1. Moriond 2003Jordan GoodmanMilagro Collaboration VHE GRBs with Milagro The Milagro Detector Why look for VHE GRBs Milagrito Result –GRB 970417a Milagro Results –GRB010921 –Future Directions Jordan Goodman University of Maryland

2. Moriond 2003Jordan GoodmanMilagro Collaboration Techniques in TeV Astrophysics Low energy threshold Good background rejection Small field of view Low duty cycle Good for sensitive studies of known sources. High energy threshold poor background rejection Large field of view (~2sr) High duty cycle (>90%) Good for all sky monitor and for investigation of transient sources.

3. Moriond 2003Jordan GoodmanMilagro Collaboration Observing the High Energy Sky 10 9 10 11 101010 10 131517 19 1 GeV1 TeV1 PeV1 EeV Satellites Fly’s Eye / HiRes Air Cherenkov Milagro EAS Arrays Solar Arrays Akeno /Auger Milagro Water-Cherenkov Detector Threshold ~300 GeV Wide-angle  /hadron Separation 24 Hour – all year operation

4. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Site Located near Los Alamos, NM, USA 8650’ Elevation 60m X 80m X 8m covered pond

5. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro 8 m 80m 50m 450 Top Layer 8” PMTs 273 Bottom Layer 8” PMTs

6. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Gamma-Ray Detector ● Altitude - 8692 ft ● Two layers of PMTs: - Top layer used to reconstruct shower direction to ~0.7 degrees. - Bottom layer used for background rejection. ● Water is used as the detection medium - allows for a large sensitive area.

7. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Site Counting House Pond Utility Building (PUB) Office Trailer Testing Trailer

8. Moriond 2003Jordan GoodmanMilagro Collaboration Milagrito A prototype for the full Milagro detector Single layer of 230 PMTs with no muon detection Milagrito operated at >250Hz from Feb 97 to April 98 (>85% livetime) More than 9 billion events - 9 Terabytes

9. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro

10. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro

11. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Outriggers

12. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Energy Response

13. Moriond 2003Jordan GoodmanMilagro Collaboration Gamma / Hadron Separation in Milagro Gammas (MC) Data Gammas (MC)

14. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Sensitivity Due to increasing energy threshold and decreasing sensitivity, we only look for GRB with zenith angles less than 45 degrees. Energy threshold is not well defined. Even though our peak sensitivity is at a few TeV, we have substantial sensitivity at lower energies.

15. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro EGRET at 100 MeV Milagro at 1 TeV

16. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro Results Data taken in the Crab Nebula region with 6  at the position of the Crab Signal map of Mrk 421 during the 2001 flare (1/17/01-4/26/01). The circle shows the position of Mrk 421 with our angular bin. The center corresponds to ~5 

17. Moriond 2003Jordan GoodmanMilagro Collaboration GRB Afterglows ● Faint, fading X-ray, optical or radio afterglows have been discovered in ~50 GRB. ● Identification of absorption or emission lines in afterglow or host galaxy allows determination of distance. Implies that GRB are enormously energetic: 10 48 -10 54 ergs. (c.f. Solar rest mass = ~2 x 10 54 ergs) Most are too distant to be observable at TeV energies - however sample suffers severe selection biases and is probably not representative of the distance to the entire class of GRB. GRB leave behind a faint declining echo at lower wavelengths. Important as it allows us to better measure the location of the source

18. Moriond 2003Jordan GoodmanMilagro Collaboration High Energy Afterglow ● In one GRB, EGRET observed emission above 30 MeV for more than an hour after the prompt emission. ● 18 GeV photon was observed (the highest ever seen by EGRET from a GRB). ● Due to Earth occultation, it is unknown for how long the high energy emission lasted. Unlike optical/X-ray afterglows, gamma-ray luminosity did not decrease with time -> additional processes contributing to high energy emission?

19. Moriond 2003Jordan GoodmanMilagro Collaboration GRB Paradigm (Piran 2001) Produce lots and lots of energy in a small region of space. Hypernova- death of a massive star merging of close compact binaries (neutron stars or black holes)

20. Moriond 2003Jordan GoodmanMilagro Collaboration Emission Models ● Series of shells produced by the central engine collide, forming shocks. ● Electrons accelerated at these shocks produce synchrotron radiation. ● Depending on the physical parameters in the emission region, there may also be a second higher energy component due to inverse Compton emission, proton synchrotron emission, or photopion reactions.

