COVER SHEET FOR PROPOSAL TO THE NATIONAL SCIENCE FOUNDATIONFOR NSF USE ONLY

NSF PROPOSAL NUMBER

DATE RECEIVED NUMBER OF COPIES DIVISION ASSIGNED FUND CODE DUNS# (Data Universal Numbering System) FILE LOCATION

FOR CONSIDERATION BY NSF ORGANIZATION UNIT(S) (Indicate the most specific unit known, i.e. program, division, etc.)

PROGRAM ANNOUNCEMENT/SOLICITATION NO./CLOSING DATE/if not in response to a program announcement/solicitation enter NSF 11-1

EMPLOYER IDENTIFICATION NUMBER (EIN) ORTAXPAYER IDENTIFICATION NUMBER (TIN)

SHOW PREVIOUS AWARD NO. IF THIS ISA RENEWALAN ACCOMPLISHMENT-BASED RENEWAL

IS THIS PROPOSAL BEING SUBMITTED TO ANOTHER FEDERALAGENCY? YES NO IF YES, LIST ACRONYM(S)

NAME OF ORGANIZATION TO WHICH AWARD SHOULD BE MADE ADDRESS OF AWARDEE ORGANIZATION, INCLUDING 9 DIGIT ZIP CODE

AWARDEE ORGANIZATION CODE (IF KNOWN)

IS AWARDEE ORGANIZATION (Check All That Apply) SMALL BUSINESS MINORITY BUSINESS IF THIS IS A PRELIMINARY PROPOSAL(See GPG II.C For Definitions) FOR-PROFIT ORGANIZATION WOMAN-OWNED BUSINESS THEN CHECK HERE

NAME OF PRIMARY PLACE OF PERF ADDRESS OF PRIMARY PLACE OF PERF, INCLUDING 9 DIGIT ZIP CODE

TITLE OF PROPOSED PROJECT

REQUESTED AMOUNT$

PROPOSED DURATION (1-60 MONTHS)

monthsREQUESTED STARTING DATE SHOW RELATED PRELIMINARY PROPOSAL NO.

IF APPLICABLE

CHECK APPROPRIATE BOX(ES) IF THIS PROPOSAL INCLUDES ANY OF THE ITEMS LISTED BELOWBEGINNING INVESTIGATOR (GPG I.G.2)DISCLOSURE OF LOBBYING ACTIVITIES (GPG II.C.1.e)PROPRIETARY & PRIVILEGED INFORMATION (GPG I.D, II.C.1.d)HISTORIC PLACES (GPG II.C.2.j)EAGER* (GPG II.D.2) RAPID** (GPG II.D.1)VERTEBRATE ANIMALS (GPG II.D.6) IACUC App. DatePHS Animal Welfare Assurance Number

HUMAN SUBJECTS (GPG II.D.7) Human Subjects Assurance NumberExemption Subsection or IRB App. DateINTERNATIONAL COOPERATIVE ACTIVITIES: COUNTRY/COUNTRIES INVOLVED(GPG II.C.2.j)

HIGH RESOLUTION GRAPHICS/OTHER GRAPHICS WHERE EXACT COLORREPRESENTATION IS REQUIRED FOR PROPER INTERPRETATION (GPG I.G.1)

PI/PD DEPARTMENT PI/PD POSTAL ADDRESS

PI/PD FAX NUMBER

NAMES (TYPED) High Degree Yr of Degree Telephone Number Electronic Mail Address

PI/PD NAME

CO-PI/PD

CO-PI/PD

CO-PI/PD

CO-PI/PD

Page 1 of 2

1211356AST - EXTRAGALACTIC ASTRON & COSMOLO

NSF 05-608 11/15/11

996000354

University of Hawaii

0016105000

ORS, Sakamaki D-200HONOLULU, HI 968222303US

Institute for AstronomyInstitute for Astronomy Institute for Astronomy2680 Woodlawn DriveHonolulu ,HI ,968221839 ,US.

Merger Modeling with IDENTIKIT

312,226 36 07/01/12

Institute for Astronomy

808-988-2790

2680 Woodlawn Drive

Honolulu, HI 96822United States

Joshua E Barnes PhD 1984 808-956-8138 barnes@ifa.hawaii.edu

965088057

Electronic Signature

11/14/2011 1 03020000 AST 1217 07/03/2012 5:03pm S

CERTIFICATION PAGE

Certification for Authorized Organizational Representative or Individual Applicant:By signing and submitting this proposal, the Authorized Organizational Representative or Individual Applicant is: (1) certifying that statements made herein are true and complete to the best of his/her knowledge; and (2) agreeing to accept the obligation to comply with NSF award terms and conditions if an award is made as a result of this application. Further, the applicant is hereby providing certifications regarding debarment and suspension, drug-free workplace, lobbying activities (see below), responsible conduct of research, nondiscrimination, and flood hazard insurance (when applicable) as set forth in the NSF Proposal & Award Policies & Procedures Guide, Part I: the Grant Proposal Guide (GPG) (NSF 11-1). Willful provision of false information in this application and its supporting documents or in reports required under an ensuing award is a criminal offense (U. S. Code, Title 18, Section 1001).

Conflict of Interest Certification In addition, if the applicant institution employs more than fifty persons, by electronically signing the NSF Proposal Cover Sheet, the Authorized Organizational Representative of the applicant institution is certifying that the institution has implemented a written and enforced conflict of interest policy that is consistent with the provisions of the NSF Proposal & Award Policies & Procedures Guide, Part II, Award & Administration Guide (AAG) Chapter IV.A; that to the best of his/her knowledge, all financial disclosures required by that conflict of interest policy have been made; and that all identified conflicts of interest will have been satisfactorily managed, reduced or eliminated prior to the institution’s expenditure of any funds under the award, in accordance with the institution’s conflict of interest policy. Conflicts which cannot be satisfactorily managed, reduced or eliminated must be disclosed to NSF.

Drug Free Work Place Certification By electronically signing the NSF Proposal Cover Sheet, the Authorized Organizational Representative or Individual Applicant is providing the Drug Free Work Place Certification contained in Exhibit II-3 of the Grant Proposal Guide.

Debarment and Suspension Certification (If answer "yes", please provide explanation.)

Is the organization or its principals presently debarred, suspended, proposed for debarment, declared ineligible, or voluntarily excluded from covered transactions by any Federal department or agency? Yes No

By electronically signing the NSF Proposal Cover Sheet, the Authorized Organizational Representative or Individual Applicant is providing the Debarment and Suspension Certification contained in Exhibit II-4 of the Grant Proposal Guide.

Certification Regarding LobbyingThe following certification is required for an award of a Federal contract, grant, or cooperative agreement exceeding $100,000 and for an award of a Federal loan or a commitment providing for the United States to insure or guarantee a loan exceeding $150,000.

