COVER SHEET FOR PROPOSAL TO THE NATIONAL SCIENCE FOUNDATION

Transcription

1 COVER SHEET FOR PROPOSAL TO THE NATIONAL SCIENCE FOUNDATION PROGRAM ANNOUNCEMENT/SOLICITATION NO./CLOSING DATE/if not in response to a program announcement/solicitation enter NSF NSF FOR CONSIDERATION BY NSF ORGANIZATION UNIT(S) (Indicate the most specific unit known, i.e. program, division, etc.) FOR NSF USE ONLY NSF PROPOSAL NUMBER PHY - ASTROPHYSICS & COSMOLOGY THEOR DATE RECEIVED NUMBER OF COPIES DIVISION ASSIGNED FUND CODE DUNS# (Data Universal Numbering System) FILE LOCATION EMPLOYER IDENTIFICATION NUMBER (EIN) OR TAXPAYER IDENTIFICATION NUMBER (TIN) NAME OF ORGANIZATION TO WHICH AWARD SHOULD BE MADE University of Maryland College Park AWARDEE ORGANIZATION CODE (IF KNOWN) NAME OF PERFORMING ORGANIZATION, IF DIFFERENT FROM ABOVE SHOW PREVIOUS AWARD NO. IF THIS IS A RENEWAL AN ACCOMPLISHMENT-BASED RENEWAL IS THIS PROPOSAL BEING SUBMITTED TO ANOTHER FEDERAL AGENCY? YES NO IF YES, LIST ACRONYM(S) DoE ADDRESS OF AWARDEE ORGANIZATION, INCLUDING 9 DIGIT ZIP CODE University of Maryland College Park 3112 Lee Building College Park, MD ADDRESS OF PERFORMING ORGANIZATION, IF DIFFERENT, INCLUDING 9 DIGIT ZIP CODE PERFORMING 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 TITLE OF PROPOSED PROJECT HAWC - A Wide-Field TeV Gamma-Ray Observatory REQUESTED AMOUNT PROPOSED DURATION (1-60 MONTHS) REQUESTED STARTING DATE SHOW RELATED PRELIMINARY PROPOSAL NO. $ 6,043, months 05/01/07 IF APPLICABLE CHECK APPROPRIATE BOX(ES) IF THIS PROPOSAL INCLUDES ANY OF THE ITEMS LISTED BELOW BEGINNING INVESTIGATOR (GPG I.A) DISCLOSURE OF LOBBYING ACTIVITIES (GPG II.C) HUMAN SUBJECTS (GPG II.D.6) Exemption Subsection or IRB App. Date PROPRIETARY & PRIVILEGED INFORMATION (GPG I.B, II.C.1.d) HISTORIC PLACES (GPG II.C.2.j) INTERNATIONAL COOPERATIVE ACTIVITIES: COUNTRY/COUNTRIES INVOLVED (GPG II.C.2.j) SMALL GRANT FOR EXPLOR. RESEARCH (SGER) (GPG II.D.1) VERTEBRATE ANIMALS (GPG II.D.5) IACUC App. Date HIGH RESOLUTION GRAPHICS/OTHER GRAPHICS WHERE EXACT COLOR REPRESENTATION IS REQUIRED FOR PROPER INTERPRETATION (GPG I.G.1) PI/PD DEPARTMENT Department of Physics PI/PD FAX NUMBER PI/PD POSTAL ADDRESS 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 Department of Physics University of Maryland College Park, MD United States Jordan A Goodman PhD goodman@umdgrb.umd.edu Brenda L Dingus PhD dingus@lanl.gov David B Kieda PhD kieda@physics.utah.edu James T Linnemann PhD linnemann@pa.msu.edu Constantine Sinnis Ph.D Gus@lanl.gov Page 1 of 2

2 Project Summary Scientific Justification Very-high-energy gamma-ray astrophysics studies the most extreme environments in the universe: regions of space under the influence of extreme gravitational and magnetic fields. There are two types of instruments that can measure the non-thermal radiation in the energy range above 0 GeV: imaging atmospheric Cherenkov telescopes (IACTs) such as HESS and VERITAS and extensive air shower (EAS) arrays such as Milagro. Recent data from Milagro demonstrates the important role of a wide-field instrument. There is a need for a next generation, all-sky, TeV gamma-ray observatory to be operational during the lifetime of the GLAST observatory, the VERITAS/HESS IACT arrays, and the IceCube neutrino observatory. We propose to build the HAWC (High Altitude Water Cherenkov) observatory, a next generation EAS array based upon technology that has been developed and proven with the Milagro detector. The HAWC observatory combines the Milagro water Cherenkov technology with a very high altitude site. Re-deploying the existing Milagro photomultiplier tubes (PMTs) and electronics in a different configuration at an altitude above 4000m will lead to a sensitivity increase of a factor of ~15 over Milagro. This dramatic improvement is due to three things: the increased altitude, the increased physical area, and the optical isolation of the PMTs. As a result of these improvements, HAWC will detect a 5σ signal from the Crab Nebula in a single 4-hr transit (compared to ~5 months for Milagro). HAWC will maintain this sensitivity over most of the northern sky. This sensitivity will enable very high energy gamma ray studies that are unattainable with the current suite of instruments. 1) HAWC will monitor (for >4 hours every day), every point in almost 2π sr of the sky. Over a 5 year observation period HAWC will perform an unbiased sky survey with a detection threshold of ~20 mcrab, enabling the monitoring of known sources, the discovery of new sources of known types, and perhaps most importantly the discovery of new classes of TeV gamma ray sources. 2) The sensitivity of HAWC to extended sources surpasses that of IACTs for sources larger than 0.25 o. HESS measurements clearly point to the existence of such objects within our Galaxy; since they observe many diffuse galactic sources clustered around their sensitivity limit, they may only be seeing the tip of the iceberg. 3) With HAWC s unsurpassed sensitivity at high energies, we will answer many questions about the origin and propagation of cosmic rays. 4) With the sensitivity to detect a flux of 5 times that of the Crab in just minutes over the entire overhead sky, HAWC will observe many flares from AGN. While an important measurement in its own right, this will also enable many multi-wavelength observations (in particular with X-ray telescopes) of these flares. 5) HAWC s sensitivity to the prompt emission from gamma-ray bursts is unique. With HAWC s low energy threshold, GRBs with a TeV fluence comparable to their kev fluence will be detectable to a redshift of ~1, while for closer GRBs much lower fluences can be detected. HAWC s sensitivity to transient phenomena will extend the field of time-domain astrophysics to TeV energies. Broader Impacts Because HAWC is an all-sky instrument it will serve as a TeV finder telescope for IACTs and IceCube. Beyond astrophysics HAWC will establish a new international scientific collaboration, with either Mexico or China. HAWC will provide strong scientific and educational opportunities for undergraduate students, graduate students, post-doctoral fellows, and underrepresented minorities. In addition, HAWC scientists will continue their extensive work bringing the excitement of particle astrophysics to high school students and the general public in the US and in the host country. Most importantly, HAWC will provide a training ground for the next generation of scientists in techniques of particle astrophysics, high energy physics and the mining of large datasets.

3 TABLE OF CONTENTS For font size and page formatting specifications, see GPG section II.C. Total No. of Pages Page No.* (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 Prior NSF Support) (not to exceed 15 pages) (Exceed only if allowed by a specific program announcement/solicitation or if approved in advance 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 Documentation Appendix (List below. ) (Include only if allowed by a specific program announcement/ solicitation or if approved in advance by the appropriate NSF Assistant 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.

4 1 Introduction Sources of TeV gamma rays are some of the most extreme objects in the Universe: relativistic jets from supermassive black holes, relativistic winds from rapidly spinning neutron stars, shocks formed by the remnants of supernova explosions, and merging compact objects. Details of the acceleration processes are still a mystery. Are electrons or hadrons accelerated? What are the magnetic fields in the acceleration region? What is the Lorentz factor of the material ejected from the black hole environment? Why are jets so ubiquitous in the Universe? Are these same sources responsible for the ultra-high-energy cosmic rays? Gamma ray observations at the highest possible energies are critical to answering these questions. We are proposing to build HAWC, a survey instrument capable of expanding the limited catalog of TeV gamma ray sources and of daily monitoring this catalog for flaring activity. HAWC will not only detect new sources, but will also promptly initiate campaigns of deeper searches for TeV gamma ray and neutrino emission as well as multi-wavelength observations. HAWC combines water Cherenkov technology with an extreme altitude site. Re-deploying the existing Milagro photomultiplier tubes (PMTs) and electronics in a different configuration at an altitude above 4000m leads to a factor of ~15 increase in sensitivity. As a result of these improvements the HAWC detector will see a 5σ signal from a Crab-like source in one 4-hr transit maintaining this sensitivity over 2π sr. This new detector can be built for approximately the cost of the original Milagro instrument when adjusted for inflation. Table 1 summarizes the improvement expected in HAWC relative to Milagro, currently the most sensitive large field of view TeV gamma-ray observatory. Milagro HAWC Detector Area 4000 m 2 (surface) 22,500 m m 2 (muon) Time to 5σ on the Crab 120 days 1 day Median Energy 4 TeV 1 TeV Angular Resolution o 0.25 o 0.50 o Hadron Rejection eff. 90% 95%-99%* Gamma-Ray Efficiency 50% 50% Q for gamma/hadron rejection * Time to detect 5 Crab flare at 5σ 5 days minutes Eff. Area at 0 GeV 5 m 2 0 m 2 Eff. Area at 1 TeV 3 m 2 20x 3 m 2 Eff Area at TeV 20x 3 m 2 50x 3 m 2 Volume of Universe where -6 erg/cm 2 GRB 2 Gpc 3 47 Gpc 3 detectable Flux Sensitivity to a Crab-like source (1 year) (5σ detection) 625 mcrab 50 mcrab Table 1 Comparison of the Milagro detector to the proposed HAWC detector. * denotes trigger multiplicities of 50 and 200 pmts. National Academy studies have repeatedly recognized the need for gamma-ray observatories. This interdisciplinary field combines the knowledge of astrophysics with the detectors of particle physics and the theories of extreme gravity and cosmology. The report of the Committee on Elementary Particle Physics in the 21 st Century, Revealing the Hidden Nature of Space and Time: Charting the Course for Elementary Particle Physics (EPP 20) 1, lists the field of gamma-ray astrophysics in the chapter on The Experimental Opportunities, and emphasizes the importance of astroparticle physics projects where the intellectual and technological capabilities of particle physicists can make unique contributions.

5 In the National Academy study, Connecting Quarks with the Cosmos: Eleven Science Questions for the New Century 2, the understanding the sources of cosmic rays was one of the eleven questions for the next century: How do cosmic accelerators work and what are they accelerating? Following this report, the President s Office of Science and Technology Policy (OSTP) assembled a road map, A 21 st Century Frontier for Discovery: The Physics of the Universe 3, of how the various federal agencies would implement the recommendations. Earlier, in the decadal astronomy study report, Decade of Discovery in Astronomy and Astrophysics 4 the panel noted that None of the existing or planned observatories have the sensitivity to detect very many sources they are what astronomers call starved for photons. This was written before this proposal. The HAWC observatory, with its large field of view, a sensitivity comparable to the current generation of imaging air Cherenkov Telescopes (IACTs), and high duty cycle has the capability to radically alter this situation. The outline for the remainder of this proposal is as follows: In section 2 we detail the scientific motivation for HAWC. In section 3, we give results from previous support - concentrating on the results from Milagro and briefly describing other projects in which we have participated. In section 4, we present the technical design of HAWC. In section 5, we present the expected performance of HAWC. In section 6, we detail our plans for site selection, design completion and the transition from Milagro to HAWC. In section 7, we describe our current education and outreach plus our plans for HAWC. Section 8 is a summary. This is a proposal to build the entire HAWC experiment. Potential collaborators in both Mexico and China have proposals to their funding agencies for substantial contributions to this project. Once site selection is complete, the site-specific design completed, and the amount of foreign contributions known, the exact cost to the US funding agencies can be determined. The budget presented in this proposal is an upper limit to that cost, since it assumes no foreign contribution. 2 Scientific Motivation The HAWC TeV gamma-ray observatory will make contributions to a broad range of problems facing gamma-ray astronomy. We will: Perform an unbiased sky survey over 2π sr. Our source detection threshold will be 50 mcrab in a single year. Search for extended sources of TeV gamma rays with significantly better sensitivity that the current generation of IACTs. Measure the spectra of TeV sources to the highest possible energies above that accessible to GLAST or IACTs. Search for transient sources (AGN flares and GRBs) that require long observations to catch rare episodic phenomena. This section delineates how HAWC will Figure 1 Sensitivity of HAWC (1 year solid blue) and HESS (50 hrs - red), and Milagro (dashed blue) as a function of spectral index. While the sensitivity of HAWC to the Crab, a relatively soft source, is ~3 times worse than HESS, the sensitivity of both instruments to point sources with hard spectra is comparable. contribute to this rapidly growing field that investigates Nature s highest energy particle accelerators.

