FY 2005 Annual Report And Request For FY 2006 Funds

LASER INTERFEROMETER GRAVITATIONAL WAVE OBSERVATORY

LIGO Laboratory / LIGO Scientific Collaboration

LIGO-M050209-00-P LIGO November 25, 2017 FY 2005 Annual Report and Request for FY 2006 Funding

Edited by Philip Lindquist and David Beckett

California Institute of Technology Massachusetts Institute of Technology LIGO Project – MS 18-34 LIGO Project – NW17-161 1200 E. California Blvd. 175 Albany St Pasadena, CA 91125 Cambridge, MA 02139 Phone (626) 395-2129 Phone (617) 253-4824 Fax (626) 304-9834 Fax (617) 253-7014 E-mail: [email protected] E-mail: [email protected]

LIGO Hanford Observatory LIGO Livingston Observatory P.O. Box 1970 P.O. Box 940 Mail Stop S9-02 Livingston, LA 70754 Richland WA 99352 Phone 225-686-3100 Phone 509-372-8106 Fax 225-686-7189 Fax 509-372-8137 http://www.ligo.caltech.edu/

03b67c30f2049a1c280b10b8508db86c.doc LIGO LIGO-M050209-00-P

1 PROJECT SUMMARY The Laser Interferometer Gravitational-Wave Observatory (LIGO) detectors use laser interferometry to measure the distortions of space-time between free masses to directly detect passing gravitational waves. The objective is to open the field of gravitational-wave astrophysics. Scientists, engineers and staff at the California Institute of Technology (Caltech), the Massachusetts Institute of Technology (MIT), and the two observatory sites in Livingston Louisiana and Hanford Washington are commissioning and operating the LIGO detectors. Caltech has primary responsibility for the project under the terms of a Cooperative Agreement1 with the National Science Foundation (NSF). LIGO is a national facility for gravitational-wave research, providing opportunities for the broader scientific community to participate in detector development, observations, and data analysis. LIGO welcomes the participation of outside scientists in these endeavors. The LIGO Scientific Collaboration (LSC) is the organization comprising the scientific community. This includes Caltech and MIT scientists and engineers responsible for data analysis, advanced R&D and the development of advanced subsystems for LIGO. This Collaboration will exploit the initial detector and is pursuing the development of second-generation detectors. The LSC has its own management structure with shared participation in its governance, and corresponding obligations and privileges. The initial LIGO system comprises one three-interferometer detector system. The sites provide space and infrastructure support for the eventual expansion of the facility to a multiple-detector configuration. This Annual Report summarizes the progress and status during the fourth year of operations (LIGO FY 2005) and the plan for the fifth year (LIGO FY 2006) activities associated with the completion of commissioning the initial detector, commencement of long-term scientific observations, and the continuing research and development for the next generation of detectors. The LIGO fiscal year begins on October 1 and ends September 30. LIGO is requesting $32 million for the twelve months beginning October 1, 2005. Of this total, $29 million is for operations and $3 million is for Research and Development. Approximately $1.5 million of the Operating funds will be used to support the R&D effort including campus research facilities and advanced detector prototypes. The work plan and the total budget of $158 million requested over five years are consistent with NSF Cooperative Agreement No. PHY-0107417 1 issued February 2002 as modified by the NSF for fiscal years 2005 and 2006. The original request was for $160 million.

2 Work Accomplished During FY 2005 2.1 LIGO Hanford Observatory

Commissioning activities and a science run dominated Hanford operations during this year. These activities led to improvements in sensitivity, reliability and data quality for the Hanford four- and

1 NSF Cooperative Agreement No. PHY-0107417, Amendment No. 011, “Laser Interferometer Gravitational Wave Observatory LIGO,” May 10, 2005.

2 LIGO LIGO-M050209-00-P two-kilometer interferometers. Detector characterization efforts led to reductions of noise introduced by optical scattering and the phase stability of the local oscillators used in heterodyne signal recovery. Significant work was expended to understand how these signals are generated and how they respond to changes of alignment and optical path in the interferometers. A study was performed to understand how output mode cleaning can be used to improve interferometer performance. Improvements to the interferometer controls greatly improved the stability of operations and provided better immunity to the microseism. We have developed a method to stabilize small deviations in optical curvature using an auxiliary laser system and we have developed diagnostics to better understand how heat is deposited in optics by the laser beams. These changes have allowed improved duty cycle for the detectors. A number of effects that produced spurious transients and spectral lines in the data from previous science runs were diagnosed and either removed or mitigated. The fourth science run (S4), our most sensitive run to date, was conducted during FY2005. One measure of this improvement is the distance out to which one could observe with confidence the inspiral of two neutron stars. For much of S4, this distance was of order twenty-five million light years for the four-kilometer interferometer and approximately ten million light years for the two- kilometer interferometer. The improvements in low-frequency performance will also enable much better results in pulsar and stochastic background searches. Another measure of improvement is that both of these detectors produced scientific data with an 81 percent duty cycle throughout this run. We have instituted an “Astrowatch” program during commissioning periods, whereby we are archiving running data from the interferometers when they are not directly manipulated, so that we can analyze the data later if an interesting event is found by other astronomical detectors. For instance, when it was announced that a significant magnetar flare had occurred nearby in our galaxy, we were able to look back at data from our four-kilometer interferometer that was archived as part of Astrowatch, even though we were not in a science run.

3 LIGO LIGO-M050209-00-P

Figure 1 A histogram of the estimated "reach" for the three LIGO detectors (H1, H2, and L1) during the S4 science run.

2.2 LIGO Livingston Observatory

2.2.1 HEPI fully operational Commissioning of the HEPI (Hydraulic External Pre-Isolator) system was an unqualified success (fabrication and installation were discussed in the last report). Its operation has rendered the L1 interferometer nearly immune to noise from forest logging and from the ocean storm-driven microseism. In addition, the HEPI system itself, which is among the more complex interferometer subsystems, has proven extremely reliable and robust. This is especially significant since many engineering features of the HEPI control systems were conceived as prototypes for Advanced LIGO active systems.

Figure 1: XX hour locked stretch spanning time of high logging activity

2.2.2 Interferometer sensitivity advances With HEPI, the L1 duty factor for both science and commissioning activities is now comparable to those of Hanford’s H1 and H2. This has accelerated the pace of L1 detector sensitivity and reliability improvement.

The stable platform afforded by HEPI permitted complete commissioning of all degrees of freedom in the wavefront sensing alignment system. In turn, this achievement itself yielded an even more stable platform, permitting upgrades to beam centering controls, auxiliary interferometer length control loops, alignment automation, and diagnostics.

In this final category, the L1 noise analysis model was extended not only to include newly discovered noise terms, but to directly interface with the interferometer global diagnostics system. This allows the noise model to collect and automatically integrate parameter and source measurements and compute a near-realtime budget of noise contributions. Scientists and engineers analyzing detector performance can now instantly isolate excess noise terms, even those arising from relatively rare parameter combinations. The improved diagnostic tools have further accelerated the pace of commissioning and detector improvement, and are now being ported to the H1 and H2 machines. One unanticipated noise interaction previously limiting performance of H1 and L1 was an anomalous sensitivity to RF phase noise in local oscillators used for fringe interrogation. This was initially discovered on L1, and limited performance above about 1 kHz. An ultralow-noise crystal oscillator and distribution system was developed to mitigate this problem and installed early in the S4 run; the high frequency noise is now consistent with photon shot noise at high frequencies. Improved oscillators have now also been ported to H1 and H2 machines. The next frontier for L1 is operation at full design laser power. The Livingston interferometer’s optics are comparatively low in absorption to begin with, leading to a natural role as pathfinder in exploring high power. In addition a Thermal Compensator System like that on H1 was installed; this uses an auxiliary carbon dioxide laser to “paint” a complementary heat pattern on the test masses and reduce absorption-induced gradients. Optimization of this system combined with incremental power improvements (to approximately 3.5 W entering the mode cleaner) yielded record sensitivity. Recent results at high power indicate binary inspiral range in excess of 10 megaparsecs. Despite some setbacks due to laser component failures, L1 is now beginning operation at full design power, approximately 6.5 W entering the mode cleaner cavity. Efforts are underway to fully harness this increase and realize reduced shot noise.

Figure 2: Recent noise analysis from L1 interferometer

2.2.3 S4 observing run successful Again largely due to the success of HEPI, L1’s performance goals for the S4 science run were achieved with some margin and comparatively early, permitting thorough pre-run characterization and leading to very high data quality and noise stationarity. The integrated 75 percent duty factor for the run, while quite welcome after the dismal 23 percent logged in S3, was heavily biased by a catastrophic Livingston power transmission line failure early in the run; the latter half of the run showed asymptotic uptime exceeding 80 percent, similar to both Hanford machines. We are thus confident that the unusual seismic environment in Livingston is no longer an issue.

Figure 3: L1 cumulative uptime vs. calendar date for S4 observing run

2.2.4 Data analysis speed and depth improved The Livingston Beowulf parallel computing cluster is now fully subscribed with analysis of S3 and S4 data, primarily burst and inspiral analysis. The cluster is poised for an expansion to 240 nodes later this summer; a corresponding upgrade to the power and cooling facilities is in preparation. In addition the new Gigabit Ethernet fiber connection between LLO and the LSU campus was used to stream real-time high-bandwidth interferometer data to Caltech and other analysis centers during and after the S4 run. Online analysis functions also made substantial advances in speed and efficiency prior to the S4 run, such that preliminary analysis results were available in near real time. This led to identification and rectification of instrumental issues within the first few days of the run, greatly improving the overall quality of the data set.

2.2.5 Outreach and education The LIGO Science Education Center program (reported separately under grant #OUTREACH#) is now in full swing, strongly leveraging the existing LLO outreach offerings provided through LIGO operations. We took delivery of the first group of seven hands-on interactive exhibits from the San Francisco Exploratorium, focusing on the core themes of wave motion, oscillations and resonance. In addition to enhancing our existing program of school field trips and hands-on science activities at the site (approximately #2,000# students participated this year), these exhibits helped us kick off our program of LIGO science-based professional development workshops for K-12 teachers. We have received excellent feedback from both teachers and administrators involved in the first three

7 LIGO LIGO-M050209-00-P workshops. Two K-12 science teachers from the Baton Rouge area have joined us in the Research Experience for Teachers (RET) program, and are working on curriculum development centered on the exhibits and LIGO science topics. In addition this year two of the five Summer Undergraduate Research Fellows (SURFs) resident at the LLO site are working on projects related to education. The site preparation for the new Science Education Center is underway and the architectural design of the facility has been completed. At this writing we are soliciting competitive bids for construction, and expect groundbreaking this fall. The tentative construction schedule targets completion and opening in time for the fall school term of 2006.

QuickTim e™ and a TIFF (LZW) decompress or are needed to s ee this picture.

Figure 4: Site plan for the new LIGO Science Education Center (Eskew+Dumez+Ripple, architects). The center is to be fully integrated with the existing LIGO autidorium and conference facility (shown shaded at left).

QuickTim e™ and a TIFF (LZW) decom pres sor are needed to s ee this picture.

Figure 5: Science Education Center perspective view. The southern facade will be a wind- driven kinetic exhibit comprising coupled pendulums, designed in collaboration with the San Francisco Exploratorium.

2.3 Safety No significant safety problems or issues were found during the annual safety audits of the Hanford and Livingston Observatories and the LIGO Caltech laboratories. The LIGO program continues to place a high priority on personnel and system safety. The observatories continue to conduct weekly “all-hands” safety meetings, which include review of emergency safety training and procedures, review and critique of safety related incidents and the associated “lessons-learned.” The Control Room operations continue to show significant improvement. In addition to increased training and opportunity to participate in the commissioning work, the operators have access to an improved system for recording operational experience and data along with the use of electronic check lists and other "note-book" information for operator reference if faced with operational questions or problems. The operation of the laser safety interlock systems continues to be satisfactory. There have been a few cases of "hardware-glitches" which have prompted some procedural and operational modifications.

2.4 Detector Commissioning Detector commissioning in the past year has yielded significant improvements in the sensitivity and reliability of all three interferometers. Detector noise levels are now within “spitting distance” of the performance goal. The improvements can be classified by frequency band:  Above ~100 Hz, noise has been reduced by increasing the input laser power, closing in on the original design level of 6 W injected at the interferometer’s recycling mirror.

 Below ~100 Hz, noise has been lowered by reducing the noise in auxiliary control loops in the interferometer, and implementing lower noise control electronics. One key to higher power operation and improved shot noise sensitivity has been the implementation of an active thermal compensation system (TCS). This system uses CO2 lasers to externally heat parts of the input tests masses, providing a controlled thermal lens used to optimize the modal properties of the recycling cavity. First exercised on H1 in 2004, these compensation systems are now in use on all three interferometers. By carefully tuning the compensation beams, several benefits are achieved: an increase in the optical gain of the interferometer; a reduction of the sensitivity to phase noise on the RF modulation; a reduction in the anti-symmetric port orthogonal-phase signal level, which could otherwise saturate the detectors. Several other tasks have been completed, or are in progress, to enable operation at higher power: three of the Lightwave Electronics, Inc. lasers have been refurbished to regain their original output power of 10 W; the chain of optical elements between the laser and the vacuum system has been tuned up to increase the optical efficiency; photodetectors have been improved so that they can handle higher power; active beam pointing stabilization is being implemented on some detection tables to deal with thermally induced beam drift. A major task is the replacement of H1’s X-arm input test mass (ITMX). From early on, H1 has seemed to exhibit higher than expected optical absorption in its core optics. Several types of measurements over the past year have identified ITMX as the most absorptive optic; it is estimated that ITMX absorbs about 10x more power than assumed in the original design. With the full incident laser power, this absorption stretches the limits of the TCS’s ability to compensate. The strategic choice was made to replace ITMX with a spare, measured as having low absorption in Caltech’s optics test facility. Higher dynamic range, wider bandwidth laser power stabilization servos have been implemented on all interferometers. This has also been important to reducing high frequency noise. Lower frequency noise reductions were the result of several improvements:  Reduced noise in the Michelson and power recycling cavity degrees-of-freedom, through improved shot noise sensitivity in these channels, and better servo loop design.  Reduced noise from the wavefront sensor (WFS) alignment system, through noise reductions in these channels and better de-coupling from the gravitational wave channel.  Reduced noise in the electronics that drive the test mass coil actuators.  New custom-made digital-to-analog converters for the most sensitive channels, with 30-40 dB lower noise than the previous model.  Reduction of scattered light coming from the end test mass transmitted beams, through the use of better optics and simpler layouts. In addition to the noise reductions mentioned above, other improvements have been made to the WFS alignment system, which determines and controls the optimal orientation of the interferometer optics. In this past year the bandwidth of these angular controls has been increased significantly, by an order of magnitude or more. Since the interferometer is a highly coupled optical system, this required careful de-coupling of the sensed degrees-of-freedom in the control

10 LIGO LIGO-M050209-00-P matrix for these loops. The benefit has been a more stable interferometer, with greater suppression of fluctuations that would otherwise saturate detectors. Many of the more routine commissioning tasks have been automated over the past year, providing better reproducibility, more uniformity among the three interferometers, and increased efficiency of commissioning. A prime example of this automation is the interferometer noise analysis model. This model, or ‘noise budget’, calculates and displays the equivalent displacement noise due to all known sources of interferometer noise, both technical (e.g., electronics noise) and fundamental (e.g., thermal noise). An example of such a noise budget plot is shown in REFERENCE TO L1 NOISE BUDGET IN LLO SECTION. In the past, data for each of the noise sources were collected and calculated by hand, a process which typically required a specific knowledgeable person for each interferometer. Now, having interfaced the model with the interferometer global diagnostics system, anyone can collect and automatically integrate parameter and source measurements and compute a near-real-time budget of noise contributions. Scientists and engineers analyzing detector performance can now instantly isolate excess noise terms, even those arising from relatively rare parameter combinations. The improved diagnostic tools have further accelerated the pace of commissioning and detector improvement. New high-frequency gravitational wave channels have been added to all three interferometers. The idea is that the interferometers have a sensitivity to gravitational waves around the free-spectral- range frequency of the arm cavities that is comparable to (though somewhat reduced from) their sensitivity at low frequencies. Thus fast data acquisition channels now collect and store data around 37.5 kHz for H1 and L1, and around 75 kHz for H2.