Certification for Contracts, Grants, Loans and Cooperative AgreementsThe undersigned certifies, to the best of his or her knowledge and belief, that:(1) No federal appropriated funds have been paid or will be paid, by or on behalf of the undersigned, to any person for influencing or attempting to influence an officer or employee of any agency, a Member of Congress, an officer or employee of Congress, or an employee of a Member of Congress in connection with the awarding of any federal contract, the making of any Federal grant, the making of any Federal loan, the entering into of any cooperative agreement, and the extension, continuation, renewal, amendment, or modification of any Federal contract, grant, loan, or cooperative agreement.(2) If any funds other than Federal appropriated funds have been paid or will be paid to any person for influencing or attempting to influence an officer or employee of any agency, a Member of Congress, an officer or employee of Congress, or an employee of a Member of Congress in connection with this Federal contract, grant, loan, or cooperative agreement, the undersigned shall complete and submit Standard Form-LLL, ‘‘Disclosure of Lobbying Activities,’’ in accordance with its instructions.(3) The undersigned shall require that the language of this certification be included in the award documents for all subawards at all tiers including subcontracts, subgrants, and contracts under grants, loans, and cooperative agreements and that all subrecipients shall certify and disclose accordingly.

This certification is a material representation of fact upon which reliance was placed when this transaction was made or entered into. Submission of this certification is a prerequisite for making or entering into this transaction imposed by section 1352, Title 31, U.S. Code. Any person who fails to file the required certification shall be subject to a civil penalty of not lessthan $10,000 and not more than $100,000 for each such failure.

Certification Regarding Nondiscrimination By electronically signing the NSF Proposal Cover Sheet, the Authorized Organizational Representative is providing the Certification Regarding Nondiscrimination contained in Exhibit II-6 of the Grant Proposal Guide.

Certification Regarding Flood Hazard Insurance Two sections of the National Flood Insurance Act of 1968 (42 USC §4012a and §4106) bar Federal agencies from giving financial assistance for acquisition or construction purposes in any area identified by the Federal Emergency Management Agency (FEMA) as having special flood hazards unless the: (1) community in which that area is located participates in the national flood insurance program; and(2) building (and any related equipment) is covered by adequate flood insurance.

By electronically signing the NSF Proposal Cover Sheet, the Authorized Organizational Representative or Individual Applicant located in FEMA-designated special flood hazard areas is certifying that adequate flood insurance has been or will be obtained in the following situations: (1) for NSF grants for the construction of a building or facility, regardless of the dollar amount of the grant; and(2) for other NSF Grants when more than $25,000 has been budgeted in the proposal for repair, alteration or improvement (construction) of a building or facility.

Certification Regarding Responsible Conduct of Research (RCR) (This certification is not applicable to proposals for conferences, symposia, and workshops.) By electronically signing the NSF Proposal Cover Sheet, the Authorized Organizational Representative of the applicant institution is certifying that, in accordance with the NSF Proposal & Award Policies & Procedures Guide, Part II, Award & Administration Guide (AAG) Chapter IV.B., the institution has a plan in place to provide appropriate training and oversight in the responsible and ethical conduct of research to undergraduates, graduate students and postdoctoral researchers who will be supported by NSF to conduct research. The undersigned shall require that the language of this certification be included in any award documents for all subawards at all tiers.

AUTHORIZED ORGANIZATIONAL REPRESENTATIVE SIGNATURE DATENAME

TELEPHONE NUMBER ELECTRONIC MAIL ADDRESS FAX NUMBER

fm1207rrs-07

* EAGER - EArly-concept Grants for Exploratory Research ** RAPID - Grants for Rapid Response Research

Page 2 of 2

Naomi R Mitake Nov 14 2011 4:53PMElectronic Signature

808-956-3105 mitake@hawaii.edu 808-956-9081

Merger Modeling with IDENTIKIT

IDENTIKIT simulates disk-galaxy encounters and mergers using a combination of self-consistentand test-particle techniques. Whereas a standard N-body simulation of a galaxy encounter imposesa specific encounter geometry, a single IDENTIKIT simulation can simultaneously model outcomesfor all possible disk orientations. Initially developed as a computational shortcut, IDENTIKIT alsoo!ers new and interesting ways to investigate galaxy encounters. Specific research subjects coveredin this proposal are:

• Modeling real galaxies. After testing on artificial data, IDENTIKIT is now ready to move“beyond the lab” by modeling real systems. Preliminary models have now been developedfor several well-observed mergers; the experience gained can help clarify which encounters aremost amenable to modeling and the kinds of observational data necessary.

• Automatic reconstruction. IDENTIKIT defines a mapping from the observable morphol-ogy and kinematics back to the initial conditions. This can provide direct solutions for theinitial orientations of the interacting disks; it also o!ers a way to automate part or even allof the matching process.

• Uniqueness and uncertainty. Given the finite resolution and S/N of real data, and ourlimited knowledge of the true structure of the progenitor galaxies, it’s inevitable that matchesto real systems will be uncertain. This investigation will delimit these uncertainties andexamine the possible existence of multiple solutions.

• Mass models. Tidal features provide direct probes of halo structure and potential welldepth. By averaging all disk orientations, results independent of encounter geometry willplace this relationship on a firmer footing. Simulated encounters with various mass ratios,halo profiles, etc will be reconstructed to examine sensitivity to these parameters. Finally,the results will be applied to models of real interacting galaxies.

Intellectual Merit. This project will further develop a powerful methodology for data-basedmodeling of real interacting galaxies. A direct consequence will be a substantial increase in thenumber of systems with detailed dynamical models. The research will address issues of uniqueness,with a particular focus on constraining dark matter.Broader Impacts. This project will have impacts both within and beyond the astronomical com-munity. Within the community, it provides a web-based tool for dynamical modeling of interactinggalaxies, and makes the software and numerical data needed publicly available. The proposed workprovides a training-ground for astronomy PhD students to obtain experience in both numericalsimulation and analysis of observational data. Finally, an interactive tutorial on galaxy mergingwill be developed for outreach to high-school students in Hawai‘i.

TABLE OF CONTENTSFor font size and page formatting specifications, see GPG section II.B.2.

Total No. of Page No.*Pages (Optional)*

Cover Sheet for Proposal to the National Science Foundation

Project Summary (not to exceed 1 page)

Table of Contents

Project Description (Including Results from PriorNSF Support) (not to exceed 15 pages) (Exceed only if allowed by aspecific program announcement/solicitation or if approved inadvance by the appropriate NSF Assistant Director or designee)

References Cited

Biographical Sketches (Not to exceed 2 pages each)

Budget (Plus up to 3 pages of budget justification)

Current and Pending Support

Facilities, Equipment and Other Resources

Special Information/Supplementary Documents(Data Management Plan, Mentoring Plan and Other Supplementary Documents)

Appendix (List below. )(Include only if allowed by a specific program announcement/solicitation or if approved in advance by the appropriate NSFAssistant Director or designee)

Appendix Items:

*Proposers may select any numbering mechanism for the proposal. The entire proposal however, must be paginated.Complete both columns only if the proposal is numbered consecutively.

1

1

14

1

2

6

1

2

2

Merger Modeling with IDENTIKIT

1 Scientific Background

Tidal encounters and mergers are engines of galaxy evolution. Non-equilibrium features such asbridges and tails (Toomre & Toomre 1972) unambiguously signal recent and ongoing tidal inter-actions between disk galaxies. Galaxies which fall together as a result of their mutual attractioneventually merge due to dynamical friction, forming “spheroidal heaps of stars” (Toomre 1977).Dark halos increase galactic merger cross-sections and accelerate orbital decay (White 1978); halosalso act as momentum sinks, enabling galaxies to merge even before their tails have dispersed, andabsorbing the initial angular momentum of the luminous components (Barnes 1988).