6 2.1 All Sky Survey One of the primary goals of HAWC is to perform a complete, unbiased TeV survey of half of the sky. An all-sky, high-duty cycle instrument such as HAWC, is the only instrument suited to perform such a survey. For example, if HESS or VERITAS were to devote a year of observation time to survey the sky, they could spend roughly 7 minutes per field-of-view and the sensitivity to a steady source would be ~0 mcrab. In contrast, HAWC will spend over 1500 hours per year viewing every source and achieve a flux sensitivity of 50 mcrab over the entire hemisphere (25 mcrab in 4 years). For comparison both COS-B and Milagro attained a level of sensitivity near 1 Crab and both instruments detected a small number of sources. With a sensitivity of about mcrab EGRET detected roughly 370 sources. More recently, HESS performed a survey of 0 square degrees around the Galactic plane at a sensitivity level of 30 mcrab and detected 18 new sources. HAWC will reach this sensitivity level in 2 years over 20,000 sq. degrees. While a direct prediction of the number of sources detectable by HAWC is impossible to make, it is clear that HAWC will significantly increase the number of TeV sources. Figure 1 shows the sensitivity of HAWC, Milagro, and HESS to point sources as a function of the energy spectrum of the source. Because the background rejection capability of HAWC improves with energy the sensitivity of HAWC is better for sources with harder spectra. Moreover, the large exposure of HAWC results in improved photon statistics at high energies relative to IACTS. With an effective area of ~50,000 m 2 above TeV, HAWC will detect ~20 gamma rays annually from the Crab with energies above 0 TeV (assuming an unbroken power law spectrum). For hard-spectrum sources (~E -2 ) HAWC has better sensitivity above 700 GeV than does GLAST above GeV. For hard spectrum sources without a high energy cut off, the exposure of HAWC will make it possible to detect and measure spectra of even dim sources well beyond 50 TeV. 2.2 Extended TeV Sources Gamma rays are produced when higher energy particles interact with matter or radiation fields. Because these particles can escape from the acceleration region, many sources of very-high-energy gamma rays are extended. HESS has observed that most Galactic sources are extended. Figure 2 shows the sensitivity as a function of source size for HAWC and IACTs. The ability to detect of a source of γ-rays depends upon the Signalto-Noise Ratio S/N, where the signal S is the number of γ-rays recorded from a source and the noise N is the square root of the number of background events recorded over the larger of the detector angular resolution θ resolution or the source size θ source. Therefore, if the source is larger than the angular resolution of the detector, it is the source size and not the angular resolution Figure 2 - HAWC survey vs. IACT sensitivity as a function of the source extent. The units are given in the integral Crab flux with the IACT threshold of 200 GeV and the HAWC (>200 multiplicity trigger) median energy of 3.8 TeV. The HESS sources are plotted with the >200 GeV integral fluxes. The Milagro source in the Cygnus region is shown both for a median energy of 15 TeV and for 200 GeV by extrapolating using a differential spectrum of index -2.3, which is the average of the HESS galactic sources. Note that the HESS sources cluster just above their sensitivity. that determines the sensitivity of the detector. In this case, for a fixed source flux, the surface brightness of the source is decreasing as the source extent increases, and the flux sensitivity decreases. For HAWC, the sensitivity begins to decrease for a source with angular extent of 0.25 o.

7 However, since HAWC has a significantly larger observation time on each source compared to an IACT our flux sensitivity to extended sources is much better than the current generation of IACTs. Figure 2 also shows the angular extent of the sources detected by HESS. The fact that many sources are clustered around the HESS sensitivity limit indicates that they only be seeing the tip of the iceberg. The results from their deep survey (in green) prove this hypothesis. With its dramatically better sensitivity to extended sources HAWC will detect many such sources. Figure 3 shows the sensitivity of HAWC to sources smaller than 0.25 degrees (including point sources) after 2 years of operation. Also shown is the expected sky coverage and sensitivity of the HESS Galactic Center survey 5 and the expected exposure of the VERITAS Galactic Plane Survey during a 2 year exposure. The figure shows the difference in the IACT sensitivity for point (red) sources and for sources extended by 0.25 O (green), typical of the size of the majority of Galactic sources observed by the HESS Survey 6. While the point source sensitivity of HAWC is approximately 80-95% of the IACT array Galactic surveys, it is roughly 3 times better than IACTs for Figure 3 - Flux limits for HAWC versus the HESS sky survey and the proposed VERITAS sky survey over the next two years. HAWC is assumed to be located in Mexico (19 O N). IACT sensitivity is shown for point (red) sources and for sources extended by 0.25 O (green). A Crab-like spectrum is assumed. The HAWC limits will be lower for harder sources. sources extended by 0.25 o or more. HAWC provides substantially improved sensitivity over IACT surveys for diffuse sources, and this sensitivity extends over a 2π sr for the same 2-year observation period. There is strong synergy between the better flux sensitivity of HAWC to extended sources and the superior angular resolution of IACTs over their narrow field of view. HAWC will identify a large number of extended sources, undetectable with IACTs. The IACTs will then target these regions for deep observations to provide high-resolution morphological and spectral information. The combination of the diffuse sensitivity of HAWC with the deeper, higher angular resolution IACT follow-up observations provides the most efficient way to map the entire sky over all angular scales. 2.3 The Galactic Plane The galactic plane is the brightest feature in the GeV and TeV sky (as seen by EGRET and Milagro). While some of the emission is likely due to point sources, a large fraction is due to cosmic ray interactions with the matter and radiation fields in the Galaxy and gamma-ray observations are the most direct probe of the flux and spectrum of cosmic rays outside our solar neighborhood. Cosmic ray hadrons interacting with matter produce neutral pions that decay to give gamma rays. Cosmic ray electrons create high-energy gamma rays through inverse Compton scattering with infrared photons. Morphological comparisons of the diffuse gamma ray emission and the hydrogen column density constrain the contributions from each of these processes. Inversely, given a model of the galactic cosmic ray distribution, the neutral pion component can be used to probe the molecular and atomic hydrogen density and thereby measure the ratio of CO to molecular hydrogen in various regions of the galaxy as was done with EGRET 7. In addition, processes not directly related to cosmic ray production may also contribute to the diffuse emission. For example, self-annihilating super-symmetric dark matter could play a significant role 8. The Gaussian width of the galactic plane as measured by Milagro (see Figure 11 of Section 3) is ~2.5 degrees. Therefore, as seen in Figure 3, HAWC will have much better sensitivity to this emission than IACTs. The diffuse GeV and TeV gamma-ray flux recorded with EGRET and

8 Milagro respectively are above predictions based upon the assumption that local cosmic rays are representative of the entire volume of the Galaxy. The local cosmic ray density would have to be un-typically low at Earth 9 or the spectrum softer than in other regions of the Galaxy to explain these observations. With deeper, higher angular resolution observations HAWC will make significant progress in our understanding of Galactic cosmic rays. Regions with harder spectra could be identified as cosmic ray acceleration regions. Spectral studies will reveal new emission processes or identify inverse Compton components from their synchrotron cooling cutoff, thereby constraining the local magnetic field. Gamma rays from secondary neutral pions have energies an order of magnitude less than the primary protons. Above TeV, HAWC will be probing the knee of the cosmic ray spectrum throughout the galaxy. With HAWC observations we will make significant progress in understanding the acceleration and propagation of Galactic cosmic rays Extended Galactic Sources Several supernova remnants and pulsar wind nebulae have now been detected with HESS. As shown on Figure 1, each of these objects could be detected with HAWC. Therefore, HAWC will be able to perform a systematic search for similar objects in the northern hemisphere. HAWC along with GLAST will serve as a search light for IACT experiments that will achieve higher angular resolution observations of these sources once discovered. It is likely that many of the Galactic sources detected by HESS are sites of cosmic-ray acceleration. All of them have X-ray counterparts, suggesting that the gamma rays result from inverse Compton scattering. They show little or no morphological correlation with local molecular clouds even when a purely inverse Compton spectrum does not fully account for the measured spectral energy density. HAWC s deep exposure will yield spectral measurements of these cosmic-ray accelerators between and 0 TeV. In this energy range, the inverse Compton component should cut-off as a result of synchrotron cooling. A detection at such high energies is strong evidence for a hadronic component and therefore a signature of a cosmic-ray accelerator. Most supernova remnants and pulsar wind nebulae detected with HESS have a resolved angular extension. Figure 2 suggests the existence of lower surface brightness objects that cannot be detected with IACTs but will be detectable by HAWC. One such source could be the Milagro source in the Cygnus region (see section 3.1.2). Sub-fields of the Cygnus region have been observed with IACTs but the very extended emission revealed by Milagro remained unnoticed. These sources probably have hard power law differential spectra with indices close to 2, like most Galactic objects detected with HESS. The HAWC sensitivity shown in Figure 2 assumes a Crab-like source of differential spectrum of index 2.6 would improve by a factor of ~2 as illustrated by Figure 1. Other possible Galactic sources are the EGRET unidentified objects at low and mid Galactic latitudes. While it is unclear if these sources emit gamma rays up to very high energies (VHE), even improved upper limits are useful in understanding the nature of these sources. Nearby molecular clouds emit gamma rays due to cosmic ray interactions with matter. Because they are nearby the cosmic-ray flux is presumed to be the same as at Earth, so the gamma-ray flux uniquely determines the only free parameter, which is the ratio of CO to molecular hydrogen. Because they are at high galactic latitudes and of small diameter, they may still be undiscovered, but they could contribute to the EGRET and GLAST unidentified sources. There are also predictions of as yet unobserved classes of VHE gamma-ray sources. For example, if the WIMP mass is above ~0 GeV, nearby clumps of dark matter would emit VHE gamma rays from neutralino annihilations. 11 The range of size of these clumps varies depending upon the cosmological model. And the flux of gamma rays also depends on the cross section for annihilation and the neutralino mass. However, a few nearby clumps could be visible to HAWC.

9 2.4 Extragalactic Extended Sources Some fraction of the immense gravitational energy in a cluster of galaxies is predicted to result in shocks that will accelerate electrons and protons to very high energies. Gamma rays are then produced by inverse Compton scattering of high-energy electrons on the cosmic microwave background and by protons interactions with matter. Mechanisms such as accretion 12 and merger shocks 13 have been considered, the former are more efficient at producing the highest energy particles. While there are individual galaxy clusters that are interesting to consider, it is difficult to determine which clusters are most likely to emit gamma rays because the production mechanism depends upon the history and masses of the clusters which are not known. There are over 300 galaxy clusters in the northern sky nearer than z=0.1. The angular extent of the emission is expected to be up to 1 degree in some cases. HAWC can easily survey all 300 clusters and determine which clusters emit VHE gamma rays. Other extragalactic sources include nearby galaxies. VHE gamma rays should be produced by cosmic ray interactions with matter (just as in our Galaxy). However, some galaxies may have an enhanced cosmic ray flux, such as starburst galaxies, or different relative electron and hadron fluxes and spectra than our galaxy 14. Also, nearby AGN could have gamma-ray halos extending a fraction of a degree that are produced by pair production of even higher energy gamma rays near the source 15. These halos can come from AGN that do not have jets pointed towards Earth and are a possible signature of UHECR sources. Figure 4: Comparison of the flux necessary for a GLAST detection of 5 γ-rays above GeV with the HAWC 5σ detection threshold for a source differential photon flux of spectral index -2 that is cut off due to extragalactic background absorption. The absorption is calculated assuming the model of Kneiske, et al. 2005, and the energy at which the flux is attenuated by 1/e is 700, 260, and 170 GeV for z=0.1, 0.3, and0.5, respectively. The gap between the lines on the left and right is due to the Earth blocking the view of the source. 2.5 Transient Sources There are three known types of transient TeV gamma-ray sources: active galactic nuclei (AGN), gamma-ray bursts (GRBs), and x-ray binaries/micro-quasars. While there may be common features to the acceleration of particles in these sources, observations of multiple sources from each of these classes is required to understand these objects. An all-sky high-duty cycle observatory such as HAWC will detect and monitor these sources over long time periods. GLAST will also monitor the sky; however, not all sources will be bright at the same energies. For example, the GLAST energy band may fall between the synchrotron and inverse Compton peak fluxes and therefore, be strongly suppressed. Figure 4 demonstrates the ability of HAWC to observe shorter time scale variations than GLAST and extend the energy range of observations beyond those of GLAST. If a gamma-ray burst is detected by HAWC and GLAST, the overlap will yield observations over 7 orders of magnitude in energy, all in the gamma-ray band. Why is an all-sky gamma-ray monitor needed to carry out such a research program? 1. AGN can flare to over times their quiescent luminosity. The paucity of detected AGN may well be due to the low duty cycle of the AGN themselves, coupled with the small total observation time. An all-sky monitor can observe every AGN in the same hemisphere every day. Therefore, even phenomena that are rare for single AGN can be observed frequently when observing a large population of objects.

10 2. GRBs occur roughly twice per day throughout the universe. Since they occur at random times and from random directions, only an all-sky, large duty cycle instrument can detect the prompt emission from these objects. 3. X-ray binary systems and micro-quasars can have long orbital periods (hours to years) and the high-energy emission is correlated with the orbital phase. Only by studying these objects over several orbital periods can we understand the mechanisms behind the acceleration process and the local conditions where particle acceleration occurs. 4. The discovery of new classes of objects unobserved at other wavelengths, is a major strength of all-sky monitors. Unbiased sky surveys hold the promise of not only detecting new sources of known classes, but of detecting new types of objects. 5. Since HAWC will always view every source (in its hemisphere) every day, multiwavelength and multi-messenger (with IceCube) observations will be limited only by the ability collect data with other detectors the TeV data from HAWC will always be available. This will result in a large number of such observations which have proven essential for understanding the extreme sources that are bright at TeV energies. What are the characteristics of HAWC that suit it to complete such an observational program? 1. The instantaneous field-of-view of HAWC is approximately 2 steradians. With the rotation of the Earth HAWC will view over 2π steradians of the sky every day. 2. Based upon our experience with Milagro we known that we can achieve a duty cycle of 90%-95%. 3. The sensitivity of HAWC is such that a source with a flux 5x that of the Crab Nebula can be detected in a minute period. Given previous studies of AGN flaring structure at TeV energies, HAWC is guaranteed a large number of such detections. 4. The low-energy response and background rejection abilities of HAWC are such that a 0 second long GRB, with a fluence as emitted at the source above -5 ergs cm -2 can be detected out to a redshift of Active Galactic Nuclei (AGN) Active galactic nuclei are supermassive black holes (~ 8 times the mass of the Sun) that emit jets of relativistic particles along their rotation axis. The jets from these objects can be extremely well collimated and can extend for millions of light years before dispersing. Different classes of AGN exist and the observational properties are thought to be related to the angle between the rotation axis and our line of sight. If the relativistic jet is pointed towards the Earth (within 5- degrees) the emission lines are either dim or unobservable, the objects are highly variable and emit much of their energy in gamma rays. Such AGN are known as BL Lacertae objects (or BL Lacs). Particle acceleration occurs in shocks that propagate along the jets. In this scenario the acceleration mechanism is a first order Fermi process. AGN spectra show a characteristic twohump structure, with a peak in the TeV band and one in the x-ray band. At present there is no clear consensus if electrons or protons (or both) are accelerated. In electron models, gamma rays are produced by inverse Compton scattering with lower-energy photons. In the synchrotron self-compton (SSC) model the low-energy photons are a result of synchrotron radiation of the same population of electrons. In hadronic models the gamma rays are produced by a hadronic cascade that originates with a pγ interaction. This requires protons with energies of ~ 18 ev, making such objects natural sources of UHECRs. In general, rapid variability favors electron acceleration while higher energies favor proton acceleration. Since the shocks can be moving at extreme relativistic velocities (bulk Lorentz factors near 40) both the duration and the maximum attainable particle energy are not by themselves conclusive proof of either model. Instead multi-wavelength and multi-messenger studies are required a natural by product of an all-sky monitor.

