Kathryn M. Zurek University of Michigan

Physics at the Frontiers

Wednesday, June 1, 2011 Fermilab Frontiers

Wednesday, June 1, 2011 Watch (and Work) List

Top AFB B physics anomalies W + jets

CoGeNT/DAMA anomalies MiniBooNe Anomaly XENON and LUX below Cosmic Radiation Anomaly Higgs pole MINOS ν / ν¯ discrepancy AMS and cosmic positrons

New Physics solutions require out-of-the-box models

Wednesday, June 1, 2011 Watch (and Work) List

Top AFB B physics anomalies W + jets

CoGeNT/DAMA anomalies MiniBooNe Anomaly XENON and LUX below Cosmic Radiation Anomaly Higgs pole MINOS ν / ν¯ discrepancy AMS and cosmic positrons

New Physics solutions require out-of-the-box models

Wednesday, June 1, 2011 Cosmic Frontier

• “Discovery” of light dark matter 3

DAMA, NaI

CoGeNT, Ge

CoGeNT

Wednesday, June 1, 2011

FIG. 3: Low-energy spectrum after all cuts, prior to efficiency corrections. Arrows indicate expected energies for all viable cosmogenic peaks (see text). Inset: Expanded threshold region, showing the 65Zn and 68Ge L-shell EC peaks. Over- lapped on the spectrum are the sigmoids for triggering ef- FIG. 4: Top panel: 90% C.L. WIMP exclusion limits from ficiency (dotted), trigger + microphonic PSD cuts (dashed) CoGeNT overlaid on Fig. 1 from [6]: green shaded patches and trigger + PSD + rise time cuts (solid), obtained via high- denote the phase space favoring the DAMA/LIBRA annual statistics electronic pulser calibrations. Also shown are ref- modulation (the dashed contour includes ion channeling). erence signals (exponentials) from 7 GeV/c2 and 10 GeV/c2 Their exact position has been subject to revisions [7]. The −4 WIMPs with spin-independent coupling σSI = 10 pb. violet band is the region supporting the two CDMS candi- date events. The scatter plot and the blue hatched region represent the supersymmetric models in [8] and their uncertainties, respectively. Models including WIMPs with mχ ∼7- Fig. 3 displays Soudan spectra following the rise time 11 GeV/cm2 provide a good fit to CoGeNT data (red contour, cut, which generates a factor 2-3 reduction in background see text). The relevance of XENON10 constraints in this low- (Fig. 2). Modest PSD cuts applied against microphonics mass region has been questioned [14]. Bottom panel: Limits are as described in [1]. This residual spectrum is domi- on axio-electric coupling gaee¯ for pseudoscalars of mass ma composing a dark isothermal galactic halo (see text). nated by events in the bulk of the crystal, like those from neutron scattering, cosmogenic activation, or dark matter particle interactions. Several cosmogenic peaks are noticed, many for the first time. All cosmogenic prod- towards threshold that rejected events exhibit. A sec- ucts capable of producing a monochromatic signature are ond source of possibly unaccounted for low-energy back- indicated. Observable activities are incipient for all. ground are the L-shell EC activities from observed cos- 65 We employ methods identical to those in [1] to ob- mogenics lighter than Zn. These are expected to contain Weakly Interacting Massive Particle (WIMP) and tribute < 15% of the counting rate in the 0.5-0.9 keVee ∼ Axion-Like Particle (ALP) dark matter limits from these region (their L-shell/K-shell EC ratio is 1/8 [5]). A spectra. The energy region employed to extract WIMP third possibility, quantitatively discussed below, consists limits is 0.4-3.2 keVee (from threshold to full range of of recoils from unvetoed muon-induced neutrons. the highest-gain digitization channel). A correction is Fig. 4 (top) displays the extracted sensitivity in spin- applied to compensate for signal acceptance loss from independent coupling (σSI) vs. WIMP mass (mχ). For 2 cumulative data cuts (solid sigmoid in Fig. 3, inset). mχ in the range ∼7-11 GeV/c the WIMP contribu- In addition to a calculated response function for each tion to the model acquires a finite value with a 90% WIMP mass [1], we adopt a free exponential plus a confidence interval incompatible with zero. The bound- constant as a background model to fit the data, with aries of this interval define the red contour in Fig. 4. two Gaussians to account for 65Zn and 68Ge L-shell However, the null hypothesis (no WIMP component in EC. The energy resolution is as in [1], with parameters the model) fits the data with a similar reduced chi- 2 σn=69.4 eV and F =0.29. The assumption of an irre- square χ /dof =20.4/20 (for example, the best fit for 2 2 ducible monotonically-decreasing background is justified, mχ = 9 GeV/c provides χ /dof =20.1/18 at σSI = given the mentioned possibility of a minor contamination 6.7 × 10−41cm2). It has been recently emphasized [6] from residual surface events and the rising concentration that light WIMP models [1, 8, 9] provide a common ex- Hunt for Dark Matter