Interacting galaxies have complex structures, and regions of intense star formation as well asextreme dust extinction are common. Detailed modeling of specific interacting systems o!ers apowerful tool to help interpret observational data, disentangling projection e!ects and resolvingthree-dimensional morphology. Moreover, dynamical modeling provides access to the past historyand future evolution of merging galaxies. A dynamical model of a merger can serve as a frameworkunifying disparate observations into a coherent picture.

Driven by evidence linking collisions to various forms of galactic activity (Sanders & Mirabel1996), recent simulations have included increasing complex “sub-grid” prescriptions for star forma-tion, feedback, and AGN. These prescriptions should be tested by detailed data-driven modeling ofspecific systems. Unless such testing can be done, confidence in the models is misplaced. Reliableapproximations for initial conditions are needed to set up these tests. Finding initial conditions forspecific systems is labor-intensive, and faster approaches are needed. Fortunately, the problem offinding initial conditions is essentially stellar-dynamical; once a good model has been constructed,various treatments of sub-grid physics may be implemented without compromising the underlyingdynamics.

To date, modeling has shown that interacting systems can be reproduced by tidal encountersof disk galaxies, but there has been limited work on the question of uniqueness. It’s likely thatsome range of initial conditions reproduce the observed morphology and kinematics of interactinggalaxies; this range needs to be quantified. As yet, it’s not known if disjoint regions of parameterspace will often yield equally good matches, or if the range of options can usually be narroweddown to a single “island” of similar initial states.

1.1 Modeling mergers

The goal of merger modeling is to launch two normal disk galaxies on a collision course such thatthey eventually reproduce the tidal features of a specific pair of merging galaxies. Finding theright initial conditions is time-consuming. The parameter set is large, and parameters interact incomplex ways. Moreover, hours may be required to run a single self-consistent calculation, andweeks of trial and error may be needed to obtain an acceptable match.

Fig. 1 summarizes the sixteen parameters needed to specify an encounter between two galax-ies. The radial and azimuthal coordinates, which respectively represent the initial orbit and diskorientations, together specify the initial conditions. The vertical coordinate specifies the time sincepericenter and viewing parameters.

The criteria used to match models to observations are a bit subtle and require discussion. Asa rule, it’s not enough to reproduce the tidal morphology of an interacting pair of galaxies, sincedi!erent encounter geometries may produce identical morphologies (e.g., Barnes 2011). Kinematic

1

Figure 1: An abstract representation of the sixteen-dimensional param-eter space of galaxy interactions. The radial coordinate represents theinitial orbit, specified by the pericentric separation p, mass ratio µ, andeccentricity e. The azimuthal coordinate represents the disc orienta-tions, specified by inclination angles i and azimuthal angles !. Thevertical coordinate represents the parameters chosen after a simulationis run, including the time since pericenter t, viewing angles "!, lengthscale L, velocity scale V, position zero-point r0 and velocity zero-pointv0. A standard N-body simulation explores the subspace represented bythe dotted line, while an IDENTIKIT simulation can explore the entirecylindrical surface.

information on tidal bridges and tails provides much stronger constraints. Interferometric HI data,which traces both tidal morphology and kinematics, is often used to constrain merger models;H# is also useful since tidal features may contain emission-line regions. Molecular and/or stellarabsorption lines can also provide constraints, but mapping extended structures in these lines isexpensive. Models and observations are typically compared visually, for example by overplottingparticles on various projections of a HI data cube (e.g., Hibbard & Mihos 1995). Since interstellarmaterial converts between molecular, atomic, and ionized phases by processes outside the scopeof purely dynamical models, visual inspection may be more reliable than quantitative measuresderived by di!erencing model and observational data cubes. However, visual inspection is tediousand inherently subjective; robust quantitative methods would certainly be welcome.

Several groups have adapted genetic algorithms to automate the process of modeling interactinggalaxies (e.g., Whade 1998, Theys & Kohle 2001, Smith et al. 2010). Some interesting results havebeen obtained, but most attempts to date have searched only a subset of the relevant parametersand made little use of kinematic information.

1.2 Galaxy models

To fully specify the initial conditions, the seven parameters characterizing the orbit and orientationsof the two disks must be supplemented by descriptions of the internal structures of the two galaxies.This presents many possibilities – disks may include separate thin, thick, and gaseous components;bulges and halos may follow a variety of profiles, and may be spherical, flattened, rotating, ortriaxial; non-axisymmetric structures such as bars or warps may be present, etc.

The most fundamental parameters are probably ratios of mass and scale length, (mb +md) :mh

and #!1d : ah, where the su"xes “b”, “d” and “h” refer to bulge, disk, and halo components, re-

spectively. These structural ratios are not arbitrary, since the galaxies involved in interactions werepresumably normal before their encounter. Nonetheless, the “generic” galaxy models sometimesused in initial conditions for merger simulations may be too limited for data-driven modeling.

Theories of disk galaxy formation suggest that galaxy-to-galaxy variations may be quite signif-icant. For example, Mo, Mao, & White (1998) present a model in which disk masses and angularmomenta are fixed fractions of the masses and angular momenta of their surrounding dark halos.In this scenario, galaxy-to-galaxy variations arise because tidal torques impart a wide range of an-gular momenta to halos (e.g., Warren et al. 1992); thus #!1

d :ah is a random quantity. Alternatively,Dekel et al. (2009) suggest that cold gas is transported into dark halos by streams associated withthe cosmic web. In this picture, the masses and angular momenta of disks may decouple from themasses and angular momenta of their dark halos.

2

Figure 2: Screen view fromIDENTIKIT showing a match(colored points) to a simu-lated encounter (grey-scale im-ages). The top-left, top-right, and bottom-left pan-els show (X,Y ), (VZ , Y ), and(X,VZ) projections of the sys-tem. These projections wouldbe available to an observerwith an HI data-cube. Thematch was obtained by inter-actively adjusting simulationand viewing parameters un-til the points reproduced themorphology and kinematics ofthe images. The bottom-rightpanel shows a (X,Z) projec-tion as well as various param-eter values.

Tidal interactions o!er a way to probe the halos of disk galaxies. Dubinski, Mihos, & Hernquist(1996) showed that the deep potential wells associated with massive quasi-isothermal halos inhibitthe formation of tidal tails. Further studies using more realistic halo models generally suggest thatthe ratio of circular velocity to local escape speed is the key parameter governing tail formation(Springel & White 1999, Dubinski, Mihos, & Hernquist 1999). This ratio may vary from galaxy togalaxy, implying that the formation of extended tidal tails cannot be taken for granted.

1.3 IDENTIKIT 1

IDENTIKIT simulations combine test-particle and self-consistent techniques (Barnes & Hibbard2009). Each galaxy is modeled by an initially spherical configuration of massive particles withcumulative mass profile m(r), in which is embedded a spherical swarm of massless test particleson initially circular orbits. Two such models with mass ratio µ are launched towards each otheron an orbit with eccentricity e and pericentric separation p. During the ensuing encounter, themassive components interact self-consistently, closely approximating the time-dependent potentialand orbit decay of a fully self-consistent galactic collision. The test particles mimic the tidalresponse of embedded discs with all possible spin vectors; once such a simulation has been run,selecting the appropriate subset of test particles yields a good approximation to the tidal responseof any particular disc.