2.5 Data and Computing Group

2.5.1 Modeling & Simulation The simulation group supported initial LIGO commissioning by providing analytical insight into complicated problems and quantitative suggestions for solving those issues based on two software packages, “e2e” (which stands for “end-to-end”) for time dependent analyses and “FFT” (Fast Fourier Transform) for static analyses. These software packages complement each other. A major issue is the effect of absorption-induced thermal deformation of input test masses (ITM). To make the beam profile of sideband fields match that of the carrier field, a Thermal Compensation System (TCS) has been implemented. Simulation programs were used in the design and implementation for the following purposes:  to quantify the details of the beam profile induced by optical path deformation which cannot be approximated by a simple lens model  to understand the difference of the side band imbalance due to mode mismatching  to quantify the effect of an input beam profile which does not match the optimal mode profile of the core optics system  to relate these mode mismatches to various error signals and sensitivities. We measured the as-built mirror surface phase maps earlier during LIGO construction. Minor (acceptable and within specification) imperfections were noted. Nonetheless, such imperfections affect the ultimate performance of LIGO and they induce effects which do not exist in idealized interferometer models. FFT analyses using these phase maps were performed to quantify these effects. One example is the dependence of the Gouy phase on mirror tilt angles. When all mirrors

11 LIGO LIGO-M050209-00-P have idealized parabolic shapes, the Gouy phase does not depend on the mirror tilt. However, due to mirror surface aberration, the actual Gouy phase will depend on mirror tilt. This explains the observed day-to-day variation of the g-factor 2 of the arm cavity. Another commissioning challenge was the mysterious behavior of wave front sensor signals (WFS). When signals for other channels indicated improved performance of LIGO, some WFS discriminator signals became weak and non-robust. The optimal telescope configuration was modified by using analytic and simulated calculations which take thermal effects into account. By changing the telescope based on this result, the performance of WFS system was improved and the mysterious behavior disappeared. The development of the simulation environment for advanced LIGO has started. The e2e code developed for LIGO I simulation will be used with some improvements and additions. The major modification needed for advanced LIGO simulation is a fast simulation module for the dual recycling Michelson cavity (DRMC). By using an approximation that the time dependence of physical optics quantities are linear over a time step during which the field propagates through the long arm, the response of fields in the DRMC can be calculated several 100X faster. The formulation for a scalar field has been completed and the C++ code development is on the way. This AdLIGO Detector simulation package is being developed using the e2e framework. A preliminary version has been completed, which includes simplified basic subsystems, parallel chains of quad pendulums, and an LSC system with analog and digital components. Using this advanced LIGO package, the locking algorithm of the new core optics system is now being investigated.

2.6 LIGO Data Analysis and Computing The LIGO Data Analysis Systems (LDAS) [need a paragraph.]

2.6.1 Software Systems. During FY 2005, the LDAS software development group prepared seven major releases of LDAS and one minor release for distribution to the LIGO Laboratory computing centers, twice the number of releases as were made in the previous year. All releases were fully functioning LDAS software releases that typically addressed enhancement requests, performance improvements and migration to newer operating system versions. All releases were coordinated with the recently formed Data Analysis Software Working Group (DASWG) within the LIGO Scientific Collaboration. The highlights of each release are given below:  1.2.0 (Sept 8, 2004): Improved performance by ten percent.  1.3.0 (Nov 12, 2004): Added the ability to specify the number of frames per frame file, the number of seconds per frame, a flag to perform checksums on input frames, a flag to allow shorter than specified frame files when data dropouts occur, a flag to generate the checksum

2 g-factor: g1(2) = 1 - L/R1(2), g=g1*g2 where L is the cavity length and R1(2) is the radius of curvature of mirror 1(2) of the Fabry-Perot cavity. g is a quantity used to characterize the spatial quantity resonating in a FP cavity. LIGO Note T050030 by Rick Savage describes the measurement.

of output frames and a flag to allow missing data to be filled to the createRDS command. Developed a utility for scanning the beowulf cluster for Linux system errors. Enhanced the diskCacheAPI so that system administrators can specify subdirectories to ignore under actively monitored directories.  1.4.0 (Jan 13, 2005): Ported LDAS software to run under DASWG approved Fedora Core 3 Linux operating system.  1.5.0 (Feb 14, 2005): LDAS release targeted for us in the S4 Science Run. Fixed issue with creating reduced data sets from frames containing subsections of compressed data from the specified input frames. Significantly improved and automated testing of software distributions.  1.5.1 (Mar 4, 2005): Made a minor fix to the createRDS command to allow non-aligned LHO and LLO frames to be merged together.  1.6.0 (May 23, 2005): Added the ability to generate md5sums for each output RDS frame file and logging of this to a file for use in data movement between systems over the network. Changed default compression level for RDS output frames from six to one to improve performance. Improved support for GEO frames in the RDS generation command. Enhanced the FrVerify utility to verify metadata in the frame file against the name of the frame file. Updated database to IBM DB2 8.2.1.Removed dependency on a Beowulf cluster as part of an LDAS system for conducting integration and system testing. This was the first of two releases requested by DASWG is preparation for the astrowatch dry run testing of LIGO software prior to the S5 Science Run.  1.7.0 (July 7, 2005): Targeted release for the astrowatch dry run beginning on July 15th. Fixed bugs identified in the 1.6.0 release necessary for the astrowatch dry run. Includes one button full system testing from the controlMonitorAPI’s graphical user interface.  1.8.0 (Sept, 2005): Port of LDAS to the new DASWG approved Linux and Solaris operating systems that need to be in place for the S5 Science Run. During the S4 Science Run LDAS primarily carried out the process of generating Reduced Data Sets (RDS) from raw frames. These frames were then quickly distributed over the wide area network to LSC Tier II centers for prompt data analysis. The generation of custom simulation and RDS frames by individuals working in the various analysis groups became popular enough to on several occasions after the science run to cause an overflow of the local file space on LDAS system servers, prompting that for the 1.6.0 release of LDAS, all frames generated by LDAS be compressed by default to conserve disk space. The LIGO Laboratory is now collaborating in three Grid Computing projects; GriPhyN, iVDGL, and the Open Science Grid (OSG). These projects are developing large scale grid-based computing infrastructure which LIGO is leveraging through various research and development and prototyping activities. LIGO has maintained an active involvement with the Pegasus development team since the beginning of the GriPhyN Project. In this past year, working within the LSC, a demonstration application was deployed on the grid, utilizing Pegasus to carry out a distributed binary inspiral search of LIGO data using several clusters from the LSC Data Grid. This application was also demonstrated to the Grid Computing community at the Super Computing Conference in November of 2004. LIGO Laboratory also set up a small test cluster for use in the OSG Integration Test-bed activity. This provided LIGO with an inside look at the evolution of

13 LIGO LIGO-M050209-00-P iVDGL’s GRID3 infrastructure as it migrates onto the larger, more open infrastructure planned for OSG. Presently LIGO Laboratory has standing members on the OSG executive board and council working to assure that this grid technology has application for LIGO’s future data analysis needs.

2.6.2 Hardware Systems Data from engineering run E11 and the fourth science run (S4) where successfully acquired and archived at both Observatories with one copy of these data being successfully transferred to the central archive at Caltech. During S4, as with the previous S3 run, the level-3 reduced data set (RDS) was transferred from the observatories to the central archive using Internet2 with approximately 1 hour latency for near real- time joint LHO/LLO analysis and redistribution to Tier-2 sites. This transfer was accomplished using the Lightweight Data Replicator (LDR) which is based on standard Grid tools (Globus, RLS, ...). However, for S4 LDR was enhanced by the LSC to allow the successful transfer of the full frames (~10MByte/sec) and the level-1 RDS frames in addition to level-3 frames over Internet2. This obviated the need to ship tapes. The main archival system continues to utilize the SAM-QFS Hierarchical Storage Management (HSM) system at both observatories and Caltech. Compared to 1 year ago the 3 archive systems have grown as follows:  LLO) 10M files/47TByte grown to 23M files/82TB,

 LHO) 13M files/84TByte grown to 18M files/207TB,

 CIT) 14M files/200Tbyte grown to 24M files/382TB.

The size of the SAM-QFS disk cache at the observatories where approximately doubled in size to 7TB at LLO and 10TB at LHO. The computing clusters at the observatories remained at 70 and 140 dual-2.66GHz P4-Xeon nodes, but the amount of internal cluster scratch disk space was doubled to 28TB and 56TB, respectively. The cluster at Caltech grew from 210 identical nodes to include an additional 80 dual-2.8GHz P4- Xeon nodes and the internal node storage increased from 48TB to 116TB. The MIT cluster continued to run with 112 2GHz P4 nodes, however, the cluster storage was increased from 2TB to 22TB. All four of these clusters are now operated primarily through the Condor batch queue system.

2.7 LIGO Laboratory IT Support Group Security, mass storage and network connectivity have been major topics that were and are being addressed. Security awareness has increased significantly as LIGO Laboratory has moved into scientific operation supporting a modest sized collaboration. More uniformity between the Observatories is taking place in that area. All four sites are now supporting Linux as one of the approved operating systems. A rough draft of the configuration of the OS installation has been developed and is being refined. All sites have increased their mass storage capacity for General Computing user home accounts and data analysis storage.

Network connectivity has been stable. Until just recently, the Observatories had greater than 99% uptime with no major down time. The Universities have had greater than 99% uptime with no major down time. MIT: MIT has set up a number of new server systems. The wireless network has been expanded with a new NAT router setup that has also improved security by limiting access to the wireless system. Livingston: The WAN connection bandwidth at the LLO site was improved considerably after replacing the router that was installed January 2004 with a Linux router/firewall. It is possible that at a later date a Cisco device may be installed as the router, but at this time the economy and benefits of running the Linux router are more advantageous. Livingston has been using an average of 15 Mbit/sec of bandwidth with peaks of up to 90 Mbit/sec. Work is being performed to improve the data transfer rate. Several RAID arrays have been brought on line for General Computing. There is now adequate storage on the general computing LAN for software mirrors, scratch space, etc. A central logging server for all GC servers and CDS machines has been setup. This has been beneficial in troubleshooting network problems, for network security and it provides visibility into the "health" of the system in general. The configuration of the general computing and CDS network is undergoing a number of changes mandated by the need to improve security and operational convenience Hanford: LHO network usage for S4 data transfers averaged about 84 Mb/sec with a maximum daily average of 94 Mb/sec on the OC3 network. No network down time during the science run and no noticeable impact to other users on the network during the S4 run. Testing of the data transfer rates had been ongoing during E12 with maximum usage peaks at 107 Mb/s. During the engineering run E12, ESnet had reported memory allocation errors on their router that handles the LIGO traffic. ESnet was able to take to reduce or eliminate adverse affects to their router. The current OC3 contract with Amerion and PNNL has been extended until Sept. 13, 2005. PNNL, in partnership with Charter Communications, is installing DWDM equipment to establish multiple "lightwaves" to Seattle.. Arrangements are being made to carry LIGO traffic over PNNL's DWDM infrastructure in a GigE circuit. This is an unprotected circuit and LIGO has requested bids from other ISPs for backup service between PNNL in Richland and ESnet in Seattle. LIGO is also requesting bids from other ISPs for a Layer 3 WAN back up network service that does not connect to the ESnet. The PNNL/LIGO gigabit circuit will be available in late July or early August 2005. LLO and LHO General Computing are working on implementing a more consistent approach to cyber security. The differences of the WAN connections at each site are so different they cannot be identical in hardware and software, but the philosophies can be consistent. The trend for the future is to use additional NAT routers and internal networks to isolate computer systems and LANs. This can prevent an attack on one computer from spreading to other computers on site. See the separate discussion on cyber security in Section XXX(Organization Section?). CIT: The LIGO CIT network was able to support multiple large data transfers during the S4 run with margin. The core gateway router peaked at 39% of the backplane capacity. The average during the S4 run was around 20% of capacity.

The internal edge router (now up for 634 days without needing a reset) peaked at Tx 42 percent and Rx 32% of line capacity (this is the gigabit connection to the gateway router line) during the S4 run. This occurred while receiving data from the two remote Observatories and transmitting the data to the LDAS data storage system and retransmitting the data to other LSC Tier II center universities. The server systems and core network systems were operational during two major, scheduled, power outages of the GC computer room. This contributed to having over a 99% uptime of LIGO systems and services at CIT. The LIGO CIT group has started replacing SUN workstations with LINUX workstations as the SUN equipment becomes obsolete. New servers for the Engineering Data Storage vault (PDMWorks), PRISM and other PC applications have been installed along with a couple of new multiple CPU sanbox units. An upgrade and overhaul of the mail system was performed. The backup system has also been augmented to a 6TB system. The GC group has an active part in work dealing with the DCC. The Caltech group assisted PSU in starting up the LIGO.ORG mail services. The LIGO.ORG core system remains at CIT but some of the services are being shared with other collaboration sites. A couple of lower end server class or higher end workstation boxes will be dedicated to security related tasks.