11 Orphan Flare Figure 5: Simultaneous multi-wavelength campaign of involving both HEGRA and Whipple TeV observations as well as x-ray observations showed flares correlated with x-rays as well as an orphan flare. 1 The 5 sigma HAWC sensitivity will be ~1 Crab for one day s observation and 0.3 Crab for days. The orphan flare flux of 5 times the Crab could be detected in minutes. To date TeV emission has been observed from 8 AGN. EGRET observations showed that 70% of the AGN were variable. 16 Since energy loss is more rapid with increasing electron energy one would expect the TeV observations to exhibit even greater variability. Observations of Mrk 501 and Mrk 421 show large flares at TeV energies. However, no flares have been observed from the newly discovered AGN. The lack of TeV variability may simply be due to the lack of long time scale, continuous observations. TeV variability provides important information for understanding the both the acceleration process and the environment near the acceleration sites. Short time scale variability probes within -0 AU of the event horizon of the black hole. If AGN are accelerating electrons, then (in the SSC model) the TeV emission should be correlated with longer wavelength emission (in particular x-ray). While several TeV flares follow this pattern, there have been orphan TeV flares where there is no commensurate change in the x-ray flux. Figure 5 shows the data from the Whipple and HEGRA telescopes along with concurrent x-ray observations. HAWC naturally provides a mechanism for obtaining many such multi-wavelength datasets and will allow us to study orphan flares and the correlations between TeV gamma ray, x-ray, optical, and neutrino emission in detail Gamma Ray Bursts (GRBs) Emitting over 52 ergs in gamma rays, gamma ray bursts are the most energetic phenomena known in the universe. Like AGN, the emission is thought to be collimated in jets however the bulk Lorentz factor of the particle flows may be as large as 00 in GRBs. GRBs last from fractions of a second to ~00 seconds and the duration distribution is bi-modal. The distinction between short and long bursts is roughly 2 seconds and the average duration of the short bursts is 0.2 seconds and long bursts 20 seconds. It is believed that the progenitors of short and long bursts are different. The prevailing model of short bursts is the coalescence binary neutron star systems and for long bursts the collapse of a supermassive star. In both cases the energy source is the gravitational potential energy Figure 6: The redshift distribution of short, < 2 seconds, (left) and released by the accretion of long, > 2 second duration bursts (right) as measured and for matter onto a compact object. various models. (Left plot is from Guetta & Piran, A&A, 453, 823, 2006 and the right plot is from Guetta, Piran, & Waxman ApJ, 619, The highest energy gamma 412, 2005.) The shorter bursts are closer, but are not observable by ray conclusively detected from a an IACT. GRB is an 18 GeV photon

12 detected by EGRET roughly 90 minutes after the onset of the burst 17. In addition, there is evidence for a high-energy component to GRBs that extends to 200 MeV with a spectral index of This observation has been cited as evidence for proton acceleration in GRBs, with the implication that GRBs accelerate protons to energies above 18 ev. While there has been no observational evidence for a spectral break at the highest energies, detection of higher energy photons has proven elusive. The best such evidence comes from the Milagrito detector (see section 3.1.4). Theoretical considerations argue for the creation of >0 GeV gamma rays in GRBs 19,20,21. As in AGN, a first order Fermi process is believed responsible for particle acceleration, however in GRBs shocks are created when the relativistic jet interacts with shells moving at different velocities and with the circumburst medium. Gamma rays are produced by either leptonic or hadronic processes and yield multi-wavelength spectra similar to AGN, where the kev-mev emission is likely due to synchrotron radiation. Measurements of the evolution of the flux at different energies provide strong constraints on the magnetic fields, the circumburst medium, and the bulk Lorentz factors. In order to observe a GRB above 0 GeV the burst must not only be within the field of view of the observatory, but its redshift must be below ~1. Gamma rays of energy E γ emitted at a redshift z are attenuated by pair production on the extragalactic background light and the optical depth is τ ~ z 4/3 (E γ /90 GeV) 3/2 for 0.1<z<2. 22 So for a GRB at z = 0.1, 0.5 or 1, the gamma ray flux is reduced by a factor of 1/e = 0.37 at E γ = 700, 170, or 90 GeV, respectively. Figure 6 shows the redshift distribution of GRBs. While IACTs may be able to slew to the longest bursts prior to the end of emission, this is not possible for short bursts. Approximately 1/3 of all bursts observed by BATSE were short bursts and these bursts are from considerably lower redshift. The sensitivity of Milagro and HAWC to a second GRB are shown in Figure 7. While Milagro can only detect the brightest and nearest GRBs, HAWC can detect a large fraction of GRBs even at high redshift if the VHE flux is comparable to the kev-mev flux Galactic Transients An x-ray binary is a system composed of a compact object (black hole or neutron star) and a massive star. If the system exhibits jet-like behavior it is referred to as a micro-quasar. An accretion disk forms around the compact object and accretion onto the compact object is believed to be the energy source responsible for particle acceleration. The variability of these systems is usually associated with orbital phase (due to an eclipse or obscuration of the acceleration site), but there may also be non-periodic flaring behavior similar to that observed in AGN. Over 0 x-ray binaries and 12 micro-quasars have been cataloged, with orbital periods ranging from hours to years. Sensitivity to Second GRB Sensitivity to Second GRB ) -2 dn/de (ergs cm 2 E degrees -20 degrees 0- degrees -4 Detected GRBs -5-6 ) -2 dn/de (ergs cm 2 E degrees -20 degrees 0- degrees -4 Detected GRBs Redshift Redshift Figure 7 Fluence sensitivity as emitted at the source for a 5σ detection of a second GRB vs. redshift for HAWC (left) and Milagro (right). The different color lines indicate the sensitivity for GRBs at different zenith angles. The superimposed triangles indicate the kev-mev fluence and redshift of satellite detected GRBs.

13 HESS and MAGIC have detected TeV emission from three x-ray binary systems, two of which are micro-quasars. In all cases, the TeV emission is modulated with the orbital phase. The TeV emission from these sources is difficult to explain because the variability implies a small source region and hence a high photon density. Therefore, the optical depth for gammaray pair production in these sources is high, yet the TeV emission extends to high energies with a hard spectral index. Figure 8 shows the TeV lightcurve of , a binary system with a very eccentric orbit with a period of 3.4 years. Different models predict either a maximum or minimum gamma-ray flux near periastron. Unfortunately, the moon prevented HESS observations precisely at periastron; so different model predictions could not be tested. The next periastron will occur during the day, again preventing observations by IACTs. The MAGIC source, LS I , has a period of 26 days which is also difficult to monitor with an IACT (being close to the lunar period). The third source (LS 5039) has a 3.9-day period and HESS has clearly observed an orbital modulation with the maximum gamma-ray emission occurring at the inferior conjunction of the orbit (along our line-of-sight where the compact object is not eclipsed by the star) and a near minimum at periastron. In addition to these sources, there are probably other types of transient Galactic TeV sources that have remained hidden after the limited surveys of the Galactic plane. For example, a large fraction of the EGRET unidentified sources at low Galactic latitudes are variable 23. While much can be learned from the detailed study of a single object, it is only by studying a large number of objects and understanding their differences that a true understanding of these systems will arise. With HAWC we will have the opportunity to perform long-term high-duty cycle observations of all of the x-ray binaries and micro-quasars in our hemisphere and search for new types of Galactic variable sources Solar Energetic Particles Our Sun is the nearest astrophysical particle accelerator. Solar particles in excess of GeV have been recorded. HAWC provides diagnostic and discovery potential in the area of energetic solar particles and the dynamics of the inner heliosphere. Using the rates in individual photomultiplier tubes, Milagro has demonstrated its ability to measure protons and ions accelerated by coronal shocks. 24 HAWC with its greater sensitivity will be able to detect the weakest flux of protons above the rigidity cutoff of its locale. Because of the low geomagnetic latitude of both candidate sites, measuring the tails of the high-energy proton distribution will provide new diagnostic capabilities for investigating coronal shock acceleration. HAWC will work in concert with nearby ground level neutron monitors, as Milagro has done with the Climax station. Full Moon Figure 8: TeV lightcurve obtained by HESS for the binary pulsar which has a highly eccentric orbit

14 of 3.4 years that emits TeV gamma-rays near periastron 25. Three models for the TeV emission are shown, but the TeV observations at periastron were not possible due to moonlight. The next periastron will be when the source is too close to the Sun to be observed by atmospheric Cherenkov telescopes The low geographic latitudes of the candidate sites also allows HAWC to measure the highest energy solar neutrons that emanate from solar flares themselves, again providing new diagnostics about particle acceleration within a flare environment. Large-scale magnetic structures inside the inner heliosphere modulate the Galactic cosmic ray flux at Earth. Therefore, we can use measurements of the cosmic-ray flux and anisotropy to study these phenomena. Conversely, measurements of time-dependent cosmic-ray anisotropies are telltale signs of approaching coronal mass ejections not visible by other means. 2.6 Synergy with other High Energy Astrophysics Projects When new wavelength bands are explored in astronomy, previously unknown sources and unknown types of sources are discovered. For example, the EGRET catalog 26 contains over 150 previously unidentified sources, HESS has discovered several sources with no known counterparts, and Milagro has detected a new source in the Cygnus Region with no obvious counterpart. The key to understanding these new sources is multi-wavelength and multimessenger observations. A large field-of-view, high duty-cycle instrument such as HAWC is the ideal instrument to enable such observations. HAWC benefits from and can benefit observations made by other instruments. Some examples are listed in Table 2. GLAST is expected to detect over 3000 unidentified gamma-ray sources 27. Continuous observation of the sky coincident with the GLAST satellite will provide a wealth of data, particularly on transient phenomena. HAWC will provide a natural extension of the energy reach of GLAST to the TeV scale and beyond for all sources within HAWC s field of view. IACT s such as VERITAS and HESS have outstanding point source sensitivity due to their excellent angular resolution (~0.05 o ) and low energy thresholds (~0 GeV). They have the capability to image the small-scale of Galactic TeV sources. However, they are poorly suited for surveys of large regions of the sky because of their narrow FOV and limited duty cycle. HAWC will be capable of both discovering new TeV sources and providing rapid alerts of TeV flares. Furthermore, at the highest energies (-0 TeV), the IACT s will be photon starved for all but the brightest sources while HAWC s larger exposure (and better background rejection) makes it more sensitive at the highest energies. HAWC will provide a natural complement to existing and future IACT observatories, both in energy and field of view. IceCube is a neutrino detector sensitive to TeV-PeV neutrinos from astrophysical sources. HAWC and IceCube will observe the same range of energies, and the proton cascade process produces comparable fluxes of photons and neutrinos at similar energies. Both HAWC and IceCube operate with nearly 0% duty cycle and both observe very large fields of view, making them very good instruments for observing transient phenomena. Both HAWC sites would give excellent coverage of the Northern sky, which is visible to IceCube. Because neutrino emission requires hadronic acceleration, the detection of neutrinos is important for identifying the sources of the UHECRs. Neutrino emission is necessarily coincident with VHE gamma ray emission (though the gamma rays may be absorbed at the source or in transit to the earth). The most likely candidate neutrino sources, AGN and GRBs, are highly variable. HAWC s observations of flaring sources are thus very useful to select the direction and time interval from which to search for neutrino emission. Such a selection dramatically increases the sensitivity of IceCube by reducing the background.

15 Table 2 Synergistic Projects. Synergistic Example of Relevant HAWC Observations Projects GLAST Constraining TeV emission for 00s of GLAST sources Identification of GLAST sources by TeV detection Contemporaneous searches for flaring AGN ACTs Search of 2π sr for TeV sources with fluxes > 50 mcrab Prompt notification of TeV transients Monitoring of TeV sources during daylight, moonlight, and bad weather IceCube Search of 2π sr for TeV sources extending to the highest energies Detect TeV spectra and variability indicative of hadronic accelerators Identify flaring episodes to select shorter IceCube observation times, reducing background, and allowing a lower energy neutrino threshold 3 Results from Previous Support Members of the HAWC collaboration have a broad base in particle astrophysics and high energy physics experiments (D0). The majority of the collaboration was actively involved in Milagro. Others have joined with experience in GLAST, VERITAS, IceCube, Auger, and HiRes. Having these strong connections to other experiments in the field will help us realize the many synergies between HAWC and these other instruments. In this section we will concentrate on the results from Milagro as it is most closely related to the proposed work in HAWC. However, elsewhere in this proposal, we point out the impact this wide-field instrument will have on the rest of the field and particularly where we already have connections. Los Alamos, UCI, and Maryland were all supported by the NSF to work on Milagro. 3.1 Milagro Milagro is the first large area, continuously operating, water Cherenkov detector used for gamma-ray astronomy. As such Milagro is a prototype for the technology and with the detection of the Crab Nebula its sensitivity relative to other techniques can be calibrated and (most importantly for this proposal) the veracity of the Monte Carlo simulation of the response of such an instrument can be demonstrated. Milagro began physics operation in 2000 taking data with the central water reservoir. In 2004 construction of the outrigger array was completed. Before the installation of the outriggers the small size of the water reservoir limited the sensitivity of Milagro such that the Crab Nebula was observed at ~4σ in one year of operation. The completion of the outriggers enabled a large increase in the sensitivity of the instrument, enabling us to detect the Crab Nebula at over 8 standard deviations in a single year of observation. This factor of ~2 increase in sensitivity (as shown in Figure ) has dramatically changed our view of the high-energy sky. It also means that the data currently being taken now with Milagro is substantially more important than our original data and we are not simply increasing our sensitivity by the square root of time over a 6-year observational period. To date our most important observational results with Milagro have been: the first detection of TeV gamma rays from the Galactic plane 28, the mapping of the diffuse Galactic gamma-ray emission at TeV energies, including the detection of the Cygnus Region at highsignificance (over σ) 29, the discovery of a new (slightly extended) source of TeV gamma rays embedded in the Cygnus Region 30, and the possible detection of a gamma-ray burst with our prototype instrument Milagrito 31. In addition, we have detected TeV gamma rays from the active galaxy Mrk , Mrk and the Crab Nebula 34, set stringent upper limits on the prompt TeV emission from several gamma ray bursts 35, and performed the most sensitive survey of the northern hemisphere at TeV energies 36.