Fig. 2 shows a test of this technique. Here, an initially parabolic encounter between two equal-mass disk galaxies was self-consistently simulated to generate artificial data (grey-scale images).The encounter geometry, pericentric separation, time since pericenter, viewing direction, and scalefactors were chosen at random. With no knowledge of these parameters, IDENTIKIT 1 was usedto find a match to the morphology (upper left) and kinematics (upper right and lower left) of

3

the self-consistent simulation. The software instantaneously displayed the results from any specificparameter choice. Instant feedback makes trial and error relatively e"cient; it took only a fewhours to find the match shown here. All unknown parameters were accurately recovered.

Barnes & Hibbard (2009) applied this technique to an ensemble of 36 random mergers, andsuccessfully reconstructed 30 cases. In the successful cases, the unknown parameters were well-constrained; for example, encounter geometry and viewing direction were recovered with medianerrors of < 15".

1.4 IDENTIKIT 2

IDENTIKIT 2 uses the self-consistent plus test-particle simulations described above to solve forthe initial orientation of each disk. This o!ers a significant shortcut compared to the trial anderror technique used with IDENTIKIT 1. The main assumption required is that the tidal featuresassociated with each galaxy can be traced back to a single disk with a unique orientation.

Fig. 3 shows how IDENTIKIT 2 works. Consider a single region of phase space with finiteextent in X, Y , and VZ and infinite extent in the remaining dimensions. This region, hereaftercalled a “box”, is placed so as to sample the tidal material from a particular galaxy. At some timepost-encounter, a single pass through the test-particle array for that galaxy selects all particlesfalling within the box. Each particle has been labeled with its initial spin axis relative to its parentgalaxy, so the selected particles define a density distribution on the unit sphere of all possible spindirections. As a rule, this distribution is extended and does not define a unique orientation forthe parent disk. However, another box sampling tidal material from the same disk will generate adi!erent distribution, and the two distributions will overlap at the disk’s true orientation.

Given boxes tracing tidal features from both galaxies, IDENTIKIT 2 can solve for the incli-nations and azimuths of both disks, directly determining four of the sixteen parameters in Fig. 1.The remaining twelve, however, must be specified beforehand, and if the specified values are wrongthen the derived disk orientations will probably be wrong as well. But if at least three boxes –ideally more – are used to sample a disk, the e!ects of parameter mismatch are likely to destroythe mutual overlap of the distributions shown in Fig. 3. By quantifying how well the distributionsoverlap – for example, by forming the product of the densities they trace – a relative figure of meritfor di!erent solutions is obtained.

With a fairly robust method of measuring quality of fit, it’s possible to implement automaticsearching over some subset of the twelve parameters besides disk orientation. Barnes (2011) de-scribed an implementation designed for systems composed of two disk galaxies which have not yetmerged. In this version, just six parameters – the initial orbit (p, µ, e), time since pericenter (t), andvelocity scale and zero-point (V, v0) – must be specified ahead of time. In addition to a set of boxestracing the tidal features of each galaxy, the algorithm requires coordinates on the plane of the skyfor both galaxies. It performs a blind search over all possible viewing directions ("X , "Y ); for eachdirection, the length scale (L), rotation about the line of sight ("Z), and position zero-point (r0)are determined by requiring the centers of the model galaxies to match the observed positions. Fitsto both disks are scored independently as described above; the viewing direction which maximizesthe product of both scores is selected. In tests with a small ensemble of simulated random mergers,the algorithm reconstructed encounter geometry and viewing direction with median errors of < 8".

1.5 Results from prior support

Joshua Barnes is a Co-I on NSF grant 1010064, which awarded $277,336 to Lisa J. Kewley for a36-month period starting 07/10/2011 to support “A Comprehensive Study of Chemical Evolution

4

Figure 3: A proof-of-concept test of the IDENTIKIT 2 algorithm. Left: (X,Y ) and (X,VZ)projections of an encounter between two disc galaxies with spins (i1,!1) = (30, 0) and (i2,!2) =(135, 0); the system is viewed along the orbital axis. The small colored boxes show three phase-space regions distributed along the tidal tail produced by galaxy 1. Right: equal-area projection ofone hemisphere of the spin sphere; lines of constant inclination i and azimuth ! are labeled. Thecolored contour plots show which disk spins can populate each corresponding box; the functionrepresented is a smoothed density on the spin sphere of particles falling within the box. The grey-scale image on the right represents the product of these distributions; the single white contour isset at 95% of the peak value, and neatly encloses the actual disc spin (i1,!1) = (30, 0).

in Galaxy Mergers”. Initial results demonstrate flattening of radial metallicity gradients in mergers(Kewley et al. 2010). Numerical simulations show that this flattening is due to merger-inducedgas inflows as well as outward transport of previously enriched material in tidal tails (Rupke etal. 2010). Further simulations show that nuclear metallicities are partly restored by chemicalenrichment after the first inflow phase and again after final coalescence, but typically remain belowtheir pre-encounter values (Torrey et al. 2011). Integral field spectroscopy shows that many late-stage mergers exhibit strong shock emission due to galactic winds which transport enriched gasoutward (Rich et al. 2010).

This proposal, which focuses on modeling the dynamics of observed mergers, is distinct fromthe work of Kewley et al., but there are several synergistic aspects. Some of the galaxies observedby Kewley et al. are potential targets for dynamical modeling, and the spectroscopic data gatheredby Kewley et al. may provide useful information on merger kinematics. Detailed models of thesegalaxies will help to place their observed properties in a broader framework.

5

Figure 4: NGC 2623. Top-left and bottom-right panelsshow HI and optical images,respectively; North is up andWest is right. Top-right andbottom-left panels show HIposition-velocity diagrams; ve-locity increases to left anddown, respectively. Grey-scaleshows emission, while contoursshow HI absorption. Red andgreen points show a prelim-inary IDENTIKIT 1 model(Privon et al., in preparation).

2 Research Objectives

2.1 Modeling real galaxies

A key goal of this proposal is to match the morphology and kinematics of real interacting galaxies.This immediately raises several new issues. First is the limitations of the observational data. Formost systems, HI is the only available tracer of large-scale morphology and kinematics. In manycases, the resolution and signal-to-noise of the HI data are barely su"cient to trace tidal structures.Second, the observational data may display features outside the scope of simple test-particle models.For example, HI data often includes absorption features, and it’s not clear if or how such featurescan be matched by the simulations. Third, the mass models used in the IDENTIKIT simulationsmay not be good approximations to the actual mass distributions of the galaxies involved in anencounter. At this point, the simulations use generic disk galaxy models; tailoring these models tospecific systems is an important challenge.