2.8 LIGO Laboratory Cyber Security The NSF sponsored a Cyber Security Workshop in September 2004 in which all major NSF facilities and research organizations were asked to participate. LIGO Laboratory sent senior networking and systems administration personnel from each LIGO site. As a result of the workshop, LIGO has begun to place greater emphasis on computing security. As it was discussed during the NSF Workshop, this was partly spawned by several high profile compromises at various computing and research facilities across the nation. Subsequent to the workshop, the LIGO directorship developed a cyber security policy and appointed a cyber security team to oversee its implementation. We now have a Computing Security Officer and a Computing Security Coordinator. The plan was presented to NSF during the annual fall review of LIGO Laboratory. Early in 2005, principals from each LIGO Observatory and Caltech met at LHO to begin the process of reviewing LIGO Laboratory IT infrastructure for compliance with the cyber security policy. LIGO has taken a look at the various computing groups and systems that it operates and determined which areas are most critical to the operation of the observatories. A committee has been formed to manage the Critical Systems security within LIGO. The focus of computing security has primarily been applied to the critical systems to date. As an example, attention has been given to the following areas within the critical systems:  Patch management  OS selection & maintainability

 Network topology and firewalls  Logical network placement of various network enabled equipment  Password and account policies  Available services and needs  Visibility & monitoring of network traffic Due to the number and complexity of systems that we maintain, it has been decided that we should prioritize the changes that have a high security value versus the time expended, then look at the issues which require more time consuming and difficult resolutions. This means that we are focusing on firewalls, network topologies, IDSs, criteria for future OS selections, etc. Each systems administrator typically supports over a dozen operating systems, over 150 computers/CPUs, and 10 to 20 managed network switches. This makes patch management a particularly time consuming and complex task. Particular focus must be placed on scheduled maintenance windows and automation of OS patches going forward. Overall, valuable progress has been made in securing the critical systems at LIGO and the framework has been put in place to handle computing security issues going forward. As we adjust to the increased emphasis on computing security, the path going forward should become smoother. The major hindrance so far has been the lack of adequate manpower available to devote to the task. As mentioned above, the highest priority is to manage security of Laboratory Critical Systems. For this purpose, an Observatory Critical Systems Committee (OCSC) has been formed. The committee meets twice a month. Its membership represents the cybersecurity team, observatory critical systems administration, software development, diagnostics and analysis groups. The committee has established a critical systems cybersecurity installation plan and a schedule. In its first six months of operation, the OCSC has accomplished;  Network gateway hardening  Centralized syslogging  Intrusion detection and isolation  An encryption standard used to share sensitive information  An email based procedure to rapidly distribute alarms  A plan to standardize the network topology at both sites  A policy to minimize the number of operating systems and to standardize their configuration  A policy to reduce any differences between the sites' critical systems All these tasks are ongoing and have already significantly increased the level of cybersecurity at both sites.

2.9 Campus Research Facilities LIGO operates a 40-Meter prototype (1/100th the length of the actual observatories) on the Caltech campus. To prototype the Advanced LIGO optical configuration and controls and study its performance, a fully instrumented suspended-mass interferometer is needed. The 40-Meter facility fulfills this function. However, it is not possible to install full-size seismic isolation, suspensions, and optical components into the 40-Meter facility vacuum. The LIGO Advanced System Test Interferometer (LASTI) facility is designed for developing and testing full-scale advanced and improved LIGO systems, without disrupting or delaying scientific operations at the observatories. Located in a high-bay laboratory built for the purpose on the MIT campus, LASTI comprises a suite of vacuum chambers and beam tubes (with a much-reduced 16 meter baseline), seismic isolation supports, lasers, and electronic and computing infrastructure closely replicating those at the observatories.

2.9.1 40-Meter Laboratory The 40-Meter Laboratory has been rebuilt in order to fully develop and test the optical configuration and control scheme for Advanced LIGO. We are currently very close to achieving our primary goal: to acquire lock and robustly control a power- and signal-recycled Michelson interferometer with Fabry-Perot arms, and demonstrate the expected response to gravitational waves. The optical configuration of the 40-Meter interferometer mimics the one planned for Advanced LIGO: high finesse Fabry-Perot arms (1235, to be compared with the Initial LIGO arm finesse of 200); correspondingly reduced gain in the power recycling cavity (in order to reduce the thermal load on the transmissive optics in the presence of higher input laser power); a mirror at the asymmetric port of the beamsplitter in order to resonantly extract the GW sidebands from the high- finesse arm cavities and thereby increase the detection bandwidth; and a detuning of the signal extraction cavity in order to enhance sensitivity at a strategically-chosen range of frequencies. The more complex optical configuration makes it significantly more difficult to acquire full lock than for Initial LIGO, so it is essential to establish a well-defined, reliable, robust and automated procedure for lock acquisition with a full prototype such as the 40-Meter interferometer. The 40-Meter interferometer is now almost fully instrumented, with a full Initial-LIGO pre- stabilized laser, ten (single-pendulum) suspended optics with digital controllers and optical lever monitors, a 13-meter input mode cleaner, a data acquisition (DAQS) system, slow control and monitoring (EPICS) system, Global Diagnostics systems, fast (16 kHz) front-end servo controls, and a next-generation length sensing and control system. The more complex length sensing system involves multiple RF sidebands, applied within an input Mach-Zehnder interferometer. An alignment sensing system remains to be fully commissioned, but it is not necessary for lock acquisition. We are able to routinely acquire lock and control the interferometer in a variety of intermediate configurations, including the power-recycled Michelson with Fabry-Perot arm (PRFPMI, the Initial LIGO configuration) and the power- and signal-recycled (ie, dual-recycled) Michelson (DRMI) with blocked arms. The detuned signal cavity causes the usual RF signals to change radically, so adding the signal mirror to the Initial LIGO configuration is not straightforward.

Our approach is the following: (a) lock the DRMI (with blocked arms) using RF signals from the beats between the carrier and sidebands; (b) transfer control to RF signals from the beats between the sidebands only, so that it will not be disturbed when the carrier resonates in the arms; (c) unblock and lock the carrier in the arms using the transmitted light, but offset from resonance (which is much easier than "catching" the RF signal from the narrow resonance in the high-finesse arm cavities); (d) transfer control of the differential arm signal (DARM) to RF, with zero offset; (e) reduce the offset in the common-mode arm signal (CARM) until the carrier is fully resonant in the arms, and transfer control to the RF signal with zero offset. We can routinely accomplish all these steps except for the last; at this stage, all five degrees of freedom are under control, but the offset-locked arms means that the neither resonant sideband extraction nor power recycling are in full effect (although we have observed both of these effects to a reduced extent). The last step is the most difficult one, because the cavity dynamics causes lock to be lost before the carrier is fully resonant in the arms. This problem can be overcome by using a dynamically changing CARM loop servo, which is currently under development and test. We are very close. Once lock is achieved in the full configuration, we will fully characterize the interferometer, including its power buildup, expected response to GWs, sensitivity, and noise budget. We will fully automate the lock acquisition, optimization, and characterization process. We will ensure that these procedures will be directly relevant for Advanced LIGO via detailed simulations of the dynamics of both the 40-Meter and Advanced LIGO interferometers. We are collaborating with the LIGO e2e group and the VIRGO Orsay group to develop a detailed time-domain model simulation of Advanced LIGO and 40-Meter interferometers. Because the Advanced LIGO optical design calls for a detuned signal cavity, RF sidebands will be unbalanced at all exit ports. This greatly increases the already serious problem of using noisy RF sidebands as the local oscillator for extracting the GW signal. Therefore, the Advanced LIGO design will employ a DC (homodyne) detection scheme, in which a controlled amount of filtered carrier light is allowed to exit the asymmetric port to serve as a less noisy local oscillator for GW detection at DC. We have developed a first-generation DC detection chain, including an in-vacuum output mode cleaner to strip all the RF sidebands and higher-order transverse modes from the output beam, and an in-vacuum DC photodiode. (Additional in-vacuum steering mirrors, mode matching telescope, beam diagnostics and readout and control systems are also required). We plan to implement and test this DC detection system at the 40 meter lab in the coming months. Further on, we are discussing the prospect of injecting squeezed vacuum into the asymmetric port of the interferometer, using a squeezing apparatus developed at the LIGO MIT quantum measurement group. This system holds the promise of reducing the quantum noise in GW detection over a range of frequencies. The 40-Meter team continues to work closely with the LIGO Controls group, the LIGO e2e simulation group, the LSC Advanced Interferometer Configurations subgroup, and LIGO Laboratory engineers and management. Graduate students, REU (Research Experiences for Undergraduates) summer students, visiting students, and visiting scientists have contributed to all aspects of the project over the last six years. In particular, REU students have made major contributions to design of the main interferometer optical plant and the length and alignment control systems, to the configuration and commissioning of the pre-stabilized laser, digital suspension controllers, suspended-mass input mode cleaner, and

19 LIGO LIGO-M050209-00-P optical lever alignment sensing systems, and to the simulation of lock acquisition and dynamics of the dual-recycled interferometers with e2e. In the past year, the 40-Meter lab has hosted fruitful long-term visits from scientists from Japan, Perth and Canberra in Australia, Glasgow, Hamburg, and Orsay. We will continue to involve students and visitors with all aspects of the project and its goals. The laboratory continues to be a popular tour site for local students, journalists, scientific visitors, and dignitaries.

2.9.2 MIT Facilities (LASTI) The LIGO Advanced System Test Interferometer (or LASTI for short) is installed in the high bay of our laboratory in MIT's building NW17. The LASTI serves as a test-bed for designs for next- generation gravitational-wave detectors and has the unique capability of housing full-size mechanical designs for the test-mass suspension and isolation systems. It is made up of standard full-scale LIGO vacuum chambers, arrayed in an “L” configuration with arm lengths of 15 meters (in contrast to LIGO's four-kilometer arms!). The goal of the MIT LASTI facility is to support the development and testing of a range of LIGO subsystems and components at full scale in an environment that is as close to that found at the observatories as possible. The present principal use of this facility is for the testing of the External Pre-Isolator (EPI) system currently being implemented at the LIGO Livingston Observatory to alleviate the problems caused by excessive seismic noise due to logging in the area. The LIGO Advanced System Test Interferometer, or LASTI, is a facility at MIT LIGO which includes full-size LIGO vacuum chambers and infrastructure. It serves to test initial and Advanced LIGO prototypes, and to explore infrastructure/detector interactions independent of Observatory pressures. In the last year in LASTI we installed and commissioned an Advanced LIGO modecleaner prototype suspension, continued our development of improved control schemes for HEPI at Livingston, and a model of the Lightwave 10 Watt amplifier was validated using our 10 Watt laser. Further, upgrades to the facility were undertaken to ensure more efficient handling of the vacuum infrastructure, and PSL modifications were performed to allow the PSL to be used simultaneously by a number of experiments. Such experiments are expected to include displacement tests of Advanced LIGO suspensions and seismic isolation, as well as pondromotive squeezing, and laser power amplification studies. In September of last year the LASTI facility took delivery of an Advanced LIGO mode cleaner suspension. This was successfully installed on one of our HAM vacuum chamber platforms. This allowed us to critique the installation procedure for such a suspension. The HAM platform was equipped with HEPI isolators which enabled full 6 degree of freedom excitation from the platform to the suspended masses. This further validated the models that are used to design the suspensions. In addition to this a new control scheme was devised that makes the control simpler and should reduce the noise injected by stages further from the test mass. The scheme allows a modal expansion to be used that uses more modes than the number ordinarily allowed, by taking advantage of very accurate models and an estimator configuration. So far this has been successfully performed in the vertical direction using one sensor to measure the motion of 3 mirror stages. This research represents a part of student Laurent Ruet’s PhD controls/mechanical engineering thesis.

LASTI tested improvements to the HEPI system continued. The highlight of this was to use actuation correction to successfully remove the effect of actuator induced bending of the support structure from adding tilt instabilities to the horizontal geophones. This allowed significantly improved broadband isolation to be achieved. Extensive characterization has been made of the LASTI coupled mechanical system, consisting of the vacuum chambers, seismic isolation piers, and concrete facility foundation. This is to address issues of amplification of ground noise seen, both at LASTI and at the Observatories, on the tops of the seismic piers, and is helping in the refinement of the HEPI tuning at LLO and in the design of the Advanced LIGO seismic isolation system. This is a good example of the value of having a test setup of the mechanical system away from the operating instruments.

2.10 LIGO Scientific Collaboration The LIGO Scientific Collaboration (LSC) is the body of scientists, engineers and others committed to carrying out the scientific program of LIGO. It has about 500 members, some from the LIGO Laboratory and others from 40 institutions in and outside the United States. The LSC has its own governance, but reports to the LIGO Directorate (which consists of the Director and Deputy Director of the LIGO Laboratory plus the LSC Spokesperson.) Discussions during the past year have led the LSC and the LIGO Lab to restructure their relationship to allow for closer collaboration. Among the changes have been: including the LSC Spokesperson as an official member of the LIGO Directorate as mentioned above; expansion of the LIGO Oversight Committee to include members from key LSC “stakeholder” institutions; and the creation of Visiting Associate in LIGO appointments at Caltech for key LSC members, to allow them to take positions of responsibility in LIGO operations and especially for the Advanced LIGO project organization now being put together. The LSC has also taken on primary responsibility for the regular review of progress reports and annual work plans of its member groups. A new LSC Charter has been drafted to reflect these changes. During the past year, the LSC has completed several papers reporting the results of the analysis of data from the second LIGO Science Run (S2.) One paper, on the search for known pulsars, was published in Physical Review Letters. Three other papers on various S2 analyses have also been submitted for publication, and we have also completed our first paper on the third science run – a new upper limit on the stochastic background of gravitational waves. Data analysis projects have expanded to include a larger selection of potential sources than were searched for in S1. We now are able to search for known pulsars in binaries, and for unknown pulsars anywhere in the sky. Inspiral searches have produced a final result in a search for black holes in various mass ranges; good progress has been made even on the case of black holes of significant spin. We have submitted an upper limit on gravitational waves from the gamma ray burst GRB030329, and have learned how to automate this process so that large lists of GRB’s can be studied. Reports on these data analysis projects were given at GR17 in Dublin in July 2004, at GWDAW8 in Annecy, France in December 2004, at the American Physical Society meeting in Tampa in April 2005, and at the Amaldi 6 meeting in Okinawa in June 2005. During the past year, the LSC launched a major contribution to the American Physical Society’s celebration of the World Year of Physics, 2005. Called Einstein@Home, it will allow owners of computers around the world to volunteer their unused compute cycles for the purpose of analyzing data from the LIGO and GEO interferometers. The project builds on the technology developed for

21 LIGO LIGO-M050209-00-P the popular SETI@Home screen-saver program. The system is now in full operation, and has almost 50,000 users. The LSC has made good progress in collaborative data analysis with several external projects. We have carried out joint analyses with the Japanese TAMA project in searches for binary inspirals and for gravitational wave bursts; papers will report these results soon. We have also moved forward with a joint analysis with the Italian AURIGA resonant-mass gravitational wave detector. Joint work with the Italian-French Virgo interferometer has made good progress at the technical level, and prospects are excellent for joint data analysis when Virgo comes on line in the next year or two. The LSC met in August 2004 at the Hanford Observatory, in November 2004 at MIT, in March 2004 at the Livingston Observatory, and in June 2005 at the University of Michigan. Our next meeting will be in August 2005, again at Hanford.