21. Moriond 2003Jordan GoodmanMilagro Collaboration Emission Models Prompt Phase (Pilla & Loeb 1998) Afterglow Phase (Sari & Esin 2001) Luminosity of the inverse Compton component is comparable to the synchrotron luminosity.

22. Moriond 2003Jordan GoodmanMilagro Collaboration What can we learn from VHE Observations? Astrophysics: ● How high in energy does the prompt GRB emission extend? Measurements of high energy cutoffs in GRB will provide information on: - particle acceleration. - Bulk Lorentz factors at each internal shock. ● Is there a second emission component? What is its nature? ● How common are high energy afterglows such as that seen in GRB940217? Physics: - Probe density and spectrum of IR/optical intergalactic radiation fields. - Test of Lorentz invariance at high energies (quantum gravity...).

23. Moriond 2003Jordan GoodmanMilagro Collaboration Lorentz Invariance Violation Bounds on energy dependence of the speed of light can be used to place constraints on the effective energy scale for quantum gravitational effects.  t ~  ( E/E QG )  L/c E 2 -c 2 p 2 ~E 2  (E/E QG )  - This may be modified in some quantum gravity models. This has the important observational consequence that this will give rise to energy dependent delays between arrival times of photons. E 2 = m 2 c 4 +p 2 c 2 - in the Lorentz invariant case, The expected time delay is : This may be measurable for very high energy photons coming from large distances.

24. Moriond 2003Jordan GoodmanMilagro Collaboration Lorentz Invariance Violation Implications for GRB observations: Delay between the keV and VHE emission.

25. Moriond 2003Jordan GoodmanMilagro Collaboration Quantum Gravity - Observational Consequences/issues ● Delay in arrival of high energy photons relative to lower energy. Depends on ability to measure  t. - require high luminosity. - short lived events. - instruments with large collection area. ● Smearing (in time) of VHE emission. Duration of very short bursts (<1 s) will have larger duration at VHE energies and should show soft -> hard spectral evolution.  t ~  ( E/E QG )  L/c

26. Moriond 2003Jordan GoodmanMilagro Collaboration Absorption of TeV Photons  TeV +  IR -> e + e - -- Limits volume of observable Universe Density of IR background radiation is hard to measure due to foreground contamination. The density of the IR background is sensitive to the epoch of galaxy formation and other details of structure formation.

27. Moriond 2003Jordan GoodmanMilagro Collaboration Measuring the Intergalactic IR Background Look for absorption features in high energy gamma-ray spectra. Need a large number of gamma-ray sources. Need sources to be distributed over a wide range of redshifts. Need the sources to be bright. Gamma-Ray Bursts are ideal test sources!

28. Moriond 2003Jordan GoodmanMilagro Collaboration Milagrito - GRB 970417a Searching 54 Batse bursts (T90) One burst 970417a showed 18 events w/background of 3.46 This has a prob< 2.9x10 -8 Accounting for all search trials – combined accidental chance 1/150 This could mean TeV emission from GRBs Batse 1  error circle

29. Moriond 2003Jordan GoodmanMilagro Collaboration Milagrito - GRB 970417a

30. Moriond 2003Jordan GoodmanMilagro Collaboration GRB970417 ● 18 signal events with an expected background of 3.46 -> Poisson prob. 2.9e-8 (5.2  ). Prob. after correcting for size of search area: 2.8e-5 (4  ). Chance prob. of this excess in any of the 54 GRB examined for TeV emission by Milagrito: 54x2.8e-5 = 1.5e-3 (3  ). Evidence for a TeV signal from GRB970417 was seen by Milagrito (a smaller, single layer prototype of Milagro)

31. Moriond 2003Jordan GoodmanMilagro Collaboration GRB970417 ● sub-MeV observations show a weak, soft burst. ● Emission must have extended up to at least 650 GeV. - Highest energy photons ever observed from a GRB! ● First evidence for existence of second emission component.