Fig. 4 presents a preliminary model of NGC 2623 (Privon et al., in preparation) which illustratesall three of these considerations. First, while HI is clearly detected in the tails, its emission ismapped by taking the maximum voxel value along each line of sight through the data cube; summingthe emission produces noise dominated-maps. It’s not entirely clear how to compare these maps tothe the projected particle distribution, which most naturally corresponds to summed emission. Inaddition, HI emission is not detected from the bright star-forming region to the south of the mainbody of NGC 2623; optical spectroscopy may be needed to determine kinematics in this region.Second, the system exhibits HI absorption, presumably due to neutral hydrogen silhouetted againstNGC 2623’s AGN (Evans et al. 2008). It’s interesting that the velocity width of this absorptionfeature is fairly well reproduced by the width of the particle distribution, but not entirely obvious

6

that this represents a success of the model since HI on the far side of the nucleus does not contributeto the absorption. Third, while NGC 2623 probably results from a merger of comparable galaxies,the hook-shaped tail to the West might be better reproduced if its parent galaxy’s disk was initiallylarger than its partner’s.

Further experience modeling a number of di!erent systems is needed to explore these issues. Itwould also be interesting to combine HI with optical data, and to include other velocity tracers suchas H# or CO. We are currently modeling Arp 240 (NGC 5257/8); other systems on our short-listinclude NGC 34 and NGC 3256. Eventually we hope to model a number of galaxies in the GOALSsample of luminous infrared galaxies (Armus et al. 2009).

To support this modeling e!ort, the existing ensemble of IDENTIKIT simulations (Barnes &Hibbard 2009) will be expanded. Currently, we have good coverage for e = 1 (parabolic) encountersof galaxy pairs with mass ratios µ = 1 and 0.5. We will expand this ensemble to include non-parabolic orbits as well as mass ratios µ = 0.75, 0.333, and 0.25. Eventually, we will also explore avariety of galaxy models as discussed in § 2.4.

2.1.1 Morphological matching

While this proposal focuses on matching both morphology and kinematics, only a modest fractionof interacting galaxies have spatially-resolved kinematic data. It’s therefore worth asking what canbe learned by matching morphology alone. A number of studies have broadly classified mergers byinteraction stage (e.g., Evans et al. 2011), using morphological criteria such as presence of tidaltails, common envelope structure, distinct nuclei, etc. Morphological matching may provide a morerigorous way to define successive interaction stages; it can also yield some rough constraints onmass ratio and encounter geometry. Finally, morphological matching can serve as a “gateway” tomore detailed modeling; a purely morphological match is unlikely to match the kinematics of tidalfeatures, but it identifies one member of an equivalence class which includes models which do matchthe kinematics.

2.2 Automatic reconstruction

The IDENTIKIT 2 algorithm (Barnes 2011) points toward a way to automate the matching processand suggests several further topics for investigation. One objective involves the constraints used toidentify good matches; adding more constraints generally improves the quality of the solutions andallows the algorithm to search over more parameters. Recently, we have implemented constraintson the systemic velocities of individual galaxies, increasing the speed and accuracy of the algorithm.We plan to implement “no-fly” boxes which the algorithm must avoid populating; this capabilityshould help in looking for solutions defined by narrow or kinematically cold tidal features.

A second objective is to extract more information from the matching process. The algorithmsolves for disk orientation by finding the highest peak in the product of the spin distributions(grey-scale on right in Fig. 3), and assigns an overall score using the peak value. In some cases theproduct function may have secondary peaks, which represent alternate solutions, or the functionmay peak along a ridge, implying that disk angles i and ! are correlated. This information may beuseful to refine the estimated quality of the derived solution. Likewise, the algorithm determinesviewing direction by selecting the direction which maximizes the product of the scores for the twodisks, but there is a good deal more information available (e.g., Barnes 2011, Fig. 6). For example,the algorithm can determine of both disks favor the same solution for viewing direction, if one diskprovides stronger constraints than the other, or if the selected direction is a compromise betweenthe two disks.

7

Figure 5: NGC 4676. Top-leftand bottom-right panels showHI and optical images, respec-tively; North is up and West isright. Top-right and bottom-left panels show HI position-velocity diagrams; velocity in-creases to left and down, re-spectively. Boxes show con-straints used to determine diskorientation. Red and greenpoints show IDENTIKIT 2model.

A third objective is to increase the speed of the algorithm. The code is already more than anorder of magnitude faster than the first version released in 2011. It can be made even faster withoutchanging the algorithm’s behavior or accuracy. Pre-computing the mapping from particle index toinitial spin sphere, which only needs to be done once for a given IDENTIKIT simulation, couldreplace one of the algorithm’s innermost loops with a simple array access. Finally, the algorithm canbe implemented in parallel, and much of the computing could be done using a Graphical ProcessorUnit (GPU). The computer workstation requested in the budget includes a powerful GPU for thispurpose. Faster versions of the code will make searching large parameter sets feasible without amajor investment in computer power.

Fig. 5 shows that the present algorithm already has interesting real-world applications. Here,four regions are allocated to each tail of “The Mice”, NGC 4676. These regions track the HImorphology of each tail (top-left panel) as well as its line-of-sight velocity (top-right and bottom-left). In addition, the solution was constrained by requiring the simulated nuclei (yellow crosses)to match the observed positions, and fall within the observed range of systemic velocities. Sixparameters were specified a priori: the orbital eccentricity (e = 1), mass ratio (µ = 1), pericentricseparation (! 4.5 disk scale lengths), time since pericenter (! 1.25 disk rotation periods), velocityscaling and velocity zero-point are close to the values adopted in earlier models (e.g., Barnes 2004).However, the remaining ten parameters were all derived by the algorithm, which required about twominutes on a 2.16GHz processor to examine 5120 viewing directions and produce the model shownhere. This model is about as good as models of NGC 4676 derived by hand. Other “well-separated”systems which can be modeled in the same way include Arp 256, Arp 295, and Arp 298; HI data isavailable for all three.

8

Figure 6: Fourteen-parameter search for models of NGC 4676. From one panel to the next, timesince pericenter t increases left to right, and pericentric separation p increases bottom to top. Eachpanel shows a 9 " 8 grid of models in V and v0. Dot size indicates model score. The 122 highest-scoring solutions are visually classified as poor (blue; N = 30), fair (black N = 56), and good (redN = 36); low-scoring solutions are shown in grey.

2.3 Uniqueness and uncertainty

IDENTIKIT 2 is fast and robust enough to go beyond the ten-parameter search described above.For a system like NGC 4676, it’s reasonable to fix the mass ratio µ = 1, adopt an e = 1 (parabolic)initial orbit, and perform a blind search in pericentric separation p, time since pericenter t, velocityscale V, and velocity zero-point v0, as well as the angles "X and "Y defining direction of view. As thenumber of parameters to be determined increases, the algorithm yields solutions with comparableoverall scores for many parameter combinations. In Fig. 6 solutions have been visually graded aspoor (blue), fair (black), or good (red) representations of NGC 4676. It’s clear that numerical scoreis an imperfect indicator of quality, since the highest-scoring solutions (in the t = 0.75, p = 0.25panel) are classified as fair, but many good solutions do get high scores, while poor solutionsconsistently get low scores.

The larger dots in Fig. 6 are not scattered at random; most solutions fall along a diagonalfrom lower left to upper right across the panels, indicating a general correlation between timesince pericenter t and pericentric separation p. Other parameters which correlate with t includethe velocity zero-point v0, viewing angles "X and "Y , and length and velocity scale factors L andV. These correlations indicate that acceptable solutions populate an “error ellipse” in parameterspace. As a group, these solutions are relatively homogeneous; the acceptable solutions all appear tooriginate from a single connected region. Moreover, the physical time since pericenter, T = (L/V)t,is very well constrained; all of the good and almost all of the fair solutions yield T values between150 and 200Myr.