2.11 Astrophysics and Data Analysis The LSC is responsible for organizing and conducting the scientific research mission of LIGO. Members of the LSC have developed data analysis algorithms and software to search for astrophysical signals in the data collected by the LIGO interferometers, using computing resources provided by the Laboratory (i.e. LDAS), by LSC institutions through independent grants, or by outside sources. The LSC is vigorously engaged in analyzing the data from the science runs conducted so far, to search for astrophysical signals and to improve the algorithms in preparation for the higher-sensitivity data to be collected during future science runs. The LSC has established four working groups, which are organized around different kinds of the astrophysical signals that may be detected by LIGO, distinct in fundamental characteristics and in data analysis requirements. LIGO Laboratory scientists (from Caltech, MIT, and both observatories) are active in all four working groups, and serve as co-chairs for three of them. One paper on a search for pulsars has appeared in the prestigious journal Physical Review Letters. Four other papers have been submitted to the journal Physical Review D. They include two searches for burst signals (one search asssociated with the strong gamma ray burst GRB030329, and the other a general search throughout all of S2, the second science run), and two searches for signals from compact binary systems (one involving pairs of neutron stars, and one for low-mass black holes.) In addition, we are in the final stages of preparing for publication two more papers: a search for bursts during S2 using combined data from LIGO and from the TAMA 300-meter interferometer in Japan, and a first result from our S3, the third science run, on a search for a stochastic background of gravitational waves. We have many active data analysis efforts, and will be continuing to present more results over the coming months, including several from our most recent Science Run, S4, which took place in February and March of this year. And, sometime later this year, LIGO will start a long science run (6 months duration or longer) devoted to searching with instruments that have now come very close to their design sensitivity.

2.11.1 Search for Binary Inspirals The LSC Inspiral Analysis Working Group aims to detect gravitational waves emitted by a pair of compact objects (neutron stars or black holes) orbiting each other at relativistic velocities. As energy and angular momentum are carried away by gravitational radiation, the orbital distance and orbital period shrink. The emitted signal is a "chirp," rising in frequency and amplitude, which can be calculated exactly (for low mass systems) or parameterized approximately (for higher-mass systems). In either case, Wiener matched filtering is used to search for such signals in the data. Binary inspirals are of interest for population and stellar evolution studies, as well as for understanding gravitational radiation processes, and are considered a likely candidate source of short-duration gamma ray bursts. During FY 2005, the Inspiral Group completed searches for binary neutron star inspirals using data from the S2 science run. The S2 data was divided into two subsets, depending on what interferometers were collecting low-noise data at any given time, which were analyzed separately. The first analysis used times during which interferometers were operating at both LIGO sites and required coincident candidate events. The interferometer sensitivity at both sites was sufficient to reach beyond the Milky Way to the nearby large galaxies M31 (Andromeda) and M33 (Triangulum). No candidate events were found in excess of background, yielding a limit of 47 per year (at 90 percent confidence) on the rate of binary neutron star inspirals per Milky Way equivalent galaxy. A paper describing this analysis has been submitted to Phys. Rev. D. The second analysis used times when only one LIGO site was collecting low-noise data, but was done jointly with colleagues from the TAMA300 gravitational wave detector in Japan, which collected data at the same time, to require coincident candidate events at two detector sites. This analysis had somewhat poorer amplitude sensitivity but a longer observation time, resulting in net rate sensitivity comparable to the LIGO-only analysis; a paper describing this analysis is being prepared. Two additional inspiral searches have been performed using the data from the S2 run. The first was a search for inspirals of pairs of low-mass "MACHO" objects in the galactic halo, which could potentially be primordial low-mass black holes. The search detected no such objects and placed a limit on the rate of such inspirals, as described in a paper accepted for publication in Phys. Rev. D. The second additional search uses the "detection template family" of Buonanno, Chen, and Vallisneri (Phys. Rev. D 67, 024016) to search for binary black hole inspirals, for which the waveforms are not exactly known. This analysis is essentially complete, and a paper is being drafted. The same types of searches are being repeated using the data from the S3 and S4 science runs; at present, the analyses using S3 data are well advanced. A great deal of effort has gone into characterizing the S3 and S4 data, identifying periods of poor data quality and evaluating many potential "veto" conditions based on auxiliary channels and on waveform consistency checks. In addition, substantial progress has been made toward implementing a search for inspirals of binary black holes with non-negligible spin, using the detection template family proposed in a later paper by Buonanno, Chen, and Vallisneri (Phys. Rev. D 67, 104025).

2.11.2 Search for Burst Signals The goal of the Burst Working Group is to search for short signals whose exact waveform is unknown, so that matched filtering methods cannot be applied. Possible astrophysical sources of

23 LIGO LIGO-M050209-00-P such signals include supernovae and black hole mergers. However, the searches are designed to be general enough to detect other unanticipated sources as well. The analysis of 10 days of triple-coincidence low-noise data from the three LIGO detectors collected during S2 was completed and submitted for publication to Phys. Rev. D. No gravitational wave bursts were detected. The search has set a 90% upper limit of 0.26 events per day on the rate of detectable gravitational wave bursts at the instruments. The amplitude sensitivity of the search was evaluated using extensive simulation studies with ad hoc and astrophysically motivated waveforms derived from core collapse supernova and binary black hole merger simulations, and found to be nearly a factor of 20 better than the amplitude sensitivity of the S1 burst search. The extension of this analysis to the S3 data was also completed without the detection of any gravitational wave bursts. The set of simulated astrophysical waveforms used to evaluate the search sensitivity was significantly expanded in the course of the S3 analysis. (No formal publication of the result obtained from the S3 search was pursued, since the S3 triple-coincidence live time was about 20 percent less than S2 and the amplitude sensitivity improvement was less than a factor of two.) Burst Group members are also searching for gravitational wave bursts associated with gamma ray bursts and other astrophysical events detected by other means. A paper describing an analysis of LIGO data at the time of the major gamma ray burst GRB0505029 has been submitted to Phys. Rev. D. The cross-correlation technique used in this analysis is being extended to analyze the set of all GRBs occurring during science runs. Major emphasis was given in the near-real-time analysis of data from LIGO's S4 run in Feb 22 - Mar 23, 2004. This included the analysis of numerous auxiliary interferometric and environmental channels for the purpose of understanding detector artifacts that might be affecting the overall data quality and the search for bursts in particular. An initial real-time search for astrophysical events of high amplitude with respect to the detector's noise level was performed without yielding any candidates; a higher-sensitivity "offline" search is currently in progress. Group members have improved existing search methods and have implemented new ones. New multi-resolution algorithms for the detection of transients were developed. These methods demonstrated their applicability for efficient burst detection as well as for generic glitch finding within the context of detector characterization. They have been used extensively during the real- time analysis of the S4 data and will continue to contribute in the completion of the S4 analyses. The past year has seen great progress in jointly analyzing LIGO data together with data from other gravitational wave detectors. A complete search for high-frequency bursts was carried out using LIGO S2 data with data from the TAMA detector, in different combinations to increase the total observation time. While no GW bursts were detected, the rate limit obtained from this search was about a factor of two better than the rate limit from the LIGO-only S2 search; however, the amplitude sensitivity was somewhat worse. A paper describing this analysis will be submitted to Phys. Rev. D soon. During the S3 and S4 runs, the GEO detector collected data simultaneously with LIGO, and a joint burst search using the LIGO and GEO data is being carried out. Several algorithms for "coherent" analysis are being developed to combine the raw time series data from multiple detectors for optimal self-consistency tests and sensitivity. A joint working group of the LSC (LIGO plus GEO) and VIRGO has been exploring how best to incorporate VIRGO into the network of gravitational wave interferometers as its sensitivity improves. Final, LSC members are

24 LIGO LIGO-M050209-00-P working with colleagues from the AURIGA resonant "bar" detector to develop inteferometer-bar joint analysis methods.

2.11.3 Continuous Wave Signals The Continuous Waves Search Group analyzes LIGO and GEO data for evidence of periodic gravitational wave (GW) sources. The generators of continuous GWs may be spinning neutron stars, strange stars or quarks stars i) with a triaxial deformation, ii) undergoing precession, or iii) with r-mode oscillations. The group analyzes the data for GWs from known neutron stars, and performs all-sky searches for signals from unknown sources in our galaxy. One of several searches of the S2 data set was published in early 2005; direct upper limits on the emission of gravitational waves from 28 known pulsars (Phys. Rev. Lett. 94 (2005) 181103), the lowest of which were a strain of a few times 10-24, and less than 10-5 on equatorial ellipticity. All-sky broadband coherent (F-statistic) analyses and incoherent (Hough statistic) analyses of the S2 data set are nearing completion, with reviews finalized and paper drafts written. An S2 coherent F-statistic search for GWs from LMXB Sco X-1 also has final upper limits. Results of these three S2 searches will be sent as two companion papers to Phys. Rev. D. The Einstein@Home distributed computing effort to search for periodic GWs was publicly launched on Feb 19, 2005. Currently, an F-statistic all-sky analysis of the S3 data set, based on the coherent S2 analysis noted above, is underway, employing the 600 best hours of H1 data from the S3 run. At the time of writing, over 50,500 Einstein@Home users had reported results in the previous week. Incoherent methods "Powerflux" and "Stackslide" are being employed to analyze the S3 and S4 data in an all-sky, broadband search. These computationally less expensive (but intrinsically less- sensitive when compared to coherent analyses) methods are ideal for a rapid assessment of a new data set, and we gain confidence in the correctness of the result from the parallel analyses. The group is driving towards a hierarchical analysis of the S5 dataset: interleaved steps of coherent and incoherent passes of the data. Incoherent searches can rapidly produce candidates on large data sets for coherent follow-up under Einstein@Home. Likewise, Einstein@Home can make a higher- sensitivity assessment of the run, followed by incoherent and coherent stages. Final surviving candidates would have parameters estimated by the coherent time-domain code.

2.11.4 Stochastic Gravitational Wave Background The Stochastic Background Working Group performs research on LIGO science data in search for signature of a broadband, continuous background of gravitational waves, as would be produced by a collection of incoherent sources. Sources of a stochastic background could be cosmological— analogous to the cosmic microwave background radiation—or present-epoch astrophysical phenomena. An isotropic background is characterized in terms of its power spectrum, which in turn is typically described by the dimensionless quantity, ΩGW(f), which is the gravitational-wave energy density per unit logarithmic frequency, in units of the closure density of the universe. Because stochastic signals are expected to be quite weak compared to the intrinsic noise of an individual interferometer, the analysis technique that is employed is that of cross-correlating the outputs of pairs of interferometers. As the cross-correlation is performed over progressively longer

25 LIGO LIGO-M050209-00-P observation times, the technique becomes ever more sensitive to any broadband background signal that may lie below the noise floor of the individual detectors. The S2 science data have been analyzed to produce an unpublished upper limit for a constant 2 +0.007 ΩGW(f) spectrum corresponding to Ω0h100 < 0.018 -0.003. This value was obtained by analyzing the H1-L1 detector pair. By comparison, our published S1 result was still unphysical (i.e., Ω0 >> 1): 2 +4.6 Ω0h100 < 23 -4.6. The S3 data provide a much tighter upper limit on Ω0, and consequently, the group has decided to publish the S3 result rather than the S2 one. Presently, a mature draft of a Physical Review Letter article is under review by the collaboration and will be submitted for publication before the end of the 2005 fiscal year. The S3 paper will also provide for the first time  analyses for different spectral power indices, ΩGW(f) = Ω (f/f0) . Besides the great advance due to the improvement in detector sensitivities, a LIGO scientist developed a new analytical method for combining the data from the three LIGO interferometers in an optimal way that rejects possible environmentally induced correlations in the data from the two Hanford interferometers. This appeared in print and is listed in the publications below. Over the past year, the Stochastic Group has conducted a number of detailed investigations to characterize the environmental nature of instrumental cross-correlations between the two co- located Hanford interferometers. While a number of insights have been gained into the problem, to date, we have not succeeded in utilizing the H1-H2 pair for astrophysical observation. Both today and in the past, the primary source of instrumental artifacts has been the common-mode acoustic coupling of the two instruments that share the common LVEA hall at Hanford. Investigations have pursued the concept of trying to temporally resolve the instrumental and gravitational wave auto- correlation functions of signal and noise. The group is in the process of developing a time-domain analysis technique to complement the frequency-domain optimal Wiener filter that has been used in the past. This is at present still a work in progress. Last, during the past year, the group has embarked on a new type of analysis that is designed to be sensitive to possible foreground (astrophysical) sources of stochastic gravitational waves. Analogous to the CMB, such sources would be localized with a distribution that follows the local matter distribution in our galactic neighborhood. Potential sources include unknown/unresolved newly born rotating neutron stars, LMXBs. Whereas periodic source searches rely on a large- volume parametrized template space, a stochastic search would cross-correlate two detectors directly, looking for evidence of coherence in the cross-power. Such a spatially resolved or targeted search requires a modification of the analysis pipeline. The basic idea is to develop a temporally varying optimal filter that tracks specific points in the sky. A first analysis would consider looking at specific mass concentrations, e.g., the galactic center and the Virgo cluster. In an enhanced application of the technique, one could imagine producing a sky map of the stochastic GW sky. The motivation for carrying out such searches is that foreground sources may be detectable in a spatially resolved search that are otherwise lost in the 4 sky-averaged analysis we have performed to date. Roughly, one can see that the SNR would be diluted by a factor ~ /4.

2.11.5 Major Publications  “Search for Gravitational Waves from Galactic and Extra-galactic Binary Neutron Stars”, LIGO Scientific Collaboration, submitted to Physical Review D, gr-qc/0505041.