16 3.1.1 Survey of the Northern Hemisphere The sensitivity of Milagro has been dramatically improved with the addition of the outrigger array. The effect of this gain in sensitivity is best demonstrated by comparing the results of our published sky survey 37 which used approximately 00 days of data (top of Figure ) and our current data set of just over 2000 days of data (bottom of Figure ). While both the Crab Nebula and the active galaxy Mrk421 are visible in the top panel, the improvement since the outriggers is dramatic. The significance of the Crab Nebula has increased from 6σ to 14σ and the Galactic plane is now clearly visible, even in this broad map. Cygnus Region MRK 421 Crab Nebula Figure. A northern sky seen in TeV gamma rays using Milagro data. The top panel shows the sky as observed before the outriggers and the bottom panel after the completion of the outrigger array The Cygnus Region and the Discovery of a TeV γ-ray Source In 2005, we published the first detection of diffuse TeV gamma-ray emission from the inner Galaxy 38. The flux of the Milagro detection is not consistent with expectations from cosmic ray interactions if the local cosmic ray flux is indicative of the flux in the rest of the galaxy. There are several possible explanations for this excess: the local cosmic ray flux is unusually low, the local spectral index is soft relative to the rest of the Galaxy, the existence of unresolved point sources. With the more recent data, taken since the completion of the outriggers, we have refined the analysis to investigate the Cygnus Region of the Galaxy in more detail. The Cygnus Region of the Galaxy is a natural laboratory for the study of cosmic ray origins. It contains a large column density of interstellar gas that should lead to strong emission of diffuse gamma rays and is also the home of potential cosmic-ray acceleration sites (Wolf-Rayet stars 39, OB associations 40, and supernova remnants 41 ). Figure 11 shows a detailed view of the Cygnus Region in TeV gamma rays. Superimposed on the figure are contour lines indicating the matter density in the region and the location (and location errors) of the EGRET sources (all unidentified) in the region. There is definitive evidence for a new source of TeV gamma rays (MGRO J ). MGRO J is detected at over 11 standard deviations and its location is consistent with the location of two EGRET sources one of which has been tentatively identified with a pulsar wind nebula 42. Though it is not evident from this figure, this

17 source is most likely extended with a width of 0.32±0.12 degrees. The best-fit location of this source is RA= o ± 0.13 o, Dec=36.96 o ± 0.08 o. The flux of diffuse emission from the Cygnus Region is measured after subtraction of the contribution from MGRO Using the GALPROP 43 program we estimate the expected flux of diffuse gamma rays. This estimate has contributions from a pion component arising from the interactions of cosmic rays with matter in the region and from gamma rays produced by inverse Compton interactions of cosmic-ray electrons with infrared radiation in the region. We Figure 11 - The Cygnus Region of the Galaxy as seen in TeV gamma rays. The color scale is the statistical significance of the gamma-ray excess at each location (points below -2 standard deviation have been suppressed for clarity). Superimposed on the image are contours showing the matter density in the region. The probability that the two distributions are not correlated is 1.5x -6, indicating that the TeV emission is due to interactions of cosmic rays with matter in the region. The crosses show the location of the EGRET sources and their corresponding location errors. Figure 12. The top figure shows a latitude scan of the Galactic plane in 1 degree bins for the inner Galaxy. The lower figure is a longitude scan in 2 deg bins of +/- 2 deg around the Galactic plane. find that the TeV gamma ray flux is about a factor of 3 larger than that predicted by this standard model of cosmic ray production in the Galaxy. The same reference gives a model that explains the GeV excess that can also explain the TeV excess however, at energies near TeV it implies that the inverse Compton component is dominant and the inverse Compton component would not show such a strong correlation with the matter density. Other plausible explanations of this observation are that the Cygnus region contains cosmic-ray accelerators thereby increasing the cosmic-ray density in that region relative to the model predictions of GALPROP. Figure 12 shows the profiles in latitude and longitude of the inner Galaxy (from 20-0 degrees in Galactic longitude). The median energy of the detected gamma rays is ~12 TeV. It can be seen that the plane has a width of ~±2.5 o in latitude. The longitude profile shows a large excess in the Cygnus region as well as an increasing flux towards the Galactic center, however with larger error bars because of the limited view of the Galactic center from the latitude of Milagro The Spectrum of the Crab Nebula We have recently developed a technique to measure the energy of the primary gamma rays that trigger Milagro. This technique requires the outrigger array and therefore can only be applied to recent data. Using this procedure, energy spectra of gamma rays from the Crab nebula and the cosmic rays have been determined in the energy range from 1 to 0 TeV. We

18 measure the cosmic-ray spectral index to be γ = ± 0.08 near TeV, in excellent agreement with direct measurements in this energy range. We have fit the Crab data to a single power law and also to a power law with an exponential cutoff. The data is shown in Figure 13. The fit to a power law gives dn/de = (4.84±1.23) x - TeV -1 m -2 s -1 (E ) ±0.14, where E is the energy divided by TeV. A fit with the addition of an exponential cutoff changes the normalization to I 0 =(1.04±0.56) x -9 TeV -1 m -2 s -1, the power law index to γ= ± 0.39, and the exponential cut off energy is E c = 31.0 ± 26 TeV. The improvement in χ 2 from the addition of the one degree of freedom, E c, indicates a probable cutoff above TeV. As shown in Figure 13 these results are in good agreement with previous measurements and Milagro provides competitive measurements at the highest energies Gamma Ray Bursts There have been several reports of TeV counterparts to gamma ray bursts 44, 45. However, the most compelling such evidence has come from Milagrito 46, the prototype detector for Milagro. Milagrito operated for about 1 year from February 1997 to May During this time the BATSE satellite was operational and detected 54 GRBs that were within the field-of-view of Milagrito. There was a significant excess of events observed from one of these bursts (18 events detected with a background of 3.46 events). The Poisson probability of this occurrence is 3x -8, but because of the poor localization capability of BATSE, a large number of trials had to be performed to scan the BATSE error box. While the excess was found within the 1-sigma BATSE error box we had chosen an a priori search strategy that searched the 2-sigma BATSE error box. After accounting for these trials and the fact that 54 GRBs were examined the post-trial probability was 1.2x -3 that this observation was a fluctuation of the background. During the time that Milagro has been operational there has not been a satellite detector comparable to BATSE and the number of known GRBs within our fieldof-view is limited. With the launch of SWIFT, we now have a comparable number of bursts within our field of view; however, the average redshift of these bursts is about twice that of the BATSE detected bursts. (This is due to the increased sensitivity SWIFT and its smaller field of view.) During the Milagro operation we have not found any further evidence of TeV emission from GRBs. These null results can be used in two ways. First, for GRBs that are detected by satellites and for which a Figure 14. Milagro s upper limit on TeV emission for GRB0921 at z=0.45 Figure 13: Milagro Crab spectrum and comparison with other observations. The dashed line is the measured Milagro spectrum with an exponential cutoff. The Milagro data are indicated by the black triangles. Our data are in good agreement with the recent measurement of a highenergy cutoff by HESS. redshift is known we can set stringent limits on the TeV emission, accounting for the absorption by the extragalactic background light (EBL). The left panel in Figure 14 shows these results for a single burst and two different assumptions for the EBL. Even at a redshift of 0.45 the limit set by Milagro is interesting, ruling out models that predict a substantially higher TeV component 47. We can also search our entire dataset for evidence of

19 transient emission from any direction in the overhead sky. The non-observation of significant emission over any timescale can be used to set limits on TeV emission from a population of standard GRBs. These limits are highly model dependent, depending upon the redshift distribution of the bursts, assumptions about evolutionary affects in bursts (such as correlations between intensity and or durations with redshift), and the luminosity distribution of the bursts. Under the assumptions that produce the largest TeV flux we would conclude that no more than 30% of all GRBs have a TeV luminosity greater than the observed kev luminosity. 3.2 The University of Maryland Result of Previous Support The University of Maryland is supported by the NSF for the Milagro experiment. Jordan Goodman is the Milagro co-spokesman and was PI on the construction and is PI on the operations grant. Maryland was responsible for the financial control of the Milagro project which was completed within budget and on schedule. Maryland also played an important role in virtually every aspect of the project especially Andrew Smith who is full time on the project. In addition, the Maryland group is supported by NSF to work on IceCube and was previously supported to work on Super-Kamiokande. 3.3 Los Alamos National Laboratory Results of Previous Support The Experimental Astrophysics team at Los Alamos National Laboratory is supported by the DOE High Energy Physics Division, the National Science Foundation, NASA, and internal LANL funding. As the host institution for Milagro, the bulk of the LANL effort has centered on Milagro. Gus Sinnis is co-spokesman of Milagro and Brenda Dingus is the Run Manager. Here we will delineate the results of our other efforts. After September 11, 2001 the HiRes collaboration was expelled from Dugway Proving Grounds. The LANL team put together a group of Q-cleared individuals from Los Alamos to take over the operation of the HiRes detector. The National Science Foundation, through a subcontract from the University of Utah, provided the funding for this effort. Gus Sinnis has actively participated in the analysis of HiRes data concentrating his efforts in the area of the anisotropy of the arrival directions of UHECRs. Brenda Dingus at LANL also has received NASA support as an Interdisciplinary Scientist for the GLAST mission and for using Milagro to observe SWIFT bursts. Finally, LANL LDRD (Laboratory Directed Research and Development) funds have been used to develop the HAWC concept with a three-year grant of $180k/year that began in These funds have paid for simulations, site investigations, preliminary studies of an upgrade to the data acquisition system, and the cost estimates upon which this proposal is based. 3.4 University of Utah Results of Previous Support The University of Utah (UU) group consists of three faculty members who are supported by NSF to work on the VERITAS Observatory (Kieda, LeBohec) and the Pierre Auger Observatory (Mostafa). The VERITAS faculty members have primary construction responsibility for calibration systems and cabling systems and secondary responsibility for the VERITAS level 1 (constant fraction discriminator) trigger. In addition, Kieda is the leader of the VERITAS camera group. During the last 3 years, the Utah VERITAS group has designed, constructed, and installed the VERITAS calibration and cabling subsystems, and supervised the deployment and commissioning of the VERITAS-4 telescopes at the Whipple base camp. The UU VERITAS group has also pursued a scientific program involving GeV/TeV observations with the Whipple m telescope, and has completed analysis of observations of starburst and dwarf galaxies, molecular clouds in the Cygnus and Monocerus Arm, and correlative and follow-up observations of the Milagro and Tibet AS all-sky surveys.

20 The UU Auger group (Mostafa) has worked with the UNM group to develop atmospheric monitoring capabilities for the Southern Auger Observatory. Within the Auger collaboration, Prof. Mostafa has developed the analysis package for the reconstruction of hybrid events, and has published scientific analysis of gamma-ray fraction of cosmic ray events in the EeV energy range, as well as studies in EHE cosmic ray anisotropy and energy spectrum. 3.5 Michigan State Results of Previous Support Prof. Linnemann has been a member of the D0 Tevatron experiment for 22 years. He has participated in the design of the Level 1 trigger framework and calorimeter trigger, led the physics algorithms development for the D0 Level 3 trigger in , and led the hardware and software construction of the Level 2 trigger for Run II (including support from an NSF MRI grant). His two most recent students performed searches for supersymmetric particles. After a 2003 sabbatical on the Milagro experiment he has been supervising a Milagro student supported via the IGPP through Los Alamos. He has been active in developing statistical techniques and served on the organizing committee of a series of workshops for physicists and statisticians. 3.6 Pennsylvania State University Results of Previous Support The Pennsylvania State University (PSU) is presently supported by the NSF Physics Division for the analysis of IceCube data, and by NSF MREFC funds for contributions to IceCube construction. Co-PI DeYoung is responsible for the integration of AMANDA into IceCube as the nucleus of a low-energy-threshold sub-array. In this capacity he has been heavily involved in the development of online data processing and monitoring software similar to that proposed for HAWC. Prior to coming to PSU, he contributed to the design and development of the IceCube data analysis software, the IceCube simulation software, and the AMANDA and Milagro event reconstruction software. 3.7 University of New Mexico Results of Previous Support The University of New Mexico (UNM) is supported by the High Energy Physics Division of the DOE. The current DOE award number is: DE-FR02-04ER The UNM experimental particle astrophysics program has focused on the Pierre Auger experiment. Within Auger the UNM group has completed a program of R&D on atmospheric monitoring for air fluorescence detectors with the HiRes Collaboration, has provided the relative optical calibration systems for the four Auger fluorescence detector sites, as well as two light sources to monitor the aerosol phase function (normalized differential scattering cross section) and two vertical and steer-able laser facilities to monitor the aerosol vertical optical depth above the Auger Southern Observatory. 4 HAWC Technical Design The HAWC concept builds upon our experience with the Milagro detector. Milagro is the first large, uniformly instrumented, air shower array using water Cherenkov technology. The water Cherenkov technique is the best technique for EAS detection because of its superior detection efficiency, calorimetric capability, and low cost. A detailed physics case for the use of water Cherenkov technology is beyond the scope of this proposal and can be found in HAWC: A Next Generation All-Sky VHE Gamma-Ray Telescope 48. In contrast, gamma-ray detectors such as the Tibet air ASγ 49 and the ARGO observatory 50 use scintillator and resistive plate chambers respectively to detect charged particles in air showers. Gamma rays, which outnumber charged particle in EM showers by 5- times, are only detected with the addition of a layer of lead converter. As a consequence, the ARGO detector, which is larger than Milagro and located at an elevation 1700m higher, will (upon completion) have roughly the same sensitivity as Milagro 51. In addition, the Cherenkov angle in water is 41 o, therefore a PMT array

21 with a spacing of roughly ½ the water depth is sensitive over essentially 0% of its area. This is in stark contrast to a traditional EAS array that is only sensitive to particles over a few percent of the enclosed area. Finally, the water can be used as an effective shield to electromagnetic radiation. Therefore, a deep layer of PMTs (~4 meters or greater) is an effective muon detector. This is an inexpensive method to build an extremely large area muon/hadron detector. Milagro has proven the effectiveness of this method in rejecting the cosmic-ray background. Based upon the above arguments, we are proposing to build a large area water Cherenkov detector at high altitude. To keep costs to a minimum we have performed an optimization procedure and determined that the most sensitive detector consists of a single layer of PMTs spread over the largest possible area. In such a configuration there is a tradespace between detector depth and spacing. The deeper the PMTs the better the background rejection, but at the expense of the angular resolution of the instrument. While an optimal configuration is difficult to define, since different physics goals would lead to different optimizations (low-energy response versus high-energy response), we have decided on a detector design that has excellent sensitivity gains in all areas relative to Milagro. The final improvement over the Milagro design is the optical isolation of each PMT. During the operation of Milagrito (a prototype for Milagro) it was observed that light traveling horizontally across the reservoir could strike a PMT far from its point of origin. Given the large Cherenkov angle in water and the clarity of the water, this is not an infrequent occurrence. The effect is two-fold. The angular resolution suffers, since the PMT stricken by light far from its point of origin is inherently late relative to the arrival time of the shower particles. In addition, because of the abundance of nearly horizontal muons traversing the detector volume, the trigger level needs to be raised so as not to overload the data acquisition system with events triggered by these muons. In Milagro, baffles were placed on each PMT to minimize the affects of this horizontal light. In HAWC, with its much larger area (and therefore higher muon rate), a more dramatic mitigation strategy is required. Therefore, we have designed an optical isolation system such that each PMT can only detect light generated in its own detector volume. 4.1 Reservoir Design 162 m 162 m Figure 16 CAD drawing of HAWC detector building showing inner support beams to which the curtains will be attached. The HAWC detector consists of a 150m x 150m x 5m deep reservoir lined with a polypropylene-nylon liner to contain and isolate the ~125 Ml of filtered water from the ground below. The reservoir is constructed by a combination of excavation and building up soil to form a berm around the perimeter. The requirements on the slope and thickness of the berm depend on the properties of the soil and are site dependent. Milagro s 900 photomultiplier tubes will be secured on a 30x30 grid with 5m spacing. Stretching between the PMTs is an opaque curtain designed to optically isolate each sensor. The PMTs will be secured to the bottom of the pond with a weight such that the top of the photocathode is 4m below the surface of the water. The PMT/weight unit will be secured to a rope that extends to the surface; so that the PMT can be raised to the surface for maintenance. A concrete footing at the top of the berm will serve as anchor for a building that covers the pond. See Figure 15.