9

Although it would seem preferable to obtain a unique model, the ensemble of solutions forNGC 4676 shown in Fig. 6 is probably a good representation of the actual uncertainties inher-ent in modeling this system. It’s unknown if most interacting galaxies will likewise yield fairlywell-constrained solutions; NGC 4676 is a relatively simple system, and others may be harder toconstrain. Imposing more constraints appears to reduce the range of acceptable solutions; this maybe the real payo! for implementing new kinds of constraints as discussed in § 2.2.

2.4 Mass models

The results presented above use a single galaxy model with a structural mass ratio (mb+md) :mh =1 : 4. While this model is adequate for some purposes, the halo comprises just 80% of the total,which is less than expected in CDM cosmologies. Moreover, any e!ort at data-driven modeling ofreal mergers must address variations in initial galaxy structure. How sensitive are model results toinevitable discrepancies between the adopted and actual structure of the progenitor galaxies? Cantidal encounter models accurately probe the overall depth and structure of halo potential wells?

To address these questions, we are developing galaxy models with structural ratios (mb + md) :mh = 1 : 9 and 1 : 19, implying halo masses more in line with the predictions of #CDM. For eachvalue of (mb + md) :mh, we will adopt three di!erent choices for the scale ratio #!1

d :ah based onplausible variations in spin parameter $ (e.g., Mo, Mao, & White 1998). To keep the total numberof simulations manageable, we will focus on encounters between galaxies with similar halo-to-totalmass ratios. First, we will run a grid of simulations without test particles to characterize orbitaltrajectories and merger timescales as functions of (a) structural ratio (mb + md) : mh, (b) scaleratio #!1

d :ah, (c) mass ratio µ, (d) initial orbital eccentricity e, and (e) pericentric separation p.Existing work on the e!ects of halo structure on tidal features (Dubinski, Mihos, & Hernquist

1996, 1999; Springel & White 1999) has largely focused on direct, co-planar encounters whichmaximize tidal features; a systematic study varying encounter geometry as well as galaxy structurewould be bedeviled by a large number of parameters. However, IDENTIKIT o!ers an e"cient wayto fold encounter geometry into the problem, since a single simulation can simultaneously modelall possible disk orientations. Fig. 7 presents an example, based on a model with a mass ratio(mb +md) :mh = 1:9. The plot shows test particles which have attained a maximum distance fromtheir parent galaxy of at least three times their initial orbital radius; this reliably identifies particlesin tidal features. No selection for initial disk orientation is done, so the result is a “synoptic” viewof all tidal structures resulting from this encounter. Each selected particle is classified as bridge,tail, or “other” on the basis of its position relative to the two galaxies at first passage and at itsinstant of maximum distance.

The histogram on the right of Fig. 7 shows how the mass fractions µtid in bridges and tailsdepend on disk inclination i. At small inclinations (i < 30"), the bridge is relatively massive,comprising ! 8% of the disk material. However, bridge mass drops rapidly with increasing i. Thetail is less massive, amounting to ! 2.5% of the disk material for small i, but drops o! less rapidlywith increasing i. In e!ect, bridge-to-tail mass ratio is a decreasing function of inclination; thistrend has been noted before but not quantified. We will extend this analysis to other galaxy models,systemically exploring how the mass in tidal features depends on the parameters (a)—(e) listedabove. This will also create a large library of IDENTIKIT simulations with various mass modelswhich can be used in other modeling e!orts.

Following Barnes & Hibbard (2009), we will generate an ensemble of artificial data based on ran-dom mergers using di!erent galaxy models, mass ratios, orbital parameters, encounter geometries,and viewing parameters. These mergers will be fit with IDENTIKIT 2, comparing results obtainedusing “correct” and mismatched mass models. Two topics to be addressed are (a) how well fitting

10

Figure 7: Left: result of a moderately close (p = 3.2#!1d ) passage between two equal-mass galaxies,

seen shortly after apocenter. Solid lines are galaxy trajectories; the cross marks the system’s centerof mass. Points show tidal structures originating from the galaxy marked by the larger filled circle,integrated over all possible disk orientations. Color indicates type of tidal feature: red for tail,blue for bridge, and green for “other” particles. Right: histogram showing mass fractions in tidalstructures as functions of disk inclination i, using the same color coding.

to interacting systems can constrain galaxy mass models, and (b) how sensitive other parametersand derived quantities are to discrepancies between the actual and assumed mass models. Finally,we will apply these results to the task of modeling real interacting systems, attempting to constrainthe parameters of their dark halos.

3 Broader Impact

3.1 Benefits to astronomy

In addition to its capacity to address the research topics in § 2, IDENTIKIT is a resource forother astronomers interested in modeling interacting galaxies. Since November 2008, a simple webIDENTIKIT 1 interface has been available (Fig. 8). To date, this page has been accessed over 7400times; these requests have originated from over 300 di!erent domestic and foreign IP addresses,and 28 di!erent IP addresses have each generated 50 or more requests. This shows that the presentversion already has some utility for astronomical research.

The work described in this proposal will significantly increase the utility of the web interface. Tobegin with, the set of IDENTIKIT simulations accessed through the interface will be expanded toinclude non-parabolic orbits, a wider range of mass ratios, and a reasonable variety of galaxy modelsas described in § 2.1 and 2.4. Additional IDENTIKIT 1 capabilities, such as overlay of observationaldata, can be implemented in a straightforward manner. It’s also possible to implement a client-sideJava applet which can handle rotations, translations, and scale transformations locally; this would

11

Figure 8: Web interface to IDENTIKIT 1. The four panels of the display show the same projectionsseen in Fig. 2. The user can select from a menu of initial orbits and enter values for other parametersin the boxes provided. Clicking on Update displys the result. (NB: to improve particle visibilitywhen printed, color values within the display have been inverted.)

make the interface faster and much more interactive. Finally, the core functions of IDENTIKIT 2can also be provided; while full parameter searching as described in § 2.2 and 2.3 may put excessivedemands on the server, the ability to define a set of boxes and derive solutions for disk orientationshould be fairly easy to implement.

Source code for IDENTIKIT and the associated software necessary to generate the requiredsimulation files has also been available on the web since November 2008, and a few researchershave downloaded the code and set up working systems. Several barriers have been identified whichdiscouraged wider adoption. The source code as originally released depended on an older version ofNumerical Recipes in C ; a new version using the free Gnu Scientific Library will soon be available.Moreover, to generate a library of IDENTIKIT simulations, potential users had to run a seriesof calculations requiring days to weeks of computer time. As described in the attached DataManagement Plan, simulation files will be made directly available via the web. These steps shouldmake IDENTIKIT much more accessible to interested users.

3.2 Education and outreach

Galaxy collisions produce compelling images which attract great public interest. While many ofthe issues now under active investigation are rather esoteric, the basic concept of gravitationalencounters between galaxies has strong popular appeal. The IDENTIKIT software is fast enoughto support interactive web-based applications which can be used in outreach to high-school classes.