• Direct conﬂict with CDMS Ge

3 • Neutralino from MSSM not viable 4

Ε0%0.006 Ε0%"0.006

60 60 "41 "41 "41 f "42 f %1 "42 f f %!0.7"41 1. ! 10 2. ! 10 3. ! 10 "42 1. ! 10 5. ! 10n p 2. ! 10 5. ! 10"41 n p 2. ! 10 10!39 10!36 5. ! 10"42 ∆%0 keV ∆%0 keV

1. ! 10"42 50 # 50 # 3. ! 10"42 10!40 10!37

"42

" " 2. ! 10

2 40 2 40 Β Β cm cm ! 10!41 ! 10!38 tan tan n n 5. ! 10"43 "42 Σ 30 Σ 30 1. ! 10 CoGeNT 99& CoGeNT 99& !42 !39 10 DAMA 99& 10 DAMA 99& 5. ! 10"43 XENON100 90& XENON100 90& 20 20 XENON10 90& XENON10 90& CDMS 90& CDMS 90& 10!43 CoGeNT unmod QNa%0.30 10!40 CoGeNT unmod QNa%0.30 10 10 6 8 10100 12 12014 14016 18160 180 200 6 8 10 12100 14 120 16 14018 160 180 200 m GeV m GeV Χ FrandsenmA et al. Χ mA Kuﬂik, Pierce, KZ

FIG. 2: Constraints on the mA tan β plane from B τν, FIG. 3: Constraints on the mA tan β plane from B τν, + − → + − + → B Dτν and φ τ τ −. In the case of the B decays, we show B Dτν and φ τ τ −, and t bH . In the case of the → !→ " fn fp%!0.7 → ! " → fn fp%!0.7 → • Is10 !365-7a conservative GeV bound (grey mass shaded region): thewindow intersection10!36 of the 3 B decays, suggestive we show a conservative bound (grey shaded of region): the + sigma allowed regions for both∆%15B processes.keV For φ τ τ − (the intersection of the 3 sigma∆%15 allowedkeV regions for both B processes. For → + irregular red shaded region), the region below the curve is allowed at φ τ τ − (the irregular red shaded region), the region below the 2 σ by the Tevatron. The B-decay# region depends on the squark and curve→ is allowed at 2 σ by# the Tevatron. Since the B-decay region something10!37 gluino masses due to loop else? corrections to the b mass, so10 we!37 show the depends on the squark and gluino masses due to loop corrections to region corresponding to &0 =+&max. The region for &0 = &max the b mass, we show lines corresponding to &0 = &max. The region

" + " − −+

2 is shown in Fig. 3. The φ τ τ − is relatively insensitive2 to these for &0 =+&max is shown in Fig. 2. The φ τ τ − constraint is Wednesday, June 1, 2011 corrections. We also show→ in this plane contours of constant scatter- relatively insensitive to these corrections. The→ green shaded region cm cm ! 10!38 ing cross section, assuming the bound on the invisible! 10Z!38width (3.0 indicates the constraint from t bH+. We also show in this plane n n →

Σ MeV) is saturated and &0 =+&max. Σ contours of constant scattering cross section, assuming the bound on the invisible Z width (3.0 MeV) is saturated and &0 = &max. CoGeNT 99& CoGeNT 99& − 10!39 DAMA 99& 10!39 DAMA 99& XENON100 90& XENON100 90& branching fraction and production cross section in opposite di- To conclude, acquiring a large scattering cross section in rections,XENON10 even90 extreme& values of ! = ! give rise to small XENON10 90& CDMS 90& | 0| max CDMSthe MSSM90& for light WIMPs requires a very particular Higgs !40 modifications, 5%, to these curves. Examining these!40 plots, 10 CoGeNT unmod∼ QNa%0.30 10 CoGeNTbosonunmod spectrum. To achieveQNa%0.43 the largest possible cross section we can pick out the largest allowed scattering cross section, consistent with constraints, we require µ very near its bound 6 8< 10 4212 214 16 18 6 8 10 12 14 16 18 σn 5 10− cm , below the CoGeNT allowed region. at 108 GeV, sbottoms and gluino relatively light (around 350 ∼ × If the errorsm areΧ bothGeVB experiments are inflated even further GeV),m a heavyΧ GeV right-handed stop around > 1.5 TeV, and small (both experiments taken at 3.1 sigma), a fine-tuned region at A-terms. To maximize scattering, the CP∼ even Higgs boson FIG. 1: Best-fit parameterlarger tan regionsβ opens. for DAMA There the and charged CoGeNT Higgs (coloured contribution regions) is as wellwith astan exclusionβ–enhanced limits from couplings XENON10, should be as light as possi- exactly the right size to (over)cancel the standard model con- XENON100, CDMS II and the unmodulated! " CoGeNT signal. In the upper panels, ble.δ = At 0 present, keV! and bounds"fn/fp from= 1