,Tubbs,R.N.,Cox,G.C.,Mackay,C.D.,Wilson,R.W.,&Andersen,M.I.2001,A&A,368,L1 Baldwin,J.E.,Warner,P.J.,&Mackay,C.D.2008,A&A,480,589 Barth,A.J.,Ho,L.C.,Rutledge,R.E.,&Sargent,W.L.W.2004,ApJ,607,90 Batcheldor,D.,&Koekemoer,A.M.2009,PASP,121,1245 Bates,R.H.T.,&Cady,F.M.1980,OpticsCommunications,32,365 Bell,E.F.,McIntosh,D.H.,Katz,N.,&Weinberg,M.D.2003,ApJS,149,289 Bifano,T.G.,&Stewart,J.B.2005,inPresentedattheSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)Conference,Vol.5895,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,ed.M.T.Valley&M.A.Vorontsov,147 Bregman,J.N.,Temi,P.,&Bregman,J.D.2006,ApJ,647,265 Bruzual,G.,&Charlot,S.2003,MNRAS,344,1000 Cappellari,M.,Neumayer,N.,Reunanen,J.,vanderWerf,P.P.,deZeeuw,P.T.,&Rix,H.2009,MNRAS,394,660 Cox,A.N.1999,Allen'sAstrophysicalQuantities:FourthEdition(Springer) Cuby,J.,etal.2010,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7735,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Daemgen,S.,Hormuth,F.,Brandner,W.,Bergfors,C.,Janson,M.,Hippler,S.,&Henning,T.2009,A&A,498,567 Davies,B.,Kudritzki,R.,&Figer,D.F.2010,MNRAS,407,1203 Davies,R.I.,etal.2006,ApJ,646,754 Eikenberry,S.,etal.2006a,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.6269,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Eikenberry,S.,etal.2006b,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.6269,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Eikenberry,S.S.,Cameron,P.B.,Fierce,B.W.,Kull,D.M.,Dror,D.H.,Houck,J.R.,&Margon,B.2001,ApJ,561,1027 143

PAGE 144

Ferrarese,L.,&Merritt,D.2000,ApJL,539,L9 Filippenko,A.V.1982,PASP,94,715 Filippenko,A.V.,&Ho,L.C.2003,ApJL,588,L13 Ford,E.B.2009,NewAstronomy,14,406 Freedman,W.L.,&Madore,B.F.2010,ARA&A,48,673 Fried,D.L.1978,JournaloftheOpticalSocietyofAmerica(1917-1983),68,1651 Frigo,M.,&Johnson,S.G.2005,ProceedingsoftheIEEE,93,216,specialissueonProgramGeneration,Optimization,andPlatformAdaptation Garcia-Rissmann,A.,Vega,L.R.,Asari,N.V.,CidFernandes,R.,Schmitt,H.,GonzalezDelgado,R.M.,&Storchi-Bergmann,T.2005,MNRAS,359,765 Gebhardt,K.,Rich,R.M.,&Ho,L.C.2005,ApJ,634,1093 Gebhardt,K.,etal.2000,ApJL,539,L13 Greene,J.E.,&Ho,L.C.2007,ApJ,670,92 Gultekin,K.,etal.2009,ApJ,698,198 Hinkle,K.H.1978,ApJ,220,210 Hormuth,F.,Hippler,S.,Brandner,W.,Wagner,K.,&Henning,T.2008,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7014,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Howell,S.B.2000,HandbookofCCDAstronomy(CambridgeObservingHandbooksforResearchAstronomers)(CambridgeUniversityPress) Keremedjiev,M.,&Eikenberry,S.S.2011,PASP,123,213 Keremedjiev,M.,Eikenberry,S.S.,&Carson,J.C.2008,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7014,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Keremedjiev,M.,Eikenberry,S.S.,&Carson,J.C.2010,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7735,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Kern,B.,Laurence,T.A.,Martin,C.,&Dimotakis,P.E.2000,AppliedOpticsLP,39,4879 144