32. Moriond 2003Jordan GoodmanMilagro Collaboration Quantum Gravity - Observational Consequences ● Modification of the pair production threshold -> less absorption on IR background than predicted. ● Delay in arrival of high energy photons relative to lower energy. Depends on ability to measure  t. - require high luminosity. - short lived events. - instruments with large collection area. ● Smearing (in time) of VHE emission. Duration of very short bursts (<1 s) will have larger duration at TeV energies and should show soft -> hard spectral evolution.

33. Moriond 2003Jordan GoodmanMilagro Collaboration Implications of the Milagrito Observations of GRB970417 The Milagrito observation represents the highest energy photons ever observed from a GRB, and the first evidence for a second emission component. ● Redshift: Opacity is ~1 for 650 GeV photons at a redshift of ~0.1. Thus z<~0.1. Implies that the burst must have been intrinsically weak at sub-MeV energies. ● Bulk Lorentz Factor:  > 95 (assuming variability timescale of 1 s and that the sub-MeV spectrum turns over at 60 keV). Particle acceleration: ● If the VHE emission was due to inverse Compton emission, E ic,max ~ 4/3  2 e,max E soft, then the electron energies required to upscatter 60 keV photons to 650 GeV,  e > 2000.

34. Moriond 2003Jordan GoodmanMilagro Collaboration Lightcurves Cross correlation between TeV and sub-MeV lightcurves peaks at a lag of 1 s. Assuming E obs = 650 GeV,  t = 4 s and z=0.1, we can obtain a constraint on E QG which is a factor of ~70 better than previous limits (Biller 1999).

35. Moriond 2003Jordan GoodmanMilagro Collaboration Inverse Compton Models For an SSC model F ic /F syn = sqrt(  e /  B ) ~ 5 for model fits to BATSE data. However, Milagrito result implies that F ic /F syn >10. Can enhance IC emission if there is an external source of soft photons: - from optical flash - expect TeV emission to be slightly delayed. - from pulsar left behind from a precursor supernova which may occur days to months before the GRB. Alternatively the TeV emission may be dominated by radiation produced by a high energy population of protons.

36. Moriond 2003Jordan GoodmanMilagro Collaboration GRB010921 Constraints on TeV emission are most interesting for GRB with known redshift. ● GRB010921 was detected by both the WXM and Fregate instruments on HETE, beppoSAX and IPN. ● Zenith angle of 10 degrees at Milagro ● Spectrum of the host galaxy measured by Palomar indicated that z=0.45 E -2.4 differential photon spectrum corrected for absorption on intergalactic background radiation.

37. Moriond 2003Jordan GoodmanMilagro Collaboration GRB010921 Ratio of VHE to sub-MeV fluence is less than for GRB970417. Preliminary!

38. Moriond 2003Jordan GoodmanMilagro Collaboration VHE Instrument Sensitivity For observations of the prompt phase of GRB, current and future high energy gamma-ray instruments (GLAST and Milagro) are very complementary.

39. Moriond 2003Jordan GoodmanMilagro Collaboration Milagro and GLAST Sensitivity For a 1 second observation, Milagro becomes more sensitive than GLAST at ~100 GeV.

40. Moriond 2003Jordan GoodmanMilagro Collaboration How many GRB will we see at TeV energies? Luminosity function at these energies is unknown! However, assuming that all are bright at TeV energies then the distance distribution of GRB will determine how many we see. (Boettcher and Dermer 1998) 9% with z<0.3 9/year (Schmidt 1999) 0.6% with z<0.3 0.6/year These predictions are only for long duration bursts and are very uncertain at low redshifts. Evidence that there may be a population of soft (Schmidt2001) and/or weak (Norris 2002) which are very close.

41. Moriond 2003Jordan GoodmanMilagro Collaboration Conclusions ● VHE observations of GRB will provide a crucial piece in the puzzle to understand these enigmatic objects. ● EGRET observations suggest that all prompt GRB spectra may extend out to at least 10 GeV. ● Many emission models of both the prompt and afterglow phases of GRB predict VHE fluxes which are observable by the current generation of instruments. ● VHE observations are much more interesting if the burst is localised and the redshift is known. SWIFT will provide a sample of such bursts. ● GLAST + TeV ground based instruments will provide complete spectral coverage from 100 MeV - 50 TeV of both the prompt and afterglow phases of GRB.