22 Figure 15 Cross-section of the HAWC detector. The optical isolation system consists of a series of opaque curtains placed between the PMTs. With the curtains each sensor only detects light produced within its cell. This dramatically reduces the trigger rate from muons, improves the angular resolution, energy resolution, and background rejection capabilities of HAWC. A test of this system has been performed in Milagro over a 4x4 array of PMTs. We observed that the singles rates in the isolated PMTs dropped by a factor of three (in agreement with Monte Carlo simulations) and that when a PMT is struck it is much more likely to have information useful to the event reconstruction algorithms (again as expected from Monte Carlo simulations). The depth of the HAWC detector was selected as a compromise between timing resolution and gamma-hadron separation. Milagro has 2 layers of PMTs: a shallow layer at ~1.5m depth for shower plane reconstruction, and a deep layer at ~6m depth for gamma/hadron separation. In HAWC these 2 layers are combined into a single layer. For gamma/hadron separation, it is important to be able to distinguish between large and small energy depositions near the PMT, so the PMT needs to be sufficiently deep that EM particles have interacted prior to reaching the depth of the PMT and the produced Cherenkov light has diffused. This requires that the PMT depth is much greater than a radiation length in water (~40cm). The deep layer must also be much shallower than the attenuation length of Cherenkov light in water (~30m for Milagro) to maximize the detected light. Monte Carlo studies indicate that a depth of ~4m or greater is sufficient for effective gamma/hadron separation and that greater depth is not advantageous. For shower angle reconstruction the optimal PMT position is near the surface prior to the diffusion of the Cherenkov light that broadens the photon arrival time distributions which can compromise the angle fit. However, if the PMTs are too shallow, the fraction of the surface that each PMT can observe is reduced because the subtended area for the detection of vertical particles is determined by the Cherenkov angle in water, 42 O. In general, the depth of the PMTs should be slightly smaller than their spacing making it possible for a PMT to view the entire surface of its cell. Given that the depth is selected to be 4m, a separation of 5m was found to be optimal. 4.2 The Building The reservoir will be covered with a light-tight building. This is a change from Milagro which has an inflatable light-tight cover. Inflating a single cover (large enough to enclose the entire detector) for HAWC is simply not practical. The building (Figure 16) is made from prefabricated steel components that are bolted in place on site to minimize construction activities at high altitude. The roof is supported by a series of steel pillars positioned on a 15m grid. By keeping the spans between the pillars short, the size and cost of the building components is held to a minimum. The pillars act both as a support for the roof and as anchor points for the curtains. The pillars will be wrapped in polypropylene liner material to prevent direct contact between the water and steel. The roof will be sealed to prevent water leaks into

23 the building and will include an additional layer of black plastic to ensure a light-tight seal. All steel components will be galvanized or painted to prevent corrosion. The cost of the building has been determined by obtaining bids from two companies that manufacture steel buildings. Details can be found in the Budget Justification section. 4.3 Instrumentation HAWC will reuse the Hamamatsu R5912 photomultiplier tubes, the bases and encapsulations from Milagro. A single RG-59 cable provides high voltage to each PMT and carries the high frequency signal back to the front-end electronics. In HAWC, the cable will be permanently attached to the PMT housing, thus avoiding the problems that Milagro encountered with underwater connectors. The Milagro front-end electronics will be reused with minor modifications. These modules isolate and process the high frequency signals from the PMTs. The pulses are shaped and analog edges are generated at two discriminator levels, ~1/4 PE and ~5 PEs. These analog edges are subsequently digitized with TDCs. The Time-Over-Threshold (TOT) method is used to measure both pulse arrival time and amplitude with a single multi-hit TDC channel. Additionally, the front-end boards provide summed trigger signals and direct access to the analog pulses for debugging and calibration. The high voltage for the entire experiment will be provided by a single CAEN supply. Modifications will be made to the HV control system so that individual PMTs can be powered off and on independently. Currently, the failure of a single PMT deactivates 16 PMTs. The Milagro front-end electronics generate a 170ns duration square pulse for every PMT hit. Triggers are generated by discriminating an analog sum of these square wave pulses requiring a prescribed number of PMTs be hit within the trigger window. HAWC noise rate simulations indicate that an average of ~6 background hits is expected within the trigger window (based on a single hit rate of ~30kHz/PMT). The PMT singles rate has been verified by measurements with a counter system at various altitudes in Colorado. With a ~30kHz/PMT noise rate, a simple 50-PMT multiplicity trigger can be easily implemented. Though the shower plane can be reliably reconstructed for events with as few as 20 hits, for thresholds below ~50 PMTs the combinatorial background grows rapidly. This background can be reduced by shrinking the trigger window to 80-0ns or by making geometric cuts requiring that the hit PMTs are concentrated, indicating the presence of a shower core. The ability to trigger at low multiplicity, while not important for the detection of multi-tev sources (the angular resolution of these events is poor, ~1 o, and the γ/hadron separation is less effective), will greatly increase the area of HAWC below 500 GeV which increases the instrument s sensitivity to distant sources and sources with intrinsic cutoffs such as AGNs and GRBs. A smart trigger system will be built that can trigger on a reduced trigger window and hit geometry. Such a system, based upon measuring the rise-time of the trigger signal, and setting a rise-time dependent multiplicity has been successfully implemented in Milagro. Milagro uses LeCroy FASTBUS TDCs and a custom FASTBUS-VME interface for readout. The throughput of this system is limited to about ~2000Hz and ~6MB/s which is inadequate for HAWC. For HAWC we propose to use the CAEN model V1190A VME TDCs. These modules can handle the increased HAWC data rate as well as simultaneous digitization and read-out, thus eliminating the principal source of (Milagro) dead time. A number of these units have been purchased and will be evaluated in Milagro. A time stamp for each event is provided by a GPS clock that is latched by the trigger. 4.4 Online Computing and Data Processing The HAWC hardware will be read out using a single data acquisition computer. Software has been developed for Milagro to reconstruct the data in near real time by distributing it to an array (~4 computers) of client nodes. The stream of reconstructed events is then made available to online analysis programs that search the sky for transients, which can be detected

24 within 5-s of their occurrence. These clients also monitor the quality of the data, so that hardware issues can be identified and corrected immediately. The design for the Milagro online software was utilized by the IceCube experiment for their online data processing and filtering system. The HAWC online systems will be based on the Milagro design, as extended and improved by IceCube. Data will buffered locally on disk arrays. After initial filtering to remove obvious background events, the raw and reconstructed data will be transmitted via the Internet to an archive site within the United States. This operational model demands a high bandwidth network connection to the site of the experiment. Both sites under consideration have sufficient network bandwidth. HAWC will utilize the robust and flexible C++ data analysis framework developed by IceCube for both online and offline data analysis. This common framework facilitates the rapid deployment of new reconstruction and filtering techniques for online processing at the site. Reuse of the IceCube software systems will leverage the considerable software investment made in the course of IceCube construction, permitting the deployment of a robust, well-tested system with considerably less effort and cost than would be required to develop and test a new system. Computers and disk arrays will be tested for operational reliability at the ~4000m elevation. (Most commercially available computers are certified to operate up to ~3000m.) At higher elevations, CPU cooling becomes an issue. This problem can be mitigated by providing either cooler air or more air flow. Hard disks can fail at high elevations because the read head uses the pressure of the atmosphere to float over the platter. Sealed pressurized drives are readily available for high altitude operation and will be used if necessary. 4.5 Remote Operation Figure 17 Angular resolution for HAWC. Three trigger thresholds are shown. HAWC is designed to be completely remote controllable. This is critical for reliable 24/7 operation of the detector. This operational model requires reliable power and a high bandwidth internet connection. All scientific collaborators are expected to take remote shifts, monitoring and maintaining the experiment. Shift-takers are required to review the operation of the experiment and the data quality daily and respond immediately to electronic pages (text messages) indicating problems. Repairs that require travel to the site will be performed by technicians who maintain the experiment. This operational model is used by Milagro where ~95% uptime has been achieved. 4.6 Laser Calibration System The laser calibration system for HAWC is designed to provide the relative timing and pulse-height calibrations for the PMTs in the HAWC detector. The absolute energy scale will be determined using single muons passing through the detector. The system is patterned after the calibration system in Milagro. Because of the shielding effect of the curtains, the number of diffusing light sources increases from 30 (Milagro) to ~225 (HAWC). In HAWC the diffusing light sources will be positioned ~2.75m above the pond and supported from the roof structure. The proposed light source design builds on the present Milagro calibration system including the JDS Uniphase PowerChip NanoLaser operating at 532nm. The ~1ns light pulses pass through a variable attenuator to allow control of the light intensity over 4 orders of magnitude. The laser beam will be directed through a series of optical fan-outs to illuminate ¼ of the PMTs at any given time.

25 5 Detector Performance The simulation for HAWC is an extension of the Milagro simulation software package. CORSIKA 52 is used to simulate gamma ray and hadron induced atmospheric showers. A custom detector simulation using GEANT 53 is used to propagate the secondary shower particles through the HAWC detector. Cherenkov light production is simulated and individual Cherenkov photons are tracked through the detector. Detailed optical modeling of the water (absorption and scattering), reflection and absorption at surfaces, and the PMT response are included. The simulation has been thoroughly tested through comparison with Milagro data. Gamma ray and background rates are scaled from measured values in Milagro by comparing the predictions of the HAWC and Milagro simulations. By doing this, we not only remove potential systematic errors internal to the simulation from the air shower modeling, optical model, and detection efficiency, but also remove systematic errors in the measurement of gamma-ray fluxes and hadronic backgrounds provided by other experiments. 5.1 Triggering and Noise Rates The trigger rate in HAWC will be ~5 khz for a 50 PMT threshold, about 3 times higher than the Milagro trigger rate. This rate increase necessitates the upgrade of the multi-hit TDCs. However, the data rate in HAWC is only about 1½ times greater than Milagro because the average event multiplicity is smaller due to the absence of the second layer of PMTs. On average the deep PMTs are hit twice as often as the shallow PMTs in Milagro. 5.2 Angle and Core Reconstruction The position of the shower core on the ground is determined by fitting the distribution of the pulse amplitudes to a Gaussian profile. After the core is located, the PMT hit times are adjusted to account for the curvature of the shower front. Typically, the shower front curvature correction is ~0.5 o -1.0 o, so misidentification of the core position leads to poor angular resolution. The corrected PMT hit times are then fit to a plane to determine the incoming shower angle. The width of the distribution of timing residuals (from the fit shower plane) ranges from 1-3 ns depending on the pulse amplitude. The angular resolution depends on the event size. Figure 17 shows the angular resolution of the HAWC detector for three different ranges of event size. 5.3 Gamma /Hadron Separation Hadronic showers are identified through the pattern of energy deposition in the detector. While gamma-ray induced showers have compact cores with smoothly falling density, hadronic showers typically deposit large amounts of energy in distinct clumps far from the shower core. This is mostly due to the presence of hadrons and muons in hadronic showers. Figure 18 shows the distribution of deposited energy in HAWC for gamma ray and hadron induced showers. As a simple γ/hadron discriminator, we have extended the compactness parameter, C, developed for Milagro 54. Here C is defined as the total number of PMTs hit with amplitudes greater than 2 PEs divided by the largest pulse amplitude that is more than 30m from the reconstructed core position. Gamma-ray induced showers, with only small hits far from the core, have large values of C. Hadron induced showers with multiple muons and hadrons have low values for C. Figure 19 shows Compactness distribution for gamma ray and proton triggers for small (nhit>50) and large (nhit>200) events. The same figure shows the Q (increase in sensitivity) as a function of the cut level. The efficiency for retaining gamma ray induced events increases with increasing energy, while it drops for proton-induced events. Therefore, the background rejection capability of HAWC improves with increasing energy. These predictions have been verified with Milagro data on the Crab Nebula.

26 P001: E= GeV, Th = 21.2 deg, C= 2.8 P002: E= GeV, Th = 15.3 deg, C= 0.5 P003: E= GeV, Th = 20.9 deg, C= 1.3 P004: E= GeV, Th = 14.0 deg, C= 0.5 Y 15 Y 15 Y 15 Y X X X X G001: E= GeV, Th = 28.1 deg, C=11.4 G002: E= GeV, Th = 25.9 deg, C=.8 G003: E= GeV, Th = 21.1 deg, C= 7.0 G004: E= GeV, Th = 18.7 deg, C=12.4 Y 15 Y 15 Y 15 Y X X X X Figure 18 Illustration of hadron and gamma-ray induced events in HAWC. The top four panels show typical proton induced events and the bottom four panels show simulated gamma-ray events. The hit amplitudes in the 30x30 PMT array are indicated by the color scale. The large black circle indicates the position of the best fit core and the radius of the circle defines the exclusion region for the compactness cut. The smaller circles indicate the positions of EM particles (red) and muons and hadrons (blue) with energies greater than 1 GeV. In gamma-ray induced showers, the >1GeV EM particles are typically confined to the shower core, but proton induced showers are easily distinguished by their numerous high energy particles detected far from the fit core. All of the simulated proton events shown here would fail a typical g/hadron cut (C>6-8). The area enclosed by the circle around the shower core is roughly the same as the area of the bottom layer of the Milagro detector. Figure 19 Compactness distribution for small (left) and large (right) events. Defining a quality factor, Q, as the fraction of retained gammas divided by the square root of the fraction of retained hadrons, the Q value for a cut accepting all events with greater than a given compactness is shown in green. For hard cuts, C>8, with large events, the efficiency for the hadronic background is poorly measured because of the paucity of simulated hadronic events passing the cuts. Q factors as large as 5 are reliably predicted for large events. It is important to point out that despite simulating billions of air showers, the available statistics in our current simulated data set at high energies are insufficient to reliably predict the background rejection (i.e. no simulated proton events with energies above ~ TeV survive the cut criteria). HAWC may be capable of rejecting nearly all hadronic background above 5- TeV, but at this time we are unable to simulate enough high energy background events to demonstrate this. The γ/hadron capability demonstrated here should be regarded as conservative. Further study will likely reveal substantial improvement, especially at large energies.