12

In Hawai‘i, the HI STAR program uses astronomical research to promote science, mathematics,engineering and technology (STEM) education to high-school students (grades 7–12). HI STARtargets students from the outer islands and rural districts, native Hawai‘ians, and members of othergroups with low socioeconomic status.

Within the framework provided by HI STAR, we plan to develop a classroom tutorial on galaxymerger modeling. Students will be introduced to the subject of galaxy mergers via observationalimages and animations derived from numerical simulations. Merger modeling can be described usinganalogies from popular culture – for example, the problem of reconstructing a galaxy collision canbe compared to the problem of reconstructing a car crash, where the state of the debris is “read”to infer a sequence of past events. After seeing some examples, students will try to match severalinteracting systems. The tutorial will focus on morphological matching, side-stepping the addedconceptual hurdles involved in kinematic matching; since even morphological matching can yieldpotentially useful results (§ 2.1.1), the students will be doing “real astronomy”. Matching can bedone in two complementary ways:

• Individual matching. Students access IDENTIKIT via the web (Fig. 8), working individu-ally or in small groups. By trial and error, they explore a subset of the parameter space andattempt to match the appearance of several real pairs of interacting galaxies. For the firstattempt, most parameters will be fixed at values previously determined to produce a goodmatch. As students make progress through a series of examples, additional parameters maybe brought into play.

• “Genetic” matching. IDENTIKIT will generate a population of ! 102 models with initiallyrandom parameters. Via a web-based display, each student will be presented with images of! 10 to 20 possible models, and asked to pick the image most closely matching the actualdata. The parameters of the selected models will be “recombined” by a genetic algorithm toproduce a new generation of models, and students will again be asked to select the best match.Over multiple iterations, the population will converge toward ever closer approximations tothe actual system.

Before these tutorials are introduced in high-school classrooms, they will be tested in an un-dergraduate astronomy lab course developed by the PI in 2002. If they work out, somewhat moreadvanced versions may also be incorporated into the undergraduate curriculum, as possible labexercises for cloudy nights.

3.3 Graduate education

This proposal includes support for an Astronomy PhD student at the University of Hawai‘i inManoa. This student will be directly involved in the research, participating in planning and con-ducting numerical simulations, extracting maps from HI data cubes and combining these withother sources of kinematic information, using IDENTIKIT to generate models of merging galaxies,and analyzing results. This will provide excellent training in a variety of techniques, and suitablematerial for a strong PhD dissertation.

The PI for this proposal is a dissertation committee member for George C. Privon, a PhDstudent supervised by Prof. Aaron S. Evans at the University of Virginia. Travel support requestedin this proposal will enable the PI to periodically visit the University of Virginia. These visits willsupport collaboration between the PI and Mr. Privon, as well as with Prof. Evans. Mr. Privon isan early adopter of IDENTIKIT 1 and will model a number of GOALS galaxies (Armus et al. 2009)as part of his dissertation.

13

4 Work Plan

Year 1. We will complete IDENTIKIT 1 models for NGC 2623 and Arp 240; these will be publishedwith G. Privon as first author (§ 2.1). “No-fly” constraints will be implemented in IDENTIKIT 2,and a parallel/GPU version will be developed (§ 2.2). The survey of NGC 4676 models will beextended to wider passages and later times to better delimit the full extent of the “good” solutions;results will be published (§ 2.3). A suite of mass models will be constructed, tested for stability, andused to survey orbital trajectories; existing simulation libraries for µ = 1 and 0.5 will be updated touse the new mass models (§ 2.4). The results will be made available in the on-line interface (§ 3.1).A simple version of the interface will be developed for instructional applications, and the mergermodeling tutorial will be “work-shopped” for undergraduates (§ 3.2). An IfA graduate student willstart work by mid-year (§ 3.3); depending on initial level of knowledge and ability, the student maybe given charge of morphological matching (§ 2.1.1).Year 2. Models of NGC 34 and NGC 3256 using IDENTIKIT 1 will be developed and published(§ 2.1). IDENTIKIT 2 will be modified to extract more information from the matching process;Arp 256, Arp 295, and Arp 298 will be modeled (§ 2.2), and the range of acceptable fits will bedetermined (§ 2.3). Random mergers for various mass models will be constructed and fit usingIDENTIKIT 2 to test sensitivity to halo structure (§ 2.4). A client-side applet will be developedfor the on-line interface, and core elements of IDENTIKIT 2 will be implemented on-line (§ 3.1).The merger modeling tutorial will debut in selected high-school classrooms (§ 3.2). As the yearbegins, the IfA graduate student will have formalized a dissertation proposal and will be fullyinvolved in some or all of the above activities.Year 3. IDENTIKIT 1 and 2 will be in full production mode. Options for extending IDENTIKIT 2to model fully-merged galaxies will be evaluated. Models for a range of systems in di!erent mergerstages will be studied to determine uniqueness, uncertainties, and strength of constraints on darkhalo structure. The on-line interface will o!er a full range of galaxy models, mass ratios, andorbital parameters; these models will be available for download. The merger modeling tutorial willbe finalized and packaged for use by Hawai‘i high-school teachers. The IfA graduate student willcontinue work and by year’s end will complete a draft of the dissertation.

14

References

Armus, L., Mazzarella, J. M., Evans, A. S., Surace, J. A., Sanders, D. B., Iwasawa, K., Frayer, D.T., Howell, J. H., Chan, B., Petric, A., Vavilkin, T., Kim, D. C., Haan, S., Inami, H., Murphy,E. J., Appleton, P. N., Barnes, J. E., Bothun, G., Bridge, C. R., Charmandaris, V., Jensen, J. B.,Kewley, L. J., Lord, S., Madore, B. F., Marshall, J. A., Melbourne, J. E., Rich, J., Satyapal, S.,Schulz, B., Spoon, H. W. W., Sturm, E., U, V., Veilleux, S., Xu, K. (2009), PASP, 121, 559.

Barnes, J. E. (1988), ApJ, 331, 699

Barnes, J. E. (2004), MNRAS, 350, 798

Barnes, J. E. (2011), MNRAS, 413, 2860

Barnes, J. E. & Hibbard, J. E. (2009), AJ 137, 3071

Dekel, A., Birnboim, Y., Engel, G., Freundlich, J., Goerdt, T., Mumcuoglu, M., Neistein, E.,Pichon, C., Teyssier, R., Zinger, E. (2009), Nature, 457, 451

Dubinski, J., Mihos, J. C., & Hernquist, L. (1996), ApJ, 462, 576

Dubinski, J., Mihos, J. C., & Hernquist, L. (1999), ApJ, 526, 607

Evans, A. S., Vavilkin, T., Pizagno, J., Modica, F., Mazzarella, J. M., Iwasawa, K., Howell, J. H.,Surace, J. A., Armus, L., Petric, A. O., Spoon, H. W. W., Barnes, J. E., Suer, T. A., Sanders, D.B., Chan, B., Lord, S. (2008), ApJ, 675, L69

Evans et al. (2011), submitted

Hibbard, J. E. & Mihos, J. C. (1995), AJ, 110, 140

Kewley, L. J., Rupke, D., Zahid, H. J., Geller, M. J., Barton, E. J. (2010), ApJ, 721, L48

Mo, J. H., Mao, S., & White, S. D. M. (1998), MNRAS, 295, 319

Rich, J. A., Kewley, L. J., Dopita, M. A. (2011), ApJ, 734, 87.