 “Search for Gravitational Waves from Primordial Black Hole Binary Coalescences in the Galactic Halo”, LIGO Scientific Collaboration, accepted by Physical Review D, gr- qc/0505042.  “ A Search for Gravitational Waves Associated with the Gamma Ray Burst GRB030329 Using the LIGO Detectors”, LIGO Scientific Collaboration, submitted to Physical Review D, gr-qc/0501068.  “ Upper Limits on Gravitational Wave Bursts in LIGO’s Second Science Run”, LIGO Scientific Collaboration, submitted to Physical Review D, gr-qc/050529.  “ Upper Limits from the LIGO and TAMA Detectors on the Rate of Gravitational-Wave Bursts”, LIGO Scientific Collaboration and TAMA Collaboration, to be submitted to Physical Review D.  “Optimal combination of signals from collocated gravitational wave interferometers for use in searches for a stochastic background,” Albert Lazzarini, Sukanta Bose, Peter Fritschel, Martin McHugh, Tania Regimbau, Kaice Reilly, Joseph D. Romano, John T. Whelan, Stan Whitcomb, and Bernard F. Whiting, Physical Review D 70, 062001 (2004).  “Limits on Gravitational-Wave Emission from Selected Pulsars Using LIGO Data”, LIGO Scientific Collaboration, Physical Review Letters 94, 181103 (2005)  “ Upper Limits on a Stochastic Background of Gravitational Waves,” LIGO Scientific Collaboration, to be submitted to Physical Review Letters.

2.12 Advanced LIGO R&D

2.12.1 Sensing and Control The interferometry aspects of the Interferometer Sensing and Control subsystem (including the sensing scheme, lock acquisition, and readout of the gravitational wave channel) are being pursued at the Caltech 40m prototype; refer to 40m lab section of this report for progress in this area. Work has begun on identifying the infrastructure and components for the interferometer's real-time controls, as well as the slow control and monitoring system. This includes research into new architectures for the input-output converters, and some testing of new converter devices. A survey of commercial systems for networking the distributed real-time processors has also begun. We are refining the optical layout as subsystem details and conflicts arise. We have also started an integrated 3D optomechanical layout using SolidWorks CAD models. The end-to-end (e2e) modeling for Adv. LIGO is proceeding well. (see the Modeling section); we expect to be simulating lock acquisition strategies in the next year. Efforts on systems trades (e.g. the beamsplitter size determination) and requirements/interface definition are accelerating.

2.12.2 Seismic Isolation Our design contractor (ASI) completed and delivered a detailed seismic isolation mechanical design (Figure 1), for the BSC vacuum chamber, and a cost estimate. The Lab has nearly completed a Critical Design Review of the SEI subsystem to insure that the system is still "on- track" before committing to the next major prototype build for the LASTI BSC chamber. As part of this evaluation the Lab completed an independent cost estimate. The technical aspects of the review

27 LIGO LIGO-M050209-00-P are primarily addressed by testing of the SEI Technology Demonstrator system (Figure 2.) at the Stanford Engineering Test Facility (ETF) and modeling. Test results are quite promising (Figure 3), though they fall somewhat short of the original performance goal at 10 Hz. It is likely that the review committee will deem the performance shortfall inconsequential. A recommendation to LIGO Management on whether to proceed with the BSC prototype is due by July.

2.12.3 Suspensions A joint US and UK effort to design a Test Mass (quadruple pendulum) suspension has resulted in a "controls" prototype (Figure 4b). Fabrication of the 'controls’ prototype has nearly been completed (sub-assembly is underway). The intent of this prototype is to confirm the dynamics, controls, assembly tooling/procedures and installation tooling/procedures. Delivery of the quad suspension to LASTI for installation and testing is planned for early fall. A concurrent effort to design a ‘noise’ prototype (Figure 4a) led by the UK Team (under PPARC funding) is well underway. The noise prototype is a near final design version which addresses low mechanical noise elements of the design such as fused silica fibers welded to fused silica ears silicate bonded to the test mass. In support of the "noise" prototype, our UK partners have made good R&D progress on a computer controlled CO2 laser system for silica fiber/ribbon pulling and welding. Figure 6: Quadruple-Pendulum Suspensions for the Test Mass Optics

(a) "Noise" Prototype Design (b) "Controls" Prototype Deisgn (structure not shown) (in fabrication)

Characterization, performance assessment and comparison with detailed models (Figure 5) have been completed for the triple-mass, Mode Cleaner (MC) suspension, which was installed and tested in the LIGO LASTI Facility (Figure 6).

Figure 7: Comparison of Model and Experiment Transfer Function from Mass1 Yaw Drive to Mass1 Yaw

Model (blue); Measurement, control off (green); Measurement, control on (red)

Figure 6: Triple Mode Cleaner Prototype Suspension being installed into a HAM Vacuum Chamber at LASTI

2.12.4 Optics Core Optics Down Selection: The Laboratory, and the LSC, initially carried both fused silica and sapphire as potential substrates for the Advanced LIGO test masses. After a several-year study, we have selected fused silica as the test mass material substrate. This recommendation has been accepted by the LIGO Laboratory Directorate and the LIGO Scientific Collaboration (LSC). The recommendation is based on extensive studies pursued by the Glasgow-Stanford-Hobart and William Smith-LIGO team; these studies include annealing and surface treatment of fused silica to reduce mechanical loss, annealing sapphire to reduce optical loss, and mechanical loss measurements of both materials. These studies have added significantly to the body of physical knowledge of both fused silica and sapphire. The choice of fused silica was based on both performance expectations and on pragmatic considerations of the ease and confidence of success in production of complete test masses installed in suspensions. Scanner: We have implemented the first optic scanner capable of characterizing LIGO size test masses at 1064 nm. This scanner can measure coating reflectivity, integrated surface scatter, coating absorption, and bulk absorption. LASTI Test Mass: We have taken delivery of the first fused silica test mass blank. The blank will be polished for use in LASTI. The work done on this piece represents our first efforts at fabricating a full size fused silica test mass. Silica Model: The LIGO laboratory in collaboration with Hobart and William Smith and NASA Goddard have compiled measurements of the mechanical loss in fused silica from samples spanning a wide range of geometries and resonant frequency in order to model the known variation of the loss with frequency and surface-to-volume ratio. Our model matches the data well and agrees with earlier work on the frequency dependence of the loss. This improved understanding of the mechanical loss has contributed significantly to the design of advanced interferometric gravitational wave detectors, which require ultra-low loss materials for their test mass mirrors.

2.12.5 Thermal Noise Interferometer The Thermal Noise Interferometer (TNI) test bed at Caltech has been used to directly measure thermal noise in substrates and coated optics. A new experiment for measuring non-Gaussian noise in silicate bonds is in construction. As a part of the new test set the lock acquisition has been automated All hardware has been assembled and commissioned, and we are now in the noise- reduction stage.

2.12.6 Optical Coatings The Advanced LIGO coatings present challenges both to deliver the required optical performance (low optical absorption and low scatter) and to deliver low thermal noise (due to the localized mechanical loss at the interface to the laser beam). LIGO has continued the research program, begun in 2004, to lower mechanical loss (and thus thermal noise) of optical coatings suitable for LIGO use. In this effort we are working with two coating vendors; LMA, Lyon, France and CSIRO, Sydney, Australia. Working with these vendors, we have completed a program examining the effects of titania dopant in silica/tantala coatings. We observed a factor-of-two improvement in thermal noise power obtained by adding titania. The optical absorption is slightly high (about 1 ppm), but should be brought down with further work. We also looked at a coating with xenon as

32 LIGO LIGO-M050209-00-P the bombardment ion, as opposed to argon. The mechanical loss was disappointing, with a slight increase in loss compared to the argon. Another area of study has been preliminary mechanical loss measurements made on a sample with intentionally poor stoichiometry (low oxygen content). This will be measured again after the sample is annealed at our LSC collaborators at Stanford. We also began an interesting collaboration to look at alternate coating geometries that preserve reflectivity but reduce the amount of tantala and thereby the total mechanical loss in the coating. Modeling results so far have been promising.

2.12.7 Pre-stabilized Laser The Advanced LIGO Pre-stabilized Laser (AdvLIGO PSL) is a collaboration between the Albert Einstein Institute (AEI) in conjunction with Laser Zentrum Hannover (LZH) and the LIGO Laboratory, with AEI taking the lead effort. During the past 12 months, improvements were made to the laser design that was chosen at a previous LIGO Scientific Collaboration (LSC) Meeting. This year, the work has led to a demonstrated 195 W in the full injection-locked configuration, thus meeting the main requirement of 180 W of power. The conceptual design review for the AdvLIGO PSL was held on the 24th March. Both the design requirements (LIGO-T050036) and conceptual design (LIGO-T050035-0A) generated to meet those requirements were successfully reviewed with minor modifications in both cases. The review committee recommended that the AdvLIGO PSL should proceed to the preliminary design stage.

high power stage

PSS3 I 5

medium power stage ILS1 PSS2 suspended modecleaner spatial filter cavity

PMC 1 PSS1

NPRO 1 2 3

ILS2 I 1 FSS

AOM I 3 4

FSS-A1 Diagnostic tidal feedback reference FSS-A2 cavity I 4 PMC 2

Figure 1. Schematic of the AdvLIGO PSL. Figure 1 shows the optical layout and control strategy of the AdvLIGO PSL, which is similar to that of the Initial LIGO PSL. Seed light from the proven GEO600 front-end is injected into ring resonator high power stage to form the 200-W AdvLIGO Laser. The output of which is spatially filtered by a three mirror ring spatial filter cavity (denoted PMC 1 in Figure 1) before being mode- matched into the suspended mode cleaner. A small pick-off located after the spatial filter cavity samples the output for frequency stabilization of the AdvLIGO Laser.

The LIGO Laboratory is responsible for the intensity stabilization of the AdvLIGO PSL. The relative intensity noise requirement is 2 109 1/ Hz at 10 Hz. This level of stabilization requires a minimum of 80 mA of photocurrent in order not to be shot-noise limited. To this end a small number of prototype high power photodetectors have been fabricated and tested. Long term exposure tests with detected photocurrents in excess of 250 mA have been conducted for periods longer than a month with no problems encountered to date. Work is in progress to characterize the noise of the AdvLIGO Laser.

2.12.8 Thermal Compensation Recent Thermal Compensation Research for Advanced LIGO The effects of thermal aberration on interferometer performance were analyzed in detail over the past year, organized generally around the need to choose a substrate material for the Advanced LIGO test masses. We identified several thermal effects that had required additional study and quantified them for fused silica and sapphire, and composed strategies to minimize of compensate these effects. One such effect is changes in the radius of curvature of the test mass high reflectance surfaces at high arm cavity power. High power raises a mound in the center of the optic’s face, thus increasing the radius of curvature. This can in principle be corrected with a thermal compensator acting directly on the test mass, but the power stability requirements of this compensator are at the level of 10-9/Hz. This may be achievable with an incandescent compensator like a ring heater. Inhomogeneities in the coating absorption will make irregular the heating and thermal expansion of the optic’s face. Since point absorbers are the typical inhomogeneity, this effect is likely to produce small ‘blisters’ on the optic face, a phenomenon we call thermal micro roughness. These can then scatter power out of the arm cavity. Since Advanced LIGO requires less than 75ppm loss per arm, this sets a strict limit on coating uniformity. We find from current maps of absorption of coatings on fused silica that thermal micro roughness can be within this limit. The effect of thermal aberrations on the gravitational wave sidebands as they resonate in the signal recycling cavity on the way to the interferometer output are now known to set the strictest limits on thermal compensation. Full FFT analyses of the effect are still underway, but a variety of simpler models indicate that the effective scatter of power from the fundamental carrier mode by the thermal aberrations must be kept below 0.1 Percent.

We have identified three techniques to mitigate the distortions, all relying on CO2 laser projectors. One is to simply project the pattern of a fixed mask onto the optic, as is done in LIGO now. This can achieve virtually ideal compensation, but is not adaptive — it takes more than a week to make a new mask. Another technique is to use a micromirror array to build a reconfigurable reflective mask. This has high spatial resolution and can be instantly reconfigured with computer control, though it will inject noise as the pixels are switched. We are evaluating such an array now to see if it can be used with CO2 laser radiation at the requisite power. The third technique is to use a scanning technique, but to scan at frequencies above the LIGO bandwidth, using acousto-optical deflectors instead of galvanometers. While not injecting noise in the LIGO bandwidth, this would inject a large amount of noise at the scan frequency and its harmonics, about 10 kHz and up. Whether this noise injection would be acceptable is still under review.

We have learned a great deal by commissioning thermal compensation in initial LIGO. In particular, the absorbed power now being compensated in the H1 interferometer is only a factor of two less than expected in Advanced LIGO. Though the requirements will be stricter in Advanced LIGO, this experience has made us confident that thermal compensation is feasible in large interferometers, and has taught us lessons about its control in practice. Thermal compensation is also under commission in the High Power Test Facility at Gingin, Australia. This experiment is testing a new feature to be used in Advanced LIGO, an off-axis Hartmann sensor to individually measure the thermal aberration of each input test mass. Results are expected before the end of 2005.

2.12.9 Flat Top Beam experiment One way to mitigate the effect of thermal noise in the test masses and coatings is to increase the effective laser beam diameter on the test masses. Following the initial idea of larger cross section beams, proposed within the LSC, a LIGO Caltech group initiated a test effort, soon joined by the Virgo LMA laboratory with an in-kind collaboration scheme in which LMA developed and provided the exotic mirrors and LIGO provided the test interferometer. At present the exotic mirrors have been manufactured by LMA and delivered to LIGO, the test interferometer has been built and is being commissioned with standard optics.

2.12.10 Alternative Isolation Approaches A simple, single stage, passive seismic attenuation system has been designed which could be installed below the existing HAM optical benches. The development, initially intended for new Output Mode Cleaner benches, is expected to be able to provide the seismic attenuation levels required from the SEI in all HAM chambers, thus providing a potential alternative to the baseline Adv-LIGO three-stage active attenuation system. If it was found necessary, an active loop could be implemented on the passive system for increased seismic attenuation performance. If much better attenuation performance was to become necessary for future evolutions of interferometers, a second layer of attenuation could be slipped in without modifying the first layer. The same design concept could be adapted to the BSC chambers.

2.12.11 Quantum Measurement Development Work leading to sub-quantum-limited interferometric detectors is concentrating on generation of squeezed states of light (or vacuum). Two independent approaches are being pursued: (i) Squeezing of light and vacuum has been achieved with nonlinear optical media, such as LiNbO3 crystals, in the MIT QM lab, as well as other labs worldwide; and (ii) Experimental demonstration of squeezing of light and vacuum due to the coupling of light to a mirror oscillator, so-called "ponderomotive squeezing", using a high-power interferometer with low-mass suspended mirrors is underway. Theoretical work on fundamental limitations to squeezed state generation and their use in interferometry (due to optical losses, e.g.) has been carried out, to guide concentration of effort.