27 Gamma Area: θ<30 o,ntop/cxpe>5.0, Δθ<1.0 O 200PMT Trigger 80 PMT Trigger 20 PMT Trigger Pond Area Figure 20 Effective Area (after passing γ/hadron and angular resolution cuts for three different The energy resolution of surface arrays is limited by shower fluctuations in the atmosphere, as only the tails of EM showers are detected. When the effective area is less than the physical detector area, fluctuations in the shower development begin to dominate the response of the detector. It is impossible to distinguish between a low energy gamma ray that interacted deep in the atmosphere and a higher energy gamma ray that interacted higher in the atmosphere. Therefore the energy resolution of HAWC is strongly dependent upon the primary gamma ray energy. Showers with energies near or above the median, can be reconstructed with ~30-40% resolution. Figure 13 of the section demonstrates the ability of Milagro to measure of the Crab spectrum. (With a median energy of ~4 TeV, Milagro can reconstruct primary gamma ray energies above ~1 TeV). 5.4 Sky Survey Sensitivity Because the sensitivity of the detector is strongly dependent on the zenith angle of the source being studied, we compute the sensitivity by estimation of the number of signal and background events collected during a single transit of the source from horizon to horizon. As a reference source, the Crab is selected. The differential spectrum of the Crab is assumed to be (2.8 x -11 ) E particles/s/cm 2 /TeV 55, but since the results are normalized to the Milagro measurements, the sensitivity is nearly independent of the assumed source spectrum. The background rate is computed by dividing the simulated background rate for HAWC by the simulated background rate for Milagro and multiplying the ratio by the measured background rate in Milagro. The signal rates are similarly scaled. The transit declination of the source is 22 o and the detector is assumed to be located at a latitude of 37 o (the location of Milagro), so the minimum zenith angle for the source is 15 o. For a source transiting through the detector zenith the sensitivity is ~25% higher and for a source transiting only within 25 o of zenith, the sensitivity would be ~25% lower. So the numbers quoted here represent a typical sensitivity for a sky survey of ±25 o declination. For this calculation, the detector altitude assumed to be 4150m asl, the average altitude of the sites under consideration. Figure 21 Energy of gamma-ray events for a Crab-like spectrum and different multiplicities in HAWC.

28 Table 3 A comparison of the sensitivities of HAWC and Milagro. The event rates for Milagro are based on observations of the Crab. The rates for HAWC are based on simulations, scaled to the Milagro simulations and observations. Configuration Gamma Rate (events/day) Bkg. Rate (events/day) Optimal Bin Radius (deg) Significance (1 day) Std. Dev. Significance (1 year) Std. Dev. g/hadron Quality Factor Median Energy Milagro: std cuts Milagro: hard cuts Milagro: Likelihood HAWC: nhit>=50, no γ/h cuts HAWC: nhit>=50, C>8.0 HAWC: nhit>=0, C>8.0 HAWC: nhit>=150,c>8.0 HAWC: nhit>=200c>8.0 HAWC: Likelihood TeV TeV TeV TeV TeV TeV TeV TeV >~5.0 >~0-4 TeV Table 3 shows the sensitivity of Milagro and HAWC for various values of a simple cut on Compactness parameter. The sensitivity of HAWC and Milagro increase with energy due to the improvement of the background rejection and the angular resolution. In HAWC, for hard cuts, the simulated data set is too small to reliably predict the background rates. Also, the simple compactness cut is probably not optimal for large events. In these events HAWC will measure the lateral distribution of deposited energy, which should improve the hadron rejection. For Milagro we have developed a likelihood analysis methodology that weights each event by the ratio of the probability that it is a gamma ray to the probability that it is a proton. This results in a ~60% improvement in sensitivity compared to making a simple cut on the value of the compactness parameter. We have applied the same technique to the HAWC data. This results in a sensitivity such that the Crab Nebula is detected at ~6σ/transit. However, because of the difficulty in generating sufficient Monte Carlo statistics of the background at high energy, we quote a conservative estimate of the sensitivity to be 5σ/day or 0σ/year. For a 1 year survey of the overhead sky, HAWC will have a 5σ point source detection threshold of ~50 mcrab. In a single year, HAWC will survey the entire sky with the sensitivity comparable to the southern Galactic plane survey performed by HESS and the northern survey planned by VERITAS. While the Crab Nebula is commonly used as a reference source in TeV astronomy, this can be misleading due to the its relatively soft spectrum (E -2.6 ) compared to other Galactic sources (E -2.2 ). In general, when comparing the sensitivity of two instruments, the one with the lower energy threshold is relatively more sensitive to softer spectra and less sensitive to harder spectra. Figure 1 illustrates this point by comparing the sensitivity of HESS to HAWC. While HESS is considerably more sensitive for soft spectrum sources, HAWC and HESS have comparable sensitivity for hard spectrum sources. For hard spectrum sources, such as those Galactic sources detected by HESS, HAWC will have a 5σ detection sensitivity of mcrab.

29 6 Proposed Research 6.1 HAWC Construction While HAWC will be a next generation detector with unprecedented wide-field sensitivity, because we are proposing to simply move and reconfigure the Milagro PMTs and electronics at a new high altitude site, it can be built relatively quickly and at low cost. We propose to complete the project within four years of the start of funding. In the table below we describe the tasks for these four project years (PY1-4). We also include PY 0 which are tasks that we will accomplish between the time of writing this proposal and the beginning of funding in spring In PY 0 we will complete the site selection process described in this section. During PY 1 we will design the pond and building/cover and support facilities that will go at the selected site. At the end of PY 1 we will have a detailed construction plan that should be reviewed by the funding agencies. We will also have a clear idea of the local (foreign) support that will be available. At that time a firm budget with bottom-up contingency can be reviewed (rather than the top-down 20% number we are using here). During PY 0, we will study and prototype possible modifications to the electronics. We will consider the most cost effective way to improve the timing resolution and thus angular resolution. We will investigate the addition of constant fraction discriminators and whether part of the long signal transport needed for HAWC should be over optical fiber. We will also study the implementation of relay control of individual HV channels to improve the on-time of HAWC. The selected Activity PY PY PY PY PY modifications will be incorporated into the running Milagro experiment for a number of channels during PY Milagro operations HAWC will be built by moving the Milagro PMTs and electronics to the new site. As such, much of the development work on new software, digitization and readout (DAQ) can be done on Milagro. Our proposal is to continue to operate Milagro as both a development site and for physics until PY2 when the construction of the new pond is underway. Then, the PMTs will be removed and refurbished. As a Data gathering and site selection Detail design for selected site DAQ development Pond construction infrastructure power, water, comm. Building Construction Turn off Milagro Refurbish PMTs Installation of liner Install water system Install counting house Install PMT s & DAQ Start filling with water Engineering running result, PY 1 contains funding for the continued operation of Milagro. The design and contracting for the new site will take less than one year and can proceed in parallel with the modifications of the electronics and the design of the software. The construction of the reservoir and building should take less than 18 months once funds are available. Removal and refurbishment of the Milagro PMTs can be done in parallel with the site construction. Once the building and reservoir are complete, the electronics, cabling, and PMTs can be installed and tested within 3 months and water filling can begin. Engineering runs can begin as soon as the PMTs are underwater.

30 6.3 Site Selection There are currently two sites under serious consideration by the collaboration for the construction of HAWC: the Sierra Negra site in Mexico and the Yangbajing (YBJ) Cosmic Ray Observatory in Tibet, China. Both sites are available and appear to be excellent choices. Both have sufficient land above 4km permitted for scientific use, neither are green sites as both have existing major observatories there already. In addition, both sites have strong local university groups with experienced scientists eager to collaborate with us on HAWC. Both of these groups have also made strong proposals (detailed below) to support HAWC at their local site. Still at the time of writing this proposal we are carefully evaluating both sites. Since the timescale for NSF approval of this proposal is such that we plan to spend what is referred to as Project Year 0 finalizing our site selection. In Appendix 1 we describe each site in detail including the local infrastructure. We also provide a list the local collaborators including a brief description of their background. In this section below we will explain the criteria and method that will be applied for the final site selection. The primary purpose of the site selection criteria is to mitigate the risk to the project. The primary identified risks are the availability of land (including permission to utilize the land for the intended purpose), the availability of water, power, and communications, the strength and contributions of the local collaboration (including both scientific and technical expertise), the total cost of the project (including site specific infrastructure costs and local financial contributions), and the ease of access for the U.S. portion of the collaboration. Secondary considerations affect possible impacts to the scientific return of the project. These include the altitude of the observatory, the rigidity cutoff (and its affect on the expected background rates), the possibility of expansion (for the deployment of a possible outrigger array to increase the sensitive energy range of the instrument), and the longitude and latitude of the site (which impacts the overlap with experiments such as IceCube and VERITAS). Given the relative similarities between the sites, these impacts are expected to be small. A site selection subcommittee has been formed and has developed a spreadsheet with the relevant requirements. The committee is in the process of determining authoritative answers to the remaining open questions. Once this process is complete the relevant information will be presented to the collaboration and following a discussion a binding vote will be taken amongst the active collaborators. The final site selection will be determined by a 2/3- majority vote Foreign contributions At both sites the local groups are making proposals to their respective funding agencies for support of HAWC. In Tibet the IHEP group is making a proposal to the Chinese NSF for support to HAWC at the level of $1M USD. This is in addition to the funds for the land use, power, water, networking and other local infrastructure. In addition, they already have substantial facilities at YBJ that would support HAWC. Professor Hesheng Chen, the Director of the Institute of High Energy Physics in Beijing, has written a letter (appended to the end of this proposal) assuring us that they will provide the necessary land if we decide to build HAWC in YangBaJing. In Mexico, a collaboration of several groups including two from Puebla (INOUE and BUAP) plus UNAM in Mexico City have submitted a proposal for $5M USD to build the HAWC pond, building and local infrastructure at the Sierra Negra site. The disposition of both of these proposals is contingent upon US support of this proposal Project Management The HAWC project will be managed in a manner similar to Milagro. There will be an institutional board consisting of a representative from each institution, including foreign collaborators. The two spokespeople will be elected to two-year staggered terms. Financial

31 responsibility for the project will be divided by funding source. All NSF funds for construction and operation will be managed at Maryland by the PI. Subcontracts will be awarded by Maryland to collaborating institutions for specific tasks. (This has successfully provided the ability to move resources where needed in Milagro). Institutional responsibilities are detailed in the budget justification. DoE funds will be managed by Los Alamos. Foreign contributions will be managed by the lead institution in the host country. They will be responsible for local regulations and site management. 6.4 Primary Responsibilities of US Institutions on HAWC University of Maryland Project Management Data Management Los Alamos National Laboratory Construction Management Infrastructure Design Data Acquisition University of Utah Electronics Development (with MSU) Michigan State University Electronics Development (with Utah) University of New Mexico Calibration System Penn State University Online Software, Simulations University of New Hampshire Solar Physics NASA Goddard GLAST Coordination 7 Education and Outreach Particle astrophysics is a field that excites the imagination of all people, from the general public, to undergraduate and graduate students. In Milagro, we have graduated 15 PhD students on Milagro. Most of these students spent a substantial fraction of their time at Los Alamos getting a rare opportunity to partake in the design and construction of a major experiment. Over the lifetime of Milagro we have hosted over 31 undergraduate students over summer programs at Los Alamos, in both the construction of Milagro and the analysis of physics data. Through a LANL intern program and other funds we have employed several high-school students from Los Alamos, the rural community of La Cueva, and the Jemez Indian Pueblo. Each summer a group of students (EARTHWATCH) stay at Fenton Hill and are given tours of Milagro with an accompanying lecture. In addition 40 other undergraduate students have worked on various aspects of Milagro at their home institutions. Many of the undergraduate students have gone on to graduate school in physics. In fact one eventual Milagro PhD student obtained his first experience in experimental physics in the TRAC teacher program when he came to work with us for a summer on Milagro. The PI, Jordan Goodman gives, on average, eight talks each year at high schools and at least two public lectures on the subject of particle Astrophysics (PA). Last year he gave a public lecture as part of a Smithsonian course for the World year of Physics. He also works with the US Physics Olympics team. Each year for the past 20 he has given talks to the US team on PA. In addition, he works with the Maryland program for 8 th grade girls (see testimonial letters in the appendix). He also started the physics Question of the Week (see it at that is used by many teachers world-wide to expose students to interesting problems in Physics.