PAGE 145

Kormendy,J.,&Richstone,D.1995,ARA&A,33,581 Krajnovic,D.,McDermid,R.M.,Cappellari,M.,&Davies,R.L.2009,MNRAS,399,1839 Larkin,J.,etal.2006,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.6269,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Larkin,J.E.,etal.2010,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7735,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Law,N.M.2007,TheObservatory,127,71 Law,N.M.,Hodgkin,S.T.,&Mackay,C.D.2006,MNRAS,368,1917 .2008,MNRAS,384,150 Law,N.M.,Hodgkin,S.T.,Mackay,C.D.,&Baldwin,J.E.2005,AstronomischeNachrichten,326,1024 Law,N.M.,etal.2009,ApJ,692,924 LeLouarn,M.2002,MNRAS,334,865 Lombardi,G.,Zitelli,V.,Ortolani,S.,&Pedani,M.2006,PASP,118,1198 .2007,PASP,119,292 Magorrian,J.,etal.1998,AJ,115,2285 Mandel,H.,etal.2007,AstronomischeNachrichten,328,626 Mannucci,F.,Basile,F.,Poggianti,B.M.,Cimatti,A.,Daddi,E.,Pozzetti,L.,&Vanzi,L.2001,MNRAS,326,745 Maraston,C.2005,MNRAS,362,799 Margon,B.1984,ARA&A,22,507 Mason,B.D.,Wycoff,G.L.,Hartkopf,W.I.,Douglass,G.G.,&Worley,C.E.2001,AJ,122,3466 McConnell,N.J.,Ma,C.,Graham,J.R.,Gebhardt,K.,Lauer,T.R.,Wright,S.A.,&Richstone,D.O.2011,ApJ,728,100 Merz,H.2007,CUFFTvsFFTWcomparison,http://www.sharcnet.ca/merz/CUDA benchFFT/ Meyer,M.R.,Edwards,S.,Hinkle,K.H.,&Strom,S.E.1998,ApJ,508,397 145

PAGE 146

Nowak,N.,Saglia,R.P.,Thomas,J.,Bender,R.,Pannella,M.,Gebhardt,K.,&Davies,R.I.2007,MNRAS,379,909 Nowak,N.,Thomas,J.,Erwin,P.,Saglia,R.P.,Bender,R.,&Davies,R.I.2010,MNRAS,403,646 Noyola,E.,Gebhardt,K.,Kissler-Patig,M.,Lutzgendorf,N.,Jalali,B.,deZeeuw,P.T.,&Baumgardt,H.2010,ApJL,719,L60 Onken,C.A.,etal.2007,ApJ,670,105 Oscoz,A.,etal.2008,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7014,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Peebles,P.J.E.1972,ApJ,178,371 Peterson,B.M.,etal.2005,ApJ,632,799 Pinkney,J.,etal.2003,ApJ,596,903 Rampy,R.,Gavel,D.,Dillon,D.,&Thomas,S.2010,inPresentedattheSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)Conference,Vol.7736,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Rayner,J.T.,Cushing,M.C.,&Vacca,W.D.2009,ApJS,185,289 Rusli,S.P.,Thomas,J.,Erwin,P.,Saglia,R.P.,Nowak,N.,&Bender,R.2011,MNRAS,410,1223 Sheth,R.K.,etal.2003,ApJ,594,225 Simons,D.A.,&Tokunaga,A.2002,PASP,114,169 Thatte,N.,etal.2010,inSocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries,Vol.7735,SocietyofPhoto-OpticalInstrumentationEngineers(SPIE)ConferenceSeries Thomas,S.2005,NOAO/NSONewsletter,81,28 Tubbs,R.N.,Baldwin,J.E.,Mackay,C.D.,&Cox,G.C.2002,A&A,387,L21 vanDam,M.A.,etal.2006,PASP,118,310 vanderMarel,R.P.1994,ApJL,432,L91 vanderMarel,R.P.,&Anderson,J.2010,ApJ,710,1063 146

PAGE 147

147

PAGE 148

MarkKeremedjievwasborninDoylestown,PennsylvaniainJuneof1984.AlthoughhegrewupinBozeman,Montanaashisparentsmovedtherein1988.UnderthestarsandmountainsoftheBigSkystatehegrewanappreciationforastronomyandnature.MarkattendedCornellUniversitywherehedidresearchwithProfessorJimHouck,Dr.LeiHaoandDr.GregSloanandreceivedaBAwithhonorsinAstronomy.HerowedvarsitylightweightcrewandalsoreceivedtheCransonW.andEdnaB.ShelleyAwardsforExcellenceinUndergraduateandGraduateResearchinAstronomy.MarkhasspentthelastfourandahalfyearsinGainesville,FloridaattheUniversityofFloridaworkingonhisPh.D.inAstronomy.Hehasbeenactiveinthedepartment'sGraduateAstronomyOrganizationandservedasPresidentfrom2009-2010.Markisanavidtriathleteandhascompletedtwohalf-IronmanracesontheroadtoafullIronman.HeishappilymarriedtoLaurenYoungandlooksforwardtoalongandprosperousfuturewithher. 148