2.13 Actual Costs - Lindquist/Kaufman Error: Reference source not found summarizes the costs projected for FY 2005 for Operations and R&D including open encumbrances. FY 2005 was the fourth full year of costs under this cooperative agreement. Funding provided by the NSF for FY 2005 was $32 million. Projected

35 LIGO LIGO-M050209-00-P expenses and encumbrances totals $31.7 million. Approximately $4.3 million of FY 2004 funding was carried forward into FY 2004. Funding of the cooperative agreement usually requires a few weeks or months following the start of the fiscal year, and these carry-forward funds provide a buffer. Items that were delayed into FY 2005 include major R&D equipment items (e.g., core optics components and BSC and HAM chamber components for the LASTI facility at MIT). Table 1 Projected Costs and open Encumbrances for FY 2005.

2.14 Organization - Whitcomb/Lindquist The LIGO Organization is shown in Figure XX. Changes during FY 2005 include:  Prof. Barry Barish…...

3 Work Planned for FY 2006

3.1 Hanford Observatory - Raab

3.2 Livingston Observatory - Zucker

3.3 Safety

3.4 Detector Commissioning During the first few months of FY 2006, commissioning efforts will be directed towards the stability of the detectors in preparation for the S5 Science Run. By December 2005 we expect to have initiated S5. This run is expected to last for approximately 18 months, so there will be relatively little detector commissioning during the most of FY 2006, although we may undertake some short-term tasks as warranted by the detector performance.

3.5 Data and Computing

3.5.1 Modeling and Simulation The application of the simulation for the commissioning support will continue on demand. The focus of the simulation group moves toward the development of more complete simulation of the advanced LIGO. The major task in the simulation code is the implementation of the modal model version of the fast simulation of DRM. This requires (1) formulation of the field evolution calculation, (2) more modular implementation of the modal model field and (3) actual code development of the e2e module. Along with this code development, advanced LIGO simulation framework with modal decomposition of the optical fields needs to be developed. A simpler configuration, such as a FP cavity, is useful to study specific topics like the angular instability due to interaction of the optical spring with alignment control systems. The development schedule will be decided in coordination with the hardware development schedule. We originally developed the FFT code for the design of initial LIGO, and it is neither flexible nor powerful enough to be used for the advanced LIGO study. A new version is under development that will provide crucial inputs for the design of advanced LIGO. The beam in advanced LIGO cavity will be deformed due to the thermal deformation of mirrors. An improvement in the speed of the FFT code is a high priority, and if the speed of the field evolution based on FFT is sufficiently fast, the FFT code may be embedded in e2e as an option.

3.5.2 LDAS – Lazzarini We will continue to improve LDAS performance and reliability in the next year now that it is a mature and stable product. One lesson from last year has been the significant number of enhancements requested of the system to allow for smoother engineering and science run operations across the various Tier II centers of the LSC Datagrid. It is expected that enhancement requests will continue to be important in the next year and play a significant part of the development. Major operating system upgrades are now occurring on roughly a one year cycle, so LDAS software should expect to be required to port to newer Linux and Solaris operating systems

37 LIGO LIGO-M050209-00-P within the next year. With the hardware group already investigating new 64 bit computers, there is also the possibility that the software will need to be ported to full 64 bit compliance within the next year, particularly if the principal C++ compiler (GNU) provides the necessary stable version for the 64 bit hardware selected. The TclGlobus Project is almost in a position to provide sufficient components of the application programming interface (API) to allow LDAS to begin integrating Globus based Grid Security Infrastructure (GSI) protocols for user authentication and data movement, allowing users to access LDAS systems with the same grid-based certificates that are currently used to access the large compute clusters running condor within the LSC Data Grid. The first official release of the Open Science Grid software is expected to be deployed in late June of 2005. LIGO plans to set up a small cluster at Caltech that is running the OSG release and will participate in OSG test-bed activities that carry significant potential for LIGO data analysis. Example of components expected to be of interest to LIGO are the Storage Resource Management (SRM) and Privilege Project components (VOMS, GUMS, PRIMA) since they can greatly reduce the administrative burden associated with operations of a computer resource and account management. LIGO Laboratory is discussing an upgrade with Caltech for the computing facility to permit an increased number of nodes to be accommodated. The plan is to prepare for an S5 run of during one year or longer. If the facilities can be prepared on time, LIGO would upgrade the Caltech cluster and ship the existing Caltech cluster to the observatories in order to effectively double the on-site computing capacity.

3.5.3 General Computing We plans to move to LDAP, with MIT and Livingston taking leads in working out the logistics for this process. With SUN having provided the capability to connect Solaris 10 servers and LINUX units using NIS+, this has provided more time to address the logistical issues of the transition. The GC group at the LIGO locations plan to continue working on the improvement and augmentation of; data storage, network capacity and performance, computer security, application and user support.

3.6 Computing Systems Security

3.7 Campus Research Facilities

3.7.1 40-Meter Lab - Weinstein

3.7.2 LASTI - Ottoway/Shoemaker

3.8 LIGO Scientific Collaboration

3.9 Astrophysics and Data Analysis The Inspiral Analysis Working Group will complete the analysis of the data collected during the S3 and S4 science runs, carrying out searches for inspirals from binary neutron stars, primordial low-

38 LIGO LIGO-M050209-00-P mass binary black holes, and binary black hole systems with total mass up to roughly 100 times the mass of the Sun. The search pipelines will be expanded to include GEO data and to perform coherent self-consistency checks using the data from multiple detectors simultaneously. A cooperative effort to explore methods and procedures for extending the detector network to include VIRGO is underway. Group members will also complete the implementation of software to use an expanded parameter space to search for binary black hole systems with significant spin. Near-real- time data processing will be improved for rapid analysis of the upcoming S5 data run as well as for searching for inspiral gravitational wave signals associated with gamma-ray bursts and/or with merger "bursts" and subsequent quasi-normal ring-downs. Hierarchical template search strategies will be incorporated into analysis pipelines to improve the computational efficiency of searches with large parameter spaces. The Burst Working Group will complete the search for unmodeled gravitational wave bursts in data from the S4 run, with particular attention to selecting high-quality data and rejecting environmental disturbances using auxiliary-channel vetoes. It will also complete the search for gravitational-wave bursts at the times of gamma ray bursts detected during the S2, S3, and S4 runs, as well as in association with other astrophysical events. Coherent analysis methods will be implemented and refined to simultaneously analyze data from LIGO together with GEO and VIRGO. Further improvements will be made to analysis methods and data characterization techniques. Data from the S5 run will be analyzed rapidly, with the help of additional automated software tools currently being developed. The Continuous Waves Search Group will complete analyses of the S3 and S4 science runs. The Einstein@home F statistic coherent search will finish the S3 all-sky broadband search for periodic GWs using H1 data with a 10h coherent integration time, and then make a similar but extended analysis of H1 and L1 data from the S4 run. The latter search will make use of a new metric for parameter spaces and a longer coherent integration time. Incoherent searches will be employed to complete analyses of the S4 data set. The coherent time-domain code will analyze the data from S3 and S4 science runs to search for GWs from known spinning isolated and binary neutron stars. Coherent F statistic search code (originally employed on S2 for Sco X-1), and StackSlide code modified for binary systems, will be used to analyze S4 data to search for GWs from LMXB systems. The Stochastic Group will continue to search for a stochastic background of gravitational waves. Using the data from the recently completed S4 run, the group expects to achieve a sensitivity ΩGW of less than 10-4, reflecting the dramatic gains for this type of search as the individual interferometer sensitivities are improved. When they reach design sensitivity, the LIGO detectors -6 should be sensitive to ΩGW ~ 10 , a level that is below the constraint on a cosmological background of gravitational radiation provided by the big-bang nucleosynthesis (BBN) model. The Stochastic Group will also carry out its first analysis for directionally targeted searches to look for a stochastic background from specific locations, such as the Virgo Cluster. These searches will target an astrophysical stochastic background, due, for example, to the random superposition of many weak signals from rotating stars or binary systems.

3.10 Advanced LIGO R&D We expect funding for Advanced LIGO in FY 2007 or FY 2008. Our research and development program is designed to support this program through research on critical development issues. We will also use limited equipment funds to procure long lead hardware.

3.10.1 Sensing and Control The sensing and control subsystem will focus most of its attention in 2006 on the tests to be undertaken on the 40m prototype at Caltech where a test of the unmodulated DC readout scheme will be underway. Work with the end-to-end model will proceed to understand the interferometer locking and steady-state control. Tests of digital-to-analog and analog-to-digital converters will also be pursued, to handle the dynamic range and input noise requirements for Advanced LIGO.

3.10.2 Seismic Isolation During FY 2006 we will fabricate, assemble and test a full-scale, BSC chamber version, prototype of the Adv. LIGO Seismic Isolation system. The installation will be at the MIT LASTI Facility. We will also complete the design of a full-scale, HAM chamber version and begin the fabrication.

3.10.3 Suspensions A series of incremental Preliminary Design Reviews are planned from June through the fall of 2005. The quad suspension, "controls" prototype will be installed and tested at the LIGO LASTI facility late 2005. The design of the quad suspension, "noise" prototype will be completed and fabrication will be nearing completion, with an expected delivery to LASTI in the fall of 2006.

3.10.4 Thermal Noise Interferometer Silica: We plan to examine the effects of mechanical loss versus annealing parameters, including peak temperature, ramp down time, and dwell time. This is a follow up of results from a number of groups showing silica mechanical loss can be a function of annealing. We will explore possible tradeoffs between optical absorption and mechanical loss. This follows from discussions with Heraeus, a fused silica manufacturer, relating fictive temperature (which is possibly related to mechanical loss) in silica to annealing and absorption. LASTI Polish: We will polish and coat the first full size test mass. This piece will be ultimately be used in the LASTI experiment, being integrated with a prototype suspension in design by the UK Advanced LIGO team. The polishing and coating of this piece will begin the process of qualifying vendors for Advanced LIGO fabrication. Excitation of Test Mass Modes: Via a parametric instability, the modes of the test mass can exchange energy with the light beam in higher order optical modes. We plan to develop the parametric instability model to a point where we can sufficiently design experiments which will look for evidence of parametric growth on suspended cavities with high Q, these experiments will most likely take place at the ACIGA Australia Gingin facility, but possibly on the LIGO detectors post S5 (or S6 or after the upgrade). Charging Studies: Charge build up on the surface of LIGO optics is expected in the advanced interferometer, both quasi-static (increasing the interaction with the environmental electric fields) and dynamically (directly producing noise forces on the test mass). We are beginning to construct our own Kelvin probe at LIGO/MIT to measure surface charging properties of silica, coatings, and sapphire. This will be done in collaboration with Trinity College, who may take on responsibility for this work over the next year.

3.10.5 Optical Coatings Our plans for the next year are to continue to try to reduce mechanical loss without increasing optical absorption. We are pursuing the use of alternate or additional dopants beyond the titania, possibly lutetium and cobalt, with LMA/Virgo. With CSIRO, we are looking at the effects of annealing including a carefully designed annealing profile to leave zero stress in the coating, a systematic change in different annealing parameters, and possibly using an ozone atmosphere during annealing. The ozone anneal would also be a continuation of the stoichiometry studies that we started this year, but now looking at oxygen rich coatings. Another path which may be pursued with Virgo collaborators involves optical coatings which do not use simple alternating layers of two materials, but instead layers of varying thickness, to try to minimize the mechanical losses while maintaining suitable optical properties. Thermal Noise Interferometer: We plan to measure broadband thermal noise of titania-doped tantala/silica coatings. We have also ordered small samples from the same coating run in order to measure the thermal conductivity and thermal expansion coefficient of the coatings. These parameters will allow us to predict coating thermoelastic-damping noise.

3.10.6 Pre-stabilized Laser Close collaboration with the Max Planck group contributing the laser will continue. At Caltech, a test setup to demonstrate the required intensity stabilization will be assembled, using in-vacuum photodiodes and pre-cleaning of the mode of the laser beam.

3.10.7 Thermal Compensation Modeling of the thermal compensation system, both for the as-built initial LIGO compensation (to test the model fidelity), and for Advanced LIGO, will continue. Experimental demonstrations of several approaches to non-homogeneous complementary heating will be undertaken, to identify the best balance of flexibility and simplicity.

3.10.8 Flat Top Beam Experiment The large-cross-section flat-top profile beams are predicted to introduce Thermal Noise improvements as large as 2 or 3. A feasibility study is being made in collaboration with Virgo to explore the possibility to design a dedicated Thermal Noise interferometer capable to measure this so far theoretical improvement.

3.10.9 Alternative Isolation Approaches We will support the installations of the anti-spring suspension in the Japanese TAMA 300 interferometer and pursue design refinements for other applications of the approach.

3.10.10 Quantum Measurement Development On the nonlinear crystal-based squeezing experiment we are nearing completion of our second squeezer with some major modifications to mitigate noise couplings encountered in our first attempt, and to design a system that is compatible with generating squeezed vacuum that can be injected into an interferometer and at the same time deriving error signals for locking the optimal

41 LIGO LIGO-M050209-00-P quadrature. This latter goal is achieved by construction of a long (few 10 cm) OPO cavity capable of intracavity modulation, the resulting RF sidebands and their harmonics are used to generate PDH-like signals for locking. In the coming year we plan to perform an interferometer test with an injected vacuum state locked to the optimal quadrature. The ponderomotive interferometer is in the construction phase. We are presently designing the suspension and control system for the low-mass miniature mirror oscillator. We are also designing the control electronics for the interferometer locking and readout system. In the coming year we hope to test a high-power, high-finesse cavity en route the full-fledged two-cavity interferometer test.