32 Milagro was one of the first Particle Astrophysics experiment to join Quarknet a program to bring active physics experiments into the classroom by involving classroom teachers in research. This was our fifth year and we have teachers setting up cosmic ray counters in their classrooms. Dr. Goodman also ran a course for high school teachers entitled What the Universe is Made Of. In Fall 2006, one of Maryland s top undergraduates (and Goldwater Scholarship recipient) is doing an independent research project on Milagro. Particle astrophysics students and faculty at UNM help run star gazing parties at our campus observatory. This program is open to all members of the community and takes place every Friday night during the spring and fall academic terms. Visitors not only "see the stars" but also have the opportunity to discuss astronomy and astrophysics topics of personal interest "with the experts". Milagro students and post-docs go on to many of the other particle astrophysics experiments. The list of these includes: VERITAS, HESS, GLAST, Swift, Auger, IceCube, HiRes, CELESTE, LIGO. We also have gotten post-docs from a number of other experiments including: Super-K, Whipple/VERITAS, HEGRA, HEP/NP experiments, Amanda/IceCube, Tibet, CGRO, Haleakala. The Utah VERITAS group has sponsored summer internships for undergraduate students to work at the Whipple Observatory on detector construction, commissioning, observations and analysis. Students spend 1 month of the internship at the University of Utah acquiring basic electronics and experimental skills, and then spend 12 weeks in Tucson during the internship. While in Tucson, the students rotated through observation shift with the Whipple -m telescope, and participated in the construction, commissioning and operation of the VERITAS -4 telescopes. Weekly lectures on topics in astronomy and high-energy astrophysics were included as part of this internship, as well visits to other National Observatories (Kitt Peak, MMT, Mt Graham). The internship attracts a large number of underrepresented minority students (approximately 50% of the internships are awarded to women). Particle and astroparticle physics faculty at Penn State are developing a professional development course on particle astrophysics for secondary school teachers, to be included in PSU s successful Science Workshops for Educators program. These summer workshops fulfill Pennsylvania continuing education requirements, and attract teachers from rural districts and those serving underrepresented minority groups. Even though HAWC will be built in either Mexico or Tibet, we will still involve undergraduates and graduate students as we did with Super-K in Japan. In addition, we will work with the local groups to do outreach in the region near the site. Magdalena Gonzalez from UNAM (a Milagro and HAWC collaborator who will work on HAWC at either site) is developing outreach materials in Spanish for use in both Mexico and the US. Our goal is to reach out to not only local populations near the site, but to target Hispanic students in the US. 8 Proposal Summary Gamma-ray astrophysics is a scientifically rich and growing field. A wide field of view TeV gamma-ray observatory will have a dramatic impact on the scientific return of this field. HAWC will have sufficient sensitivity to discover new sources. Prime candidates are HESS-like galactic sources of larger angular extent and GLAST sources that extend to higher energies. HAWC will have sufficient sensitivity to monitor sources for transient emission and alert the community. This will dramatically increase the number of multi-wavelength and multi-messenger observations. HAWC will have excellent sensitivity at the highest energies. Combined with the long observation times we will extend measured spectra to the highest energies - required to observe hadronic sources. HAWC is built on the heritage of the Milagro experiment. The Milagro experiment has proven that new sources can be discovered with this technique and that the sensitivity of

33 a water Cherenkov telescope can be accurately simulated. The high duty cycle, ~95%, and remote operation of Milagro are evidence that the HAWC observatory can be operated efficiently. HAWC will have a broad impact of the science in this multidisciplinary field and will follow in the tradition of Milagro to produce outstanding scientists and public outreach activities. The overlapping fields of view of HAWC, GLAST, and IceCube will enhance the scientific understanding of any of the sources detected by any of the three observatories. The different energy ranges of HAWC from 0 GeV-0 TeV and GLAST from 30 MeV-300 GeV are complimentary, and the energy range of IceCube matches that of HAWC to improve the likelihood of multi-messenger observations. The large field of view of HAWC will be used to find new sources, which imaging atmospheric Cherenkov telescopes and x-ray telescopes can further investigate. HAWC will follow the tradition of Milagro, which has trained 72 undergrads, 12 Ph.D. students and 12 postdocs that have populated the fields of high energy physics, astronomy, defense, and education. The HAWC collaboration has scientists actively involved in public outreach who will continue and expand upon these activities.

34 References 1 Revealing the Hidden Nature of Space and Time: Charting the Course for Elementary Particle Physics, 2 Connecting Quarks with the Cosmos: Eleven Science Questions for the New Century, 3 A 21st Century Frontier for Discovery: The Physics of the Universe, 4 Decade of Discovery in Astronomy and Astrophysics (2001), 5 Aharonian, F. et al., 2005, Science, 307 (5717), Hunter, S.D. et al. 1997, ApJ, 481, de Boer, W.; et al. 2005, A&A, 444, Strong, Andrew W.; Moskalenko, Igor V.; Reimer, Olaf 2004, ApJ, 613, Torres, Diego F.; Dame, T. M.; Digel, Seth W. 2005, ApJ Lett, 621, pp. L29-L Koushiappas, S. M.; Zentner, A. R.; Walker, T. P. 2004, Physical Review D, 69 (4), F. Miniati, 2003 MNRAS, 342, 09 & Inoue, S., Aharonian, F., Sugiyama, Naoshi 2005, ApJ Lett, 628, 9I. 13 R. C. Berrington. & C. D. Dermer 2003, ApJ, 594, Blom, Jan J.; Paglione, Timothy A. D.; Carramiñana, Alberto, 1999, ApJ, 516, Aharonian, F. A.; Coppi, P. S.; Voelk, H. J.,1994, ApJ Lett, 423, L5-L8 16 McLaughlin, M.A., Mattox, J.R., Cordes, J.M., Thompson, D.J., 1996, ApJ 473, Hurley, et al., 1994, Nature, 372, Gonzalez, M.M. et al., 2003, Nature 2003, 424, Razzaque, S. & Mészáros, P. 2006, astro-ph/ Zhang & Mészáros, 2001, ApJ, 559,1. 21 Böttcher & Dermer, 1998 ApJ, 499, L Kneiske, T. M.; Bretz, T.; Mannheim, K.; Hartmann, D. H. 2004, A&A, 413, Torres, D. et al A&A 370, Falcone, A. et al., 2003 ApJ, 588, Aharonian, et al. 2005, A & A, 442, Hartmann, R.C. et al. 1999, ApJ Supp, 123, Chen, A., Reyes, L.C., Ritz, S., 2004, ApJ, 608, Atkins, R. et al Phys Rev Lett 95, Abdo, A. et al. 2006, in preparation. 30 Smith, A. et al. 2006, Review, Rapporteur, & Highlight Papers of 29th ICRC in Pune India,, Atkins, R. et al. 2000, ApJ Lett, 533, 119 & Atkins, R. et al. 2003, ApJ, 583, Atkins, R. et al. 1999, ApJ Lett, 525, L Atkins, R. et al. 2004, ApJ, 608, Atkins, R. et al. 2003, ApJ, 595, Atkins, R. et al. 2005, ApJ, 630, 996 & Atkins, R. et al., 2004 ApJ Lett, 604, Atkins, R. et al. 2004, ApJ, 608, Atkins, R. et al. 2004, ApJ, 608, Atkins, R. et al Phys Rev Lett 95, van der Hucht, K. A., 2001, New Astron. Rev. 45, Bochkarev, N. G. and Sitnik, T. G., 1985, Astrophys. Space Sci. 8, Green, D. A., 2004, Bull. Astron. Soc. India 32, Roberts, M. S. E., et al., 2002, ApJ. 577, L Strong, A. W., Moskalenko, I.V., and Reimer, O., 2004, ApJ, 613, Amenomori, M., et al., 1996, Astron & Astrophys., 311, 919.

35 45 Padilla, L., et al., 1998, Astron. & Astrophys., 337, Atkins, R., et al., 2000, ApJ Lett, 533, L Pe er, A. and Waxman, E., 2004, ApJ Lett, 603, L1. 48 Sinnis, G., Smith, A., and McEnery, J.E., 2004, astro-ph/ Amenomori, M. et al., 2005, ApJ 633, di Sciascio, G. et al., 2005, High Energy Gamma-Ray Astronomy: 2nd Int. Symp. in Heidelberg, eds. F. A. Aharonian, H. J. Völk, and D. Horns. AIP Conf. Proc, Cao, Z. et al., 2005, 29th International Cosmic Ray Conference, Pune 5, Atkins, R. et al. 2003, ApJ, 595, Aharonian, F. et al.,2004, ApJ, 614, 897.

36 Budget Justification The budget presented here can be broken down into two phases project year 1 (PY1), where we will complete the site-specific design of the pond, building/cover and support facilities plus do development work on DAQ and electronics using the Milagro detector and PYs 2-4 where we will actually build the HAWC detector. The budget presented here is for construction of the complete HAWC detector even though some of the cost will be born by the foreign host country. At the end of PY 1, we will have a detailed construction design and cost that should be reviewed by the funding agencies. We will also have actual commitments of local (foreign) support at that time. In the PY1 budget there are funds that will allow us to continue operate Milagro both for HAWC development and for physics until construction on the new detector begins. The continued operation of Milagro is important for two reasons: 1. We will use Milagro to prototype HAWC subsystems: in particular, the electronics and data acquisition system. This will result in a substantially faster start-up time for HAWC, especially since HAWC will be built at a remote site. 2. The physics reach of Milagro is now substantially higher than in the past (because of the outrigger array) and we feel that it is important to have a widefield instrument operating during the initial operation of GLAST and VERITAS. The major costs associated with operating Milagro are continued support of the LANL group and continued support of 1.5 technicians at UCI and UCSC. In return for the $250k NSF provides to LANL for operations, LANL provides funding to support ¾ of the co-spokesman (Sinnis) and ¾ of the Run Manager (Dingus). LANL also provides infrastructure support for Milagro: electricity, supplies, office space for collaborators and in residence students and postdocs, a Government vehicle for use by the collaboration, computer support, personnel to maintain the infrastructure at the Milagro site, support for ½ a postdoc. The funding for UCSC and UCI is to provide the support for two technicians. The UCI technician, Scott DeLay, is full time in Los Alamos and is essential for day to day Milagro operations. The UCSC technician is split between Milagro operations and HAWC development. His efforts on the building design, air conditioning requirements, and other technical details have been critical in developing this proposal and we will require his services even more in this coming year. The additional costs associated with the first year budget are for the new groups that are not currently supported to work on Milagro. This includes technician support at the University of Utah, Michigan State University (jointly responsible for the electronics development) and a halftime post-doc at Penn State (the other half is supported by IceCube). The cost for the final design of the reservoir and building represents an estimate from two engineering firms for these tasks. The HAWC construction budget is given shown below and listed under other in the budget. This cost does not include contingency. At this time we can only give a topdown estimate of the contingency of ~20%. Part of the design in the first year budget will

37 be used to perform a bottom-up evaluation of the contingency. Based on our work to date we expect the total Item Unit Cost Units Quantity Sub Total Totals Total 5.24E+06 Site Preparation 655,000 Water 600,000 ea 1 600,000 Power 50,000 ea 1 50,000 Communications 5,000 ea 1 5,000 Land Use 0 ea 1 0 Pond Preparation 3.49E+06 Reservoir 5 m 3 120, ,000 Concrete 0 yd ,000 Rebar,000 ea 1,000 Building Materials 1,350,000 ea 1 1,350,000 Building Labor 305,000 ea 1 305,000 Galvanization/painting 150,000 ea 1 150,000 Liner 600,000 ea 1 600,000 Curtains (Materials) 3,500 4mx150m ,000 Curtains (Steel) 9,000 m 6 54,000 Curtains (Labor) 50,000 ea 1 50,000 Lightning System 50,000 ea 1 50,000 Utility Build. 50,000 ea 1 50,000 PMTs 87,000 Kevlar String 1 meters 2,000 2,000 SS parts ea 2,000 20,000 PMT anchors 50 ea ,000 Refurbish 20 all 1,000 20,000 Electronics 632,200 Cables 1 meters 300, ,000 Connectors SHV 60 ea 1,000 60,000 HV Supply 35, ea 1 35,000 Front elect. Mod ea ,000 VME TDCs, ea 8 80,000 Computers 4,000 ea 40,000 Disk Storage 9,000 8TB 3 27,000 Calibration System 78,000 Laser Fiber 0.4 meters 45,000 18,000 Laser 20, ea 1 20,000 Monitoring 20, ea 1 20,000

38 Optics 20, ea 1 20,000 Water System 50,000 Additional Capacity 50, ea 1 50,000 Facility 74,520 AC Units, ton unit 3 30,000 Water Chiller, ea 1,000 Fans 45 ea ,520 Fan housing 1, ea 8 8,000 Racks 1,000 ea 8 8,000 Networking 2,000 ea 1 2,000 Environmental Mon. 5,000 ea 1 5,000 Freight 175,000 Building 150,000 ea 1 150,000 Water system 5, C-tainer 2,000 Freight 5, C-tainer 1 5,000 PMTs 5,000 C-tainer 2,000 Table 1. HAWC construction budget. project construction cost to be approximately $5.24M. By far the bulk of the costs associated with HAWC are the costs associated with the reservoir itself: the cost of the actual water reservoir, the liner, and the building that encloses the reservoir. These costs account for nearly 70% of the project cost and we expect these items to have the largest contingency. We examine each cost in detail below. The cost estimate for building the reservoir is the amount bid by a construction company in Tibet, China, for digging the reservoir, and for providing the sub-reservoir draining system, the gravel, and the sand layer that are required for a true water reservoir. During the upcoming year we will work with Boyle Engineering (the company that designed the Milagro reservoir) of Albuquerque, NM to develop a detailed reservoir design. This requires geologic and seismic studies of the actual site. Parameters such as the slope of the sides and the amount of material that can be built up as opposed to dug out are dependent upon these measurements. Since the final site has not been selected we did not feel it cost effective to proceed with this evaluation work at this time. The estimate for the liner is $600K. We conservatively chose the higher of two independent estimates. It was from the Layfield Corporation, which designed, built, and installed the current Milagro liner. The lower estimate of $475K was obtained from Field Lining Systems Inc. which provided the curtain material for the curtain tests in Milagro. Both bids include the cost of a geotextile layer that goes under the liner itself. The cost difference between the two bids is consistent with the different thickness each used for the liner material and since we will need to be able to walk and work on the liner during installation we chose the thicker liner. Perhaps the biggest unknown for us (due to lack of experience) is the cost of the building itself. We seriously considered an inflatable cover system like the one currently used for Milagro and are convinced that a building is a far better choice. A building

39 poses many fewer performance risks than an inflatable cover has no known cost disadvantages. There are serious disadvantages to a membrane cover. An inflatable cover that can be inflated in a reasonable time, ~2 hours, would require extensive engineering design work and poses uncertainties in cost and performance. Commercially available systems (for example tennis bubbles) operate at much higher pressure than the current Milagro design, requiring highstrength materials with a substantially higher cost than the current Milagro cover material. The requirement that the cover must be light-tight may be in conflict with such a high-pressure system. We have investigated alternate methods utilizing novel designs (inflating large tubes with air to push the cover up). While such a system could probably be made to work, the possible cost savings appears to be minimal and the added risk was not deemed to be worth the possible modest cost savings. For any system with an inflatable membrane cover there is a large additional cost associated with a lightning protection system, similar to that built for Milagro. It is the sum of the costs for the inflatable membrane system and the LPS that makes a building attractive from a financial point of view. The advantages of a building however are many. A straightforward construction project. A building, even of this size, can be built with standard construction techniques. The ease of access enabling repair at essentially any time. Currently in Milagro the repair scenario is a large undertaking that it occurs only once per year: meaning that we run for a significant fraction of the time with non-working channels. Plus a significant fraction of the ~5% downtime for Milagro was associated with repairs. The building will reduce this significantly. Calibration of PMTs. The timing calibration of the system is critical to the success of HAWC. The relative times of all PMTs must be calibrated to within 1 ns. This requires that each PMT can see multiple light sources and each light source is visible to several PMTs (at a minimum). With the addition of curtains this becomes difficult if one cannot place the light sources above the water. A building makes this a straightforward. Control over the water air interface. For solar studies and stability of the detector itself it is important to control the water air interface. If the water is in contact with a material with a similar index of refraction the formation of an ice layer, effectively places an air layer between the water and the cover. This leads to total internal reflection of some of the light generated inside the water volume. This has a large affect on the singles rates of the PMTs and requires one to increase the trigger threshold of the detector to keep the trigger rate at a level that the data acquisition system can tolerate. The cost estimate for the building is the result of two independent bids, one from Braemar Building Systems and the other from the Kirby Building Systems. Both quotes only include the cost of the building materials. The construction cost estimates were provided by these vendors as typical for buildings of this size constructed within the U.S. We expect the labor costs to be lower at either of the sites under consideration. The actual building design will have to interface with the reservoir, the curtain system, and the liner. This design will be carried out during the first year once the site selection process is complete.