Rupke, D. S. N., Kewley, L. J., Chien, L.-H. (2010), ApJ, 723, 1255

Sanders, D. B. & Mirabel, I. F. (1996), ARA&A, 34, 749

Smith, B. J., Carver, D. C., Pfei!er, P., Perkins, S., Barkanic, J., Fritts, S., Southerland, D.,Manchikalapudi, D., Baker, M., Luckey, J., Franklin, C., Mo!ett, A., Struck, C. (2010), ASPConference Series, 423, 227

Springel, V. & White, S. D. M. (1999), MNRAS, 307, 162

Theys, C. & Kohle, S. (2001) A&A., 370, 365

Toomre, A. (1977), “Evolution of Galaxies and Stellar Populations”, ed. B. M. Tinsley & R. B.Larson, Yale University, p. 401

Toomre, A. & Toomre J. (1972), ApJ, 178, 623

Torrey, P., Cox, T. J., Kewley, L., Hernquist, L. (2011), ApJ, in press

Warren, M. S., Quinn, P. J., Salmon, J. K., Zurek, W. H. (1992), ApJ, 399, 405

Whade, M. (1998) A&AS, 132, 417

White, S. D. M. (1978), MNRAS, 184, 185

JOSHUA EDWARD BARNES: BIOGRAPHICAL SKETCH

Professional Preparation

Harvard University, Cambridge Astronomy & Astrophysics B.A., 1979University of California, Berkeley Astronomy PhD., 1984Institute for Advanced Study, Princeton Astrophysics 1984 – 1989

Appointments

Astronomer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8/2003 – NowInstitute for Astronomy, University of Hawaii, Honolulu, HI.Graduate program chair: 1/2005 – 8/2007, 8/2008 – 6/2010.

Associate Astronomer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1/1991 – 7/2003Institute for Astronomy, University of Hawaii, Honolulu, HI.Tenure granted: 7/1995. Appointed associate chair: 1/2002.

Senior Research Associate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9/1989 – 12/1990Canadian Institute for Theoretical Astrophysics, Toronto, Canada.

Proposal-Related Publications

Barnes, J.E., “Identikit 2: an algorithm for reconstructing galaxy collisions”, Monthly Notices ofthe Royal Astronomical Society 413, 2860–2872 (2011).Barnes, J.E. & Hibbard, J.E., “Identikit 1: A Modeling Tool for Interacting Disk Galaxies”,Astronomical Journal 137, 3071–3090 (2009).Barnes, J.E., “Shock-induced star formation in a model of the Mice”, Monthly Notices of theRoyal Astronomical Society 350, 798–808 (2004).Barnes, J.E. & Hernquist, L., “Transformations of Galaxies. II. Gasdynamics in Merging DiskGalaxies”, Astrophysical Journal 471, 115–142 (1996).Barnes, J.E. & Hernquist, L., “Fueling starburst galaxies with gas-rich mergers”, AstrophysicalJournal Letters 370, L65–L68 (1991).

Other Publications

Barnes, J.E. & Hernquist, L., “Dynamics of interacting galaxies”, Annual Review of Astronomyand Astrophysics 30, 705–742 (1992).Barnes, J.E., “Transformations of Galaxies. I. Mergers of Equal-Mass Stellar Disks”, AstrophysicalJournal 393, 484–507 (1992).Barnes, J.E., “Encounters of Disk/Halo Galaxies”, Astrophysical Journal 331, 699–717 (1988).Barnes, J. & Hut, P., “A Hierarchical O(N log N) Force Calculation Algorithm”, Nature 324,446–449 (1986).

Synergistic Activities

Software: Developed hierarchical “Tree Code” for gravitational N -body simulations. Wrote firstimplementation, proved O(N log N) scaling, made code publicly available (1986). Developed com-prehensive ZENO N-body simulation environment, made code publicly available (2008).Invited lecturer: 5th Canary Islands Winter School of Astrophysics on “The Formation andEvolution of Galaxies” (1993); 26th Saas-Fee Advanced Course on “Galaxies: Interactions andInduced Star Formation” (1996). 2nd Course of the International School on Astrophysical Relativityon “Frontiers in Numerical Gravitational Astrophysics” (2008).Conference organization: Aspen Summer Workshop on “Galaxy Interactions at Low and HighRedshift” — chaired SOC (1996); IAU Symposium 186 on “Galaxy Interactions at Low and HighRedshift” — member of SOC, edited conference proceedings (1997).Course development: Proposed and launched highly successful undergraduate “Astronomy Lab-oratory” at University of Hawai‘i Manoa (2002). Proposed undergraduate “Foundations of Astro-physics” course sequence at University of Hawai‘i Manoa as first stage in implementing undergrad-uate astrophysics major (2011).

Collaborators and Other A!liations

Collaborators: Appleton, P.N. (Herschel); Armus, L. (Spitzer); Baker, A.J. (Rutgers); Bothun,G. (U. Oregon); Bridge, C.R. (CALTECH); Chan, B. (IPAC); Charmandaris, V. (U. Crete); Chien,L.H. (STScI); Contursi, A. (MPE Garching); Evans, A.S. (U. Virginia); Fernandez, X. (Columbia);Frayer, D.T. (Herschel); Genzel, R. (MPE Garching); Haan, S. (Spitzer); Howell, J.H. (Spitzer); In-ami, H. (Spitzer); Iwasawa, K. (Universitat de Barcelona); Johnston, K. (Columbia); Joseph, R.D.(IfA); Kewley, L.J. (IfA); Kim, D.C. (NRAO); Koda, J. (SUNY); Lord, S. (IPAC); Lutz, D. (MPEGarching); Madore, B.F. (IPAC); Marshall, J.A. (Spitzer); Matsuhara, H. (ISAS); Mazzarella, J.M.(IPAC); Melbourne, J.E. (CALTECH); Mihos, J.C. (Case Western); Murphy, E.J. (Spitzer); Netzer,H. (Tel-Aviv); Petric, A. (Spitzer); Privon, G.C. (U. Virginia); Rich, J. (IfA); Rupke, D.S.N. (U.Maryland); Sanders, D.B. (IfA); Schulz, B. (Herschel); Schweitzer, M. (MPE Garching); Schweizer,F. (Carnegie); Spoon, H.W.W. (Cornell); Sternberg, A. (Tel-Aviv); Stierwalt, S. (Spitzer); Stock-ton, A. (IfA); Sturm, E. (MPE Garching); Surace, J.A. (Spitzer); Tacconi, L.J. (MPE Garching);Teyssier, M. (Columbia); U, V. (IfA); van Gorkom, J.H. (Columbia); Vavilkin, T. (Stony Brook);Veilleux, S. (U. Maryland); Xu, K. (Herschel)Advisors: Davis, Marc (Berkeley); Hut, Piet (IAS); White, Simon D.M. (Max Planck).Advisees: Fulton, Eliza, MS 1998 (Baylor); Ishida, Catherine, PhD 2004 (Subaru) [with D.Sanders]; Chien, Li-Hsin, PhD 2009 (STScI). Total: 3.Postdoctoral sponsorship: Hibbard, John, 1994–7 (NRAO) [with D. Sanders]. Total: 1.