3.11 Budget Request The total funding request for LIGO Operations for FY2002 through 2006 is $158.0 million, reduced from the proposed amount (NSF Proposal 0107417) of $160 million. The proposed period of performance is October 1, 2001 through September 30, 2006. The effort covered by this request includes LIGO Operations as well as Research and Development on important advanced detector technologies. The period of performance for this budget request is FY2006 beginning October 1, 2005 and ending September 30, 2006. The total requested is $32 million reduced from the original request of $33 million submitted September 2001. Of the $32 million total requested, $3 million is to support research for the development of advanced detectors. This includes supplies and materials as well as the salaries of the incremental support staff working directly on these tasks. $29 million is for Operations. However of this $29 million, $0.75 million will support R&D in the form of prototypes and operation of the campus research facilities at Caltech (40-Meter Facility). This conforms to proposed budgets and discussions with NSF proposal review panels. LIGO activity is ongoing at four separate sites: Caltech, MIT, the Hanford Observatory in Hanford, Washington, and the Livingston Observatory in Livingston, Louisiana. A staff of 23 scientists, engineers, graduate students, undergraduates, and support personnel is proposed at MIT. This proposal for $32 million includes the effort at all four sites. MIT effort is identified as a subaward, and the MIT budget appears as a single entry in the G5 Subawards line. The MIT proposal is included as an Appendix. Actual salaries have been used where available to project direct costs. Staff Benefits Rate as of the Indirect Cost Rate Agreement negotiated with the Office of Naval Research September 14, 2004 (excludes Undergraduate and Graduate Student Salaries): 26.0 percent. Participant budget has been established for all sites to support Outreach activities as well as the Caltech Summer programs providing research opportunities for undergraduates (SURF). Graduate Research Assistantship (GRA) Benefits: Institute Policy is to provide each graduate student employee who meets a required average workweek with full tuition and fees. A portion of this cost is requested as a benefit (exempt from indirect costs) equivalent to 63.5 percent of the graduate research assistant stipend effective September 25, 2003. This rate is applicable to all federal grants and contracts, and all other awards that provide full indirect cost recovery. The GRA Tuition Remission Benefit for all non-federal awards (gifts, grants, contracts) that do not provide full indirect cost recovery is 85 percent of GRA salary.

Indirect Costs Indirect Cost Rate Agreement Agency and Date: Office of Naval Research, 8/31/00 (rates in effect at time of the proposal in September 2001). On-Campus Overhead Rate: 58.0 percent MTDC. Off-Campus Overhead Rate: 26.0 percent. Excludes Equipment, JPL Work Orders (when work is performed at JPL), Subcontract amounts in excess of $25,000, and GRA Benefits. Table 2 FY 2006 Budget Request by Proposal Budget Line and by Site

Table 3 FY 2006 Budget Request by Top Level WBS Element

3.12 Proposed Staffing Levels Table 4 summarizes the proposed Full Time Equivalent (FTE) staffing levels at the various sites. The proposed staffing is consistent with staffing levels budgeted for FY 2005 decreased slightly to reflect reduced funding.

Table 4 FY2006 Proposed Staffing by Category

3.13 Meetings Scheduled for FY 2006 The next annual review of LIGO is scheduled at Caltech from Wednesday through Friday, November 9-11, 2005. The LIGO Scientific Collaboration will hold full collaboration meetings in March 2006 and August 2006, and will hold data analysis and results meetings in November 2005 and June 2006. The LIGO Lab and the LSC will support the following meetings though member participation and representation on scientific organizing committees:  Marcel Grossmann Meeting (St. Petersburg)  Gravitational Wave Data Analysis Workshop (Brownsville, Dec 2005)  Gravitational Wave Advanced Detector Workshop (Elba, May 2006)  Workshops (TBD) at the Penn State Center for Gravitational Wave Physics The LIGO Program Advisory Committee will meet in December 205 and May 2006. The LIGO Educational Program Advisory Committee will meet in May 2006. LIGO will be represented at the Gravitational Wave International Committee meeting, to be scheduled in summer 2006.

4 Appendix A -- Organizational Partners The LIGO Project is a collaboration between the California Institute of Technology (Caltech) and the Massachusetts Institute of Technology (MIT). Caltech has primary responsibility for the project under the terms of the Cooperative Agreement. The vehicle for authorizing work at MIT is a subaward contract issued by Caltech. The LIGO Scientific Collaboration (LSC) is the body of scientists, engineers and others that will exploit the initial detector and is pursuing the development of second-generation detectors. The LSC is committed to carrying out the scientific program of LIGO. It has over 500 members, some

44 LIGO LIGO-M050209-00-P from the LIGO Laboratory and others from 30 institutions in and outside the United States. The LSC is separate from the LIGO Laboratory and has its own governance, but reports to the Laboratory Directorate for final approval of its research program, technical projects, observational physics publications, and talks announcing new observations and physics results.

5 Appendix B--Educational Outreach Activities

5.1 Hanford Observatory Education and Outreach Project Activities We project that LIGO Hanford (LHO) will host at least 2600 visitors during 2005, an increase of 500 from our 2004 total. Roughly 60 percent of the visitors are in the category that we call ‘general public,’ 20 percent come from schools and the remaining 20 percent are a combination of visits from college groups, school teacher groups and professional scientists. This year we are asking visitors to fill out a brief customer satisfaction survey at the close of their activity. So far 55 percent of our student respondents have marked that they learned “a lot” about LIGO during their field trip, and 51 percent characterized their field trip as “very worthwhile.” 57 percent of the general public respondents marked that they learned “a lot” about LIGO while here, and 70 percent remarked that they would “certainly” come to LIGO for another activity or recommend our activities to others. Analysis of the survey results continues to drive improvements in our visitor programs – better hands-on activities for students, improvement of age-appropriate presentations for school groups of different levels, more complete driving directions and more effective publicity of our events in the community. Our main outreach activities continue to be monthly drop-in public tours, privately scheduled tours for groups larger than 15, the annual LIGO Public Lecture and public astronomy events for which we collaborate with the Tri-City Astronomy Club (TCAC). We are also offering extra monthly events in 2005 to celebrate the World Year of Physics. This year LHO and TCAC have partnered with Columbia Basin College (CBC), a community college in Pasco, for public astronomy events. CBC has opened the Robert and Elisabeth Moore Observatory on its Pasco campus, an excellent facility with a 16-inch reflector telescope. LHO, CBC and TCAC sponsored a National Astronomy Day viewing on April 16 (featuring simultaneous activities at LHO and the Moore Observatory), and on May 21 the trio offered “El Cielo en Mayo” at the Moore Observatory, a Spanish language public viewing at which a team of volunteer translators assisted the astronomers. LIGO presented gravitational wave astronomy alongside the optical astronomy activities. Both spring events were hindered by bad weather but the attendance and visitor interest were excellent. We are now tracking off-site outreach contacts. LHO staff members have made over 1000 off-site contacts thus far in 2005. These have mainly come from visits to local and regional schools, from attendance at local science fairs and community festivals such as Cinco de Mayo, and from speaking engagements at local service clubs. Hanford Observatory’s Local Educator Network (LEN), the outreach advisory council, met in February and will meet again in the fall. Our 14 LEN members come from all levels of the education community (including a teacher from Portland and a professor from Montana), from the professional science community and from the business and economic development sector. Their input continues to direct the long-term goals of our outreach program, and several of the members’ organizations are important collaborators with LHO on specific outreach programs. General outreach goals for 2006:

 We will continue to develop our collaboration with Columbia Basin College and the Moore Observatory. We will ask our LEN for guidance on the possibility of seeking grant funds to implement a public astronomy program in concert with CBC that will capitalize on the obvious community interest in astronomy.  We will continue to use the public astronomy program as a means of introducing LIGO science to the public.  We will use the success of this year’s “El Cielo en Mayo” event as a launching pad for events and activities that will serve the large Hispanic community in the Tri-Cities and in outlying areas.  We will increase the volume of Native American outreach that we undertake, utilizing recent contacts with local Tribes and with Heritage University to develop and/or to contribute to programs that are valuable to these constituencies.  We will exploit the success of our World Year of Physics event series by developing a special interest theme for a series of activities in 2006.  We aspire to host at least 3000 on-site visitors in 2006.

5.2 Livingston Observatory Education and Outreach Project Activities This is a collaborative education and outreach project based on the science of LIGO and it has three main goals: (1) Establish In-service and Pre-service teacher training programs of significance; (2) Reach a broad audience of students in Louisiana and the surrounding region with vibrant science concepts’ (3) positively impact the local and regional general public on LIGO science themes. The over-arching idea is to create a seamless connection between the science conducted at LIGO and the formal education requirements of the school classroom by using interactive, hands-on exhibits, specially designed teacher professional development programs featuring these exhibits and the informal learning environment of a Science Education Center to bring rich science content into Louisiana classrooms. The four collaborative partners include: LIGO Livingston (LLO) who brings the science content to the partnership; Southern University at Baton Rouge which brings the Pre-service Teacher Training effort to the partnership; The Exploratorium of San Francisco who provides 38 interactive, hands- on exhibits and specialized training to the collaboration; and LA GEAR UP, a Louisiana state education reform agency who organizes the teacher Professional Development programs and provides formal access to schools currently struggling to meet state education standards. The Education and Outreach Program Leader at LIGO Livingston, a certified teacher of Physics and Mathematics in Louisiana, is currently attending The Center for Informal Learning and Schools (CILS), a two year program developed and administered by The Exploratorium and designed to advance the knowledge and practices of educators in the informal learning community in meeting the needs of teachers in the formal learning environment of schools. In February, 2005, a team of university educators along with LLO education staff attended the Exploratorium’s Institute for Inquiry to sharpen their skills in producing inquiry-based teaching and learning activities and programs for teachers. The material and skills acquired at that intensive workshop were used to formulate two In-service Teacher Professional Development programs (based on LIGO science themes) which will take place, summer, 2005. The Exploratorium delivered the first 12 exhibits to

LIGO in May, 2005. Seven of the exhibits were installed at LIGO Livingston and five exhibits were installed at Southern University at Baton Rouge. These exhibits are already being incorporated in teacher Professional Development workshops held at LIGO this summer. LA GEAR UP funded two teacher Professional Development efforts taking place this summer: SUBR will conduct project MISE and Louisiana Tech will conduct project Ripple. LIGO joined the Association of Science and Technology Centers (ASTC) in order to strengthen ties to the informal learning community. LA GEAR UP sent one High School physics teacher to The Exploratorium for a summer of intensive workshops as part of the Teacher Institute. That teacher is now working with the Professional Development group, at Louisiana Tech., on Project Ripple. The Exploratorium conducted three separate workshops in Louisiana for participants from the collaboration. This exhibit training was designed to extend understanding and application of the exhibits to the teacher Professional Development efforts. LLO created a Local Educator’s Network which met May 2005 to map out objectives for aggressively introducing LIGO science into local classrooms in the 2005/2006 school year. The collaboration established a LIGO Education Project Advisory Committee to provide continuing expert advice to the Collaboration and evaluate the overall program. The Project Advisory Committee’s inaugural meeting was held in May 2005. During the past year, LIGO hosted eight individual education-based workshops directed by various Louisiana education agencies. Presentations describing the LIGO Science Education Collaboration were made at a regional GEAR UP conference and at the April APS meeting in Tampa, FL. Seven undergraduate students accepted research positions at LIGO Livingston this summer. In addition, two local high school physics teachers accepted Research Experience for Teachers positions for the summer. Numerous school field trips (K-16) and general public tours were given at LIGO Livingston for a total of 1108 guests as of June 2005. Of this number, 415 were school children, 437 were educators, 79 were college students, and 187 were general visitors.

6 Appendix C--Publications

7 Appendix D--Other Specific Products (Teaching Aids) - Ingram and Thacker We continue to expand and refine the resources that we make available to teachers. In the spring of 2005 we took delivery of three exhibits from the Exploratorium in San Francisco, giving LHO a total nine interactive hands-on exhibits at the site. We offer a field trip handout that guides students through these activities during a visit. Our field trips always include a presentation about LIGO science from an LHO staff member and a walking tour of the site in addition to the hands-on activities. On the ‘Teachers Corner’ portion of our Web site we offer a set of age-appropriate classroom activities whose content meshes with the content of our field trip activities and with various aspects of LIGO science. We work with teachers to customize field trip experiences to match the age and science sophistication of the students. In 2005 we have hosted science students from upper and lower levels of high school and middle school as well as a pair of local geometry classes. We have used a summer teacher intern, Keith Plewman of McLoughlin Middle School in Pasco, to facilitate progress in our exhibit development, field trip activities and classroom activities. Keith also led in the design of a science poster for elementary classrooms that we are now piloting in several local and regional elementary classrooms.

Goals for teaching products 2006:  With the ongoing assistance of teacher interns, we will update the alignment our field trip activities and our classroom activities with Washington and Oregon state standard grade level expectations in science and math.  We will disseminate information about LIGO teaching resources through state science teacher conferences and in state science teaching journals in Washington, Oregon and Idaho.  We will develop a strategy for accessing and/or producing bilingual resource materials in situations where it will be of service to do so. LIGO Livingston has installed 12 new, interactive, hands-on exhibits based on LIGO science themes that are currently being used in three separate teacher Professional Development programs, summer 2005. For each of these exhibits, introductory activities, lesson plans, and post-visit activities are being developed. These products will be available online and at the Livingston site. The Livingston Education and Outreach staff member presented a discussion of our outreach effort at the spring 2005 APS meeting in Tampa, detailing our approach to bringing the exciting science at LIGO into the local classrooms. Livingston has completed the review of the Detailed Design documents for the new LIGO Science Education Center which will be constructed on the LIGO Livingston site. This center will feature over 40 hands-on, interactive exhibits and will be a significant science education resource for regional science educators.

8 Appendix E--Contributions:

8.1 Contributions within Discipline LIGO is attempting interferometric measurements over kilometer distances at precisions never before achieved. The technology required includes extremely stable solid state lasers, precision optics and coatings, careful seismic isolation and very quiet suspensions, and ultra low noise high- precision electronic controls. In addition methods must be developed for collecting, storing, analyzing, and sharing large amounts of data. Examples include: Resonant Sideband Extraction research has contributed to baseline output methods proposed for advanced detectors. New metrology methods have been developed for measuring precision polished and coated optics. Techniques developed for damping low frequency noise have been used by the Japanese in their TAMA prototype interferometric gravitational-wave detector. The development of large, optically homogeneous sapphire substrates may be used for advanced detectors. The development of reliable, high-power, ultra-stable Nd:YAG lasers will allow advanced detectors to achieve lower noise floors at high frequencies. Active isolation techniques will be used for advanced detectors and have already been used to attenuate anthropogenic noise sources at the Livingston Observatory.