40 The budget given for HAWC construction is the estimated total construction budget for the site in Mexico. We expect either host country to contribute significant funds towards the construction of HAWC. In China, we already have an assurance from the IHEP to provide all of the Site Preparation costs for the experiment producing a cost savings of ~$650k. The bulk of this cost is the cost of obtaining water at the Mexico site. (In Tibet the costs associated with this line will be substantially smaller). The estimate in budget assumes that a well will be dug in a nearby village and a pipeline constructed to the site. Recent geologic studies have indicated the presence of a nearby underground stream. If this or other sources of water are found, the cost of this item could go down by nearly an order of magnitude. In addition to this contribution both host countries will submit proposals to their respective funding agencies. The Chinese collaboration would submit a proposal for roughly $1M. The Mexican collaboration has already submitted a proposal for $4-5M. In either case we expect the actual cost to the U.S. to be smaller than the number given here even if some of the costs escalate from our current estimates. It should be noted that in addition to the 900 PMTs from Milagro, a significant amount of other equipment will be reused (e.g. lasers, hv supplies, etc) so they are not included in this budget.

41 FACILITIES, EQUIPMENT & OTHER RESOURCES FACILITIES: Identify the facilities to be used at each performance site listed and, as appropriate, indicate their capacities, pertinent capabilities, relative proximity, and extent of availability to the project. Use "Other" to describe the facilities at any other performance sites listed and at sites for field studies. USE additional pages as necessary. Laboratory: Milagro site at Los Alamos National Lab The final site selection has been narrowed down to two possibilities--yangbajing, China and Sierra Negra, Puebla, Mexico Clinical: Animal: Computer: Office: Other: MAJOR EQUIPMENT: List the most important items available for this project and, as appropriate identifying the location and pertinent capabilities of each. 900 Milagro PMTs The Milagro electronics 0 TB of data storage OTHER RESOURCES: Provide any information describing the other resources available for the project. Identify support services such as consultant, secretarial, machine shop, and electronics shop, and the extent to which they will be available for the project. Include an explanation of any consortium/contractual arrangements with other organizations.

42 Appendix: Site Information The final site selection has been narrowed down to two possibilities-- Yangbajing, China and Sierra Negra, Puebla-México. We considered sites in the United States -- the University of California s White Mountain Research Station and Mt. Evans which is the location of the University of Denver s Meyer-Womble Observatory. While these two sites are at high altitude, the weather in the winter is severe and would restrict access. And, probably more important, both are environmentally protected areas and it would be difficult to build such a large project at these sites or any other sites in the United States. We also researched sites in South American and visited a possible site near the Mt. Chacaltaya Cosmic Ray Observatory which is close to La Paz, Bolivia. These southern hemisphere sites are even higher altitude and allow better observations of galactic sources; however, no sites with current infrastructure and a large local group of scientists were found. We feel that the budget of HAWC is too small to be able to develop these capabilities, and so we have eliminated these southern sites from consideration. Yangbajing Cosmic Ray Observatory, Tibet The Yangbajing (YBJ) International Cosmic Ray Observatory (30.11 o N, o E, altitude=4300 m a.s.l.), Tibet, China was founded in Presently there are two major gamma-ray observatories operating at the site: the ASγ and ARGO detectors. ASγ is a large traditional scintillator array operated by a Japanese and Chinese collaboration. This detector has been operating (in different configurations) nearly continuously since the founding of the YBJ Observatory. The ARGO detector (Figure 1) consists of a large (~6500m 2 ) array of resistive plate chambers (RPC) used to record the passage of charged particles. This detector was recently commissioned and is a Chinese Italian collaboration. In addition, there are a number of smaller scale experiments operating at YBJ: a neutron monitor and a neutron telescope for solar and heliospheric studies. For both experiments the Chinese collaboration is headed by physicists from the Institute for High-Energy Physics (IHEP) in Beijing. Physicists from Tibet University handle local interactions. Figure 1. The ARGO detector at the YBJ Observatory. The primary site for HAWC would be just beyond the ARGO detector. Yangbajing, a town with ~4000 residents, is located approximately 90km from the city of Lhasa. Lhasa is now served by a major rail line with a maximum altitude of 5072m the highest railway in the world, which opened for passenger traffic in July of In addition there is regular air travel between Beijing and Lhasa (via Chengdu). From the U.S. there is

43 roughly two days of travel to arrive in Lhasa (including an overnight in Beijing). The town is home to a series of hot springs and generates 25 MWatts of power locally from two geothermal power stations. At the site of the YBJ Observatory there are dorms with roughly western style rooms and a comparable number of less modern rooms. The cost of the western rooms is approximately $20/night (which includes food prepared by a local cook). Figure 2 - Picture of EAS/γ array near the potential HAWC site in Tibet The YBJ site is located in the Tibetan plateau: essentially a large valley surrounded by high mountains. Most of the precipitation falls in the mountains, resulting in little snowfall at the observatory. The long-term weather data (seasonal temperatures, rainfall, and snowfall) are given in Table XX. <T> Year January 2.5 C -7.2 C <T> High Low 23.4 C C <ΔT> Day/Night Pressure Sunshine <Hrs>/yr 15.9 C mb % Table 1 Long-term climate data for YBJ Observatory. Rain (mm/year) Average Extreme (high/low) mm 430.5/296.8 mm Snow <Days>/yr Depth (cm) 26.4 days <7 cm Both, the Institute for High Energy Physics and Tibet University are interested in this project. The Director of IHEP (Hesheng Cheng) has given his support to the project as seen in the attached letter. The Chinese collaboration consists of Key Laboratory of Particle Astrophysics, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 0049, China X.J. Bi, Z. Cao 1, Q.B. Gou, Y.Q. Guo, H.H. He, H.B. Hu 2, H. Lu, P.R. Shen, X.D. Sheng, F. Shi, Y.H. Tan 3, H. Wang, C.Y. Chao, M. Zha, H.M. Zhang, J.L. Zhang Department of Mathematics and Physics, Tibet University, Lhasa , China T.L.Chen,Danzengluobu, X.H.Ding, Haibing Hu, Labaciren, H.J. Li, X.R. Meng, C.C.Ning, Zhaxisangzhu, Zhaxiciren, A.F. Yuan Department of Physics, Shandong University, Jinan 2500, China X.Y.Zhang Institute of Modern Physics, South West Jiaotong University, Chengdu 6031, China H.Y.Jia

44 Physics Department and Tsinghua Center for Astrophysics, Tsinghua University, Beijing 0084, China Y.-Q. Lou 1 Chinese spokesman for ARGO detector and member of the High Resolution Fly s Eye collaboration 2 Chinese spokesman for the ASγ detector 3 previous spokesman for the ASγ detector and YBJ Observatory This is a strong group with a proven track record. The Chinese were responsible for the construction of the ARGO building (which was built on time and within budget) and has operated the ASγ observatory since its inception. IHEP provides 4-5 technicians on site at the YBJ Observatory to support the construction and operation of the existing experiments. These technicians would be available to work on HAWC and assist in the maintenance and operation of the detector. Chinese funding for HAWC would flow from two sources. The IHEP has a relatively small amount of discretionary funds that would be used to provide the required site infrastructure (power, water, and communications to the building site). In addition the IHEP group would apply to the Chinese NSF for funding to support their share of the construction and operation of HAWC.

45 Sierra Negra, Puebla, Mexico Sierra Negra is located at 18.9 N, 97.4 W, which is about a 2 hour drive from Puebla, a town of population ~ 2 million. Puebla has an international airport with flights to Houston and is ~ 2 hour drive from the international airport in Mexico City. The Sierra Negra site has been developed for astrophysical projects. In particular the largest project is the >$0M Large Millimeter Telescope (LMT), which is a collaborative effort between the Mexican national laboratory, Instituto Nacional de Astrofísica Óptica y Electrónica (INAOE), and the University of Massachusetts. The LMT with a diameter of 50 meters will be the largest single dish millimeter observatory. The inauguration of the telescope will be in November More information about the telescope can be found at Figure 4 shows photos of the LMT and the possible HAWC sites. Three different sites for HAWC are being considered and are shown on the map of Figure 3. The altitude of these sites is 40 m, 4040 m, and 4300 m as shown from left to right with the third site located in The Orizaba Peak. Detailed topographic measurements are being made of each site to help determine which is most appropriate for HAWC. Despite the high altitude, the low latitude of the site keeps the weather temperate and stable with an average temperate at 4600m of ~1 o C and typical daily swings of ~2 o C. Several astrophysics projects are located near the LMT, and a consortium has been founded to share resources amongst these projects. The current list includes the LMT; RT5, a 5m radio telescope for solar and interstellar medium studies; a solar neutron monitor; an antineutron experiment; a cosmic ray detector; and a seismological detector. If HAWC is built at this site, we would also be a member of this consortium, and have been invited to join by the director of INAOE (see attached letter). The land used by the consortium is part of the Orizaba National Park that was established in No formal management plan exists for this park; however, INAOE is discussing whether it could provide this service with the government agency that oversees this park. As discussed in the attached letter, INAOE, as collaborators on HAWC, would seek permission and obtain the required environmental impact declaration. All of the previous requests for astrophysics projects at Sierra Negra have been approved. Furthermore, for the third site, a permission to use a 1km 2 for a cosmic ray experiment already exists and it would only have to be extended to HAWC without increasing the site extension. The resources shared by the consortium include roads, security, power and, in the near future, high bandwidth communications. The LMT also currently has an office and will have a dormitory in Atzitzintla (pop. 8000, 3000m elev.) that HAWC as part of the consortium could share. Additional costs would be required to extend electricity and communications for approximately 1km to the location of HAWC and are included in the budget. The HAWC power and communications requirements can be met with the existing or planned capacity. HAWC requires 125 Mliters of water, and several options are being considered for how to get this water. The budget of HAWC currently includes $600K for acquiring the water. This option uses a system of pipe ~15 km long and multiple pumps to move the water from wells at a lower elevation to the site. However, cheaper options are likely possible and are being studied, such as a ~150 m deep well near the site where an underground stream has been electrically detected, or by using gravity to pipe the water from intermittent streams to gather a fraction of the average precipitation of ~80 cm/ year. Locating HAWC at this Mexican site would bring a large and strong collaboration of scientists to work on this project. These scientists have extensive background and interest in cosmic ray physics, high energy physics, and astronomy. These institutions would provide 4 technicians (2 from BUAP and 2 from INAOE), and most of these collaborators would have at least one student working on the project. The list of collaborators and their institutions are

46 Instituto Nacional de Astrofísica Óptica y Electrónica (INAOE) 1 Alberto Carramiñana (Director of Astrophysics, γ-ray astronomy & LMT) Eduardo Mendoza (Solar Physics, Interstellar Medium) Universidad Nacional Autónoma de México (UNAM) 2 Instituto de Astronomía Magdalena González (Milagro collaborator, GRBs), Dany Page (neutron star theory), William Lee (GRB theory), Hector Hernández (optical & multiwavelength), Deborah Dultzin (AGNs), Erika Benitez (AGNs) Instituto de Física (experimental HEP, cosmic rays, ALICE at CERN, AMS, CREAM) Arturo Menchaca (Director of the Institute), Rubén Alfaro, Andres Sandoval, Ernesto Belmont Instituto de Ciencias Nucleares Lukas Nellen (AGN - Auger), G. Medina-Tanco (HECR- Auger) Instituto de Geofísica José Valdés Galicia (Director of the Institute, Chair of 2007 ICRC in Merida, Mexico, solar physics), Alejandro Lara (solar physics) Benemérita Universidad Autónoma de Puebla (Auger) Humberto Salazar, Oscar Martínez, Cesar Álvarez, Arturo Fernández Universidad Michoacana de San Nicolás de Hidalgo Luis Villaseñor (Auger) CINVESTAV Arnulfo Zepeda (Auger spokesman) Universidad de Guanajuato David Delepine (ν astrophysics), Gerardo Moreno, Marco Reyes, Luis Ureña (HEP), Victor Migenes (astronomy) University of Torino, Italy Oscar Saavedra (cosmic rays) This collaboration of Mexican scientists will write proposals to obtain funding for this project from the Mexican government. Typically these proposals can be for up to $500k, but there is a current call for proposals for larger projects, and they have submitted a preliminary proposal. Even without this current opportunity, several proposals will be submitted for different scientific investigations. This strategy was pursued for Auger and $2.5M was received. N 1 km Figure 3: Topographic map of the area around the Large Millimeter Telescope which is located on the top of Sierra Negra at an elevation of 4600m. At the upper right corner is Pico de Orizaba, the highest peak in Mexico at 5700m. The three blue boxes denote the location of possible HAWC sites. The road to the LMT passes near the center site and is mostly paved to this point. The town of Atzizintla, where INAOE has an office, is ~15 km from the site to the south. LMT 1 National laboratory in Puebla, Mexico. 2 Largest university in Mexico. Located in Mexico City.

47 Figure 3: Photos of the Sierra Negra area, starting at the top left and going clockwise. 1) Sierra Negra on the left and Pico de Orizaba on the right. The LMT is visible on the top of Sierra Negra. 2) The 50 m diameter LMT dish at 4600 m elevation. 3) Potential HAWC site at 4300m. 4) View of same site with Pico de Orizaba behind. 5) Pico de Orizaba as seen from the LMT site on Sierra Negra. The 4300 m and 40 m site are indicated by arrows.