8.2 Contributions to Other Disciplines LIGO research has contributed to a number of technology advances of importance to other disciplines. The development of reliable, high-power, ultra-stable Nd:YAG lasers benefits researchers outside of the LIGO collaboration, for example those working on space-based gravitational-wave detectors (e.g., LISA). Requirements for laser stability and power pushed development at Lightwave, Inc., and the custom- designed Nd:YAG 10 watt laser is now a part of their product line. This laser is in turn being used by TRW in a 100 kW laser as well as by Litton Industries and the US Navy. These lasers can be used for remote sensing, space communications, and for ultra-precision machining as well as basic research. We have also learned that the LIGO mode cleaner technology is being used for very substantial savings in the inspection process for fighter jet components. Technologies associated with the optics are also pushed by LIGO requirements. Mirrors developed for LIGO have resulted in dramatic improvements in the uniformity of polished optical surfaces, optical coatings, and the ability to measure these features. Our industrial partner, Research Electro- Optics, Inc., and a subsequent spin-off company, Advanced Thin Films, Inc., have used these capabilities for a variety of customers. The technology allows users to focus beams more tightly, to design laser gyroscopes with greater precision, to make better images for high-resolution telescopes. For advanced detectors we are developing sapphire crystals in collaboration with Crystal Systems, Inc., that are larger and more perfect than anything manufactured to date. Large, optically homogeneous sapphire substrates can be useful for better windows for a range of military applications where resistance to abrasion, durability, and optical quality are important. Methods to mechanically bond complex optical assemblies have been refined by LIGO and collaborators and are being applied by NASA to designs for space-based interferometric sensing devices for the LISA development effort. X-ray telescopes provided the technology for our core optics components, and the results of our second generation effort will in turn be useful for the next generation of x-ray telescopes. The metrology is also of interest for defense applications. LIGO has the largest ultra-high vacuum enclosure in the world, and it was built at exceptionally low cost and with no leaks in a close collaboration with Chicago Bridge and Iron. The fabrication techniques--development of low-outgassing steel, high reliability economical welding techniques, and environmentally friendly cleaning processes and testing methods--can be used in large-scale processing and power plants as well as in vacuum systems for research. The 10 miles of vacuum pipe is aligned to a quarter of an inch, pushing GPS m3easurement technology to new limits of precision. The next generation of lithography for the manufacture of integrated circuits will require extreme stability in the platform used. The LIGO active isolation techniques currently being developed are being applied to this problem. Precision attenuators are of interest for isolating precision instrumentation like electron and tunneling microscopes, precision crystal growing kilns, etc.

Improved and alternative methods for thermophysical property measurements have broad applications in material science. The directed-beam actuator being developed for possible application in advanced interferometers may find use in optical metrology, medical lasers, or remote sensing. Modern servo-control techniques have been pushed (Pentek Inc. is our principal industrial partner) to new limits in the quality and speed of analog-to-digital and digital-to-analog conversion. For advanced detectors, the seismic isolation systems will use multiple, cascaded control systems, and leading control experts at Stanford and MIT are engaged in development. Our collaborative controls efforts have led directly to advances in products made by TRW, Inc., who manufacture the equipment used to fabricate some of the highest density integrated circuits.

8.3 Contributions to Human Resource Development There has been a noticeable increase in the pool of qualified post doctoral scientists over the past few years. The Visitors Program, partially funded under this cooperative agreement and partially by a separate NSF grant, has been a major factor in the increased interest in the field of gravitational-wave physics.

8.4 Contributions to Resources for Research and Education LIGO is becoming a world class facility for scientific research and education. The LIGO Scientific Collaboration (LSC) made up of a significant number of university groups and institutions from the US and the rest of the world is discussed elsewhere in this report.

8.5 Contributions to Resources for Research and Education at Hanford Observatory LHO represents the LIGO Laboratory in contributing to a proposal for a national program for high school teachers and students. The “Interactions in Understanding the Universe” proposal (I2U2) seeks to build on the success of QuarkNet. LIGO is one of several large projects that will support the overall effort by authoring data access and analysis tools, and by piloting the classroom use of these with a group of local science teachers. The proposal’s intent is to bring school students and members of the public into direct contact with authentic data. Several LHO scientists have been involved in the development of Einstein@Home, an outreach effort of the LIGO Laboratory and the LSC. Goals for Resource Development, 2006:  Develop successful deliverables for LIGO’s I2U2 commitment (contingent upon the success of the proposal).  Continue to promote the use of Einstein@Home as part of our public outreach.

8.6 Contributions to Resources for Research and Education at Livingston Observatory The vibrant mixture of activities among the collaborative partners describes an unfolding effort, rich in benefit to the science education community. We are developing a seamless approach to marry informal learning in the science center to formal learning taking place in school classrooms.

The necessary training and education completed (both formal and informal) is foundational of our approach. Hands-on, interactive exhibits were purchased from the Exploratorium which demonstrate science themes that are incorporated in the LIGO experiment. These exhibits have already been linked to LA state required learning objectives and grade-level expectations. We are currently developing a menu of activities and demonstrations which emphasize selected science themes and incorporate the use of the hands-on exhibits. Teachers, at various grade levels, will be able to choose the math or science educational objectives for their visit, and using our website, download valuable pre-visit and post-visit lesson plans, demonstrations and activities that will tie in with their visit objectives.

8.7 Contributions beyond Science and Engineering We have instituted a summer teacher's program at the Hanford and Livingston Observatories. High school and college teachers and some of their students have spent summer breaks working alongside LIGO researchers in the laboratories, learning about the work currently being done towards the detection of gravitational waves. They have taken their newly acquired scientific knowledge back to their classrooms, nurturing in their students an interest in science in general and in gravitational waves in particular. This is a valuable practice funded partially under this grant. John Thacker and Dale Ingram were teachers who participated in this program during the summers at the LIGO Livingston Observatory and the LIGO Hanford Observatory respectively. They have since been hired by LIGO as Outreach Coordinators for the sites. During FY 2005, general interest in the study of gravitational-waves, and the work of the LIGO Laboratory in particular, continued to attract strong attention from a wide range of academic communities. Both professional and amateur audiences alike have shown a deep fascination for the efforts underway at the laboratory. Of course the science of LIGO, and its daring ambitions, will always be its most compelling feature. But there has also been lively, somewhat unexpected, interest from indirectly related fields, areas where the excitement of LIGO has radiated outward to capture fresh territories of the imagination. It seems LIGO isn't just for physicists any longer.

8.7.1 A HISTORICAL RETROSPECTIVE OF LIGO'S CONSTRUCTION Early in 2005, journalist Trudy Bell, a former editor of "Scientific American" and one of the founding editors of "OMNI" magazine, described how by sheer "dumb luck" she was the sole press correspondent at a Penn State conference titled "The Future of Gravity-Wave Astronomy." A great part of the program, said Bell, was devoted to LIGO, "a pioneering venture to detect gravitational waves – which are not electromagnetic at all but warpages in space-time that travel at the speed of light." "The more I listened to the physicists and astronomers talk," Bell continued, "the more I realized there is a stupendous story in the construction of LIGO spanning every engineering discipline - along the lines of David O. Woodbury's famous book 'The Glass Giant of Palomar' or Gay Talese's memorable recounting of the building of the Verazzano Narrows Bridge (the longest suspension bridge in the world). For example, the laser beams shoot down the highest-volume ultrahigh- vacuum chambers on the face of the planet – a meter wide and 4 km long. Building those and *making sure their welds were airtight* was a monumental effort; they were assembled from spiral steel only 3 mm thick, that was stretched out like a soda straw 60 feet long (the longest thing that can be trucked in the U.S.), and butt-welded on site. Because LIGO is designed to be sensitive to

51 LIGO LIGO-M050209-00-P frequencies of roughly 50 to 5000 Hz, there was a major mechanical engineering challenge of building everything in each facility so its natural resonances fell outside that range. There's a huge story in the design of seismic and thermal isolation of the whole chamber, in the real-time control system to stabilize the mirror optics (which hang like pendulums), and in eliminating noise from thunderstorms three states away... [A]s far as I can tell no one has written or is planning any article recounting the audacious [ingenuity] of constructing something three miles long capable of sub- neutron precision measurement of unknown radiation." Bell became animated by the goal of explicating LIGO from a purely engineering perspective, viewing the laboratory as largely a collection of tremendous challenges in civil, electrical, mechanical, and vacuum construction. She solicited her proposal to "The Bent," a quarterly magazine for the multidisciplinary engineering honor society Tau Beta Pi. Then she flew to Caltech to conduct extensive interviews with relevant task leaders, after which she performed an exhaustive search of the LIGO construction photo archive. She selected nearly 100 photos to illustrate her article before flying to Washington State for a tour of the Hanford Observatory and further interviews. To create an in-depth, fully-illustrated historical retrospective of LIGO's construction is a highly ambitious goal, and at this time Ms. Bell's feature is still in development. But once complete and published, plans have been made to spotlight it on the LIGO web-pages, and make it permanently available there to anyone whose interest is not only in LIGO's future, but also in its history.

8.7.2 THE SOCIOLOGY OF GRAVITY WAVE RESEARCHERS In Autumn 2004, sociologist Harry Collins published his long-awaited book "Gravity's Shadow - the Search for Gravitational Waves." The main focus of the book is not so much the science of LIGO as the scientists of LIGO, or more broadly the technique of "doing science" as practiced by those involved in gravity-wave research. Collins's goal is to elucidate the sociological dynamics of scientists at work in a large-scale project like LIGO. He cites that in a span of three decades, scientists have claimed to detect gravitational waves more than a dozen times, but that in each case the claims were subsequently demolished by ongoing scrutiny. Collins grew interested in the fundamental precepts underlying scientific research: issues of methodology, definitions of truth and corroboration, the practice of dispute and refutation, how consensus is built and destroyed, and the various ways that scientists interact with their work and each other. Collins has been following gravitational-wave research since 1972, interviewing the scientists, institutions and government agencies pushing the work forward. Collins's book, Gravity's Shadow, looks at core questions of scientific knowledge and the nature of expertise, while recounting the four-decades long effort to detect gravitational waves. It is a significant contribution not only to sociologists and historians of science but to the advance of scientific inquiry itself.

8.7.3 EINSEIN@HOME In a happy partnership of computer networking, gravity-wave research, and educational outreach, the Einstein@Home program is now in full operation. Einstein@Home works with volunteers all across the world who donate a portion of their home computer's resources to help analyze data from gravitational-wave science runs. In its search for gravity wave sources, LIGO must roam through a range of frequencies at every point in the sky. This quickly compiles a mass of data so large that most computers would take

52 LIGO LIGO-M050209-00-P years to process it. Einstein@Home divides the challenge into manageable chunks through "distributed computing," which has the potential to increase LIGO's computer power exponentially. Based on the same principle as the SETI@Home software, volunteers are asked to download a special screen-saver program to their computer that runs only when the machine is idle. The down time is used to analyze a small bloc of the data collected by the interferometers, with the results sent back to LIGO the next time the user connects to the Internet. The screen-saver program shows a rotating diagram of the celestial sphere, and a small orange search marker indicates the section of sky being analyzed. The program does not affect computer performance since it runs only when the computer is idle or when the user turns it on. Launched in February 2005, the Einstein@Home program is already linked to over 100,000 users, with more joining daily. For LIGO, this vast and growing network amounts to a virtual "super- computer" that powerfully augments our computational strength. We also have the opportunity to share the excitement of our mission with thousands of participants worldwide. Among the benefits to the user, apart from the educational exposure, is knowing that he or she is contributing to one of the great scientific efforts of our time, and the slight possibility of being recorded in history as the person on whose computer a gravity wave was first detected.

8.7.4 LIGO IN THE AMERICAN MUSEUM OF NATURAL HISTORY Also during this period, the Science Bulletins team of the American Museum of Natural History created an elaborate web-site designed to showcase every aspect of LIGO's operation and technology. The site is rich in both educational and entertaining detail. To start, there are four full- length feature articles covering key aspects of the LIGO enterprise. First is "Newton vs. Einstein vs. the Next Wave," which describes our evolving theories of gravity in scientific history. Next is "Waiting for Gravity at LIGO," which explains the great lengths researchers must go to in order to snare a gravity wave. Then comes "A Rogue’s Gallery of Gravity-Makers," listing some of the most likely astrophysical sources of the gravity waves LIGO hopes to detect. Finally there is "LIGO's Extended Family," which details some of the other gravity-wave interferometry projects throughout the world, teammates in training with whom LIGO will one day partner in an envisioned global dragnet of research. Additional to these essays, the American Museum of Natural History site offers an extended movie-clip which includes a discussion and video images of the LIGO components; an on-line virtual tour of the laboratory; and even an interactive "working" interferometer which allows visitors to manipulate the beam splitter and see directly how LIGO operates. http://sciencebulletins.amnh.org/astro/f/gravity.20041101/ The site also contains a page of educator's resources, where teachers can help students probe deeper into LIGO and gravity-wave research.

8.7.5 OPENING THE BOOK OF THE UNIVERSE And coming in autumn 2005, even people sitting at home on their couches will have the chance to become better acquainted with LIGO as they enjoy their morning coffee and blueberry scones. Dorling Kindersley publishers, the London-based firm known worldwide for its lavishly-illustrated "coffee table" books on science and technology, will be releasing a new title: "Universe." This will be an oversized, full-color hardback book that looks at the Universe and mankind's increasing

53 LIGO LIGO-M050209-00-P knowledge of it. In the section dealing with Space, Time, and General Relativity, the subject of Gravitational Ripples will be illustrated by the LIGO Laboratory. Photos and text will teach readers about the science and technology LIGO uses to hunt gravity waves, and about the deeper understanding of the cosmos this effort will eventually provide. The book is being produced in association with the Smithsonian Institution.

9 Appendix F--Training and Development

9.1 Training and Development at Hanford Observatory In the summer of 2005 LHO will host, for the second time, a two-week 3-credit WSU Tri-Cities summer course for teachers, “The Nature of Scientific Inquiry.” The 2004 course received positive feedback from the teacher participants, who felt that LIGO’s contribution to the course added significant value to their learning and to the overall quality of the experience. LHO LEN member Dr. Judy Morrison of WSU Tri-Cities is the instructor of record. LHO continues to host one-session visits by Dr. Morrison’s Teaching Methods classes during the academic year during which the prospective teachers learn about LIGO and the resources we provide for science education. In June 2005 we will also host one-day visits by two groups of teachers under the auspices of Eastern Washington University. Goals for Training and Development, 2006:  Integrate LIGO more fully into the ongoing professional development programs of our local school districts.  Train a set of volunteers to assist with field trips and visitor programs here at the site and in community venues (such as the Moore Observatory) where LHO would be participating in joint public activities.

9.2 Training and Development at Livingston Observatory Livingston Observatory’s Education and Outreach Program Leader is participating in the Informal Learning Certificate program at Exploratorium’s Center for Learning and Schools. The program includes four institutes --- Inquiry, School Sense, Learning Theory, and the summer San Francisco Bay Area Institute --- for a total of 17 days over a two-year commitment. University professors and LIGO staff provide In-service and Pre-service teacher professional development experiences to enrich their science content knowledge. Professional development providers have participated in three exhibit training workshops designed to insure maximum transfer of science content as it relates to the exhibits. In addition, we have seven undergraduate students participating in Summer Undergraduate Research Fellowships and two local high school physics teachers participating in our Research Experience for Teachers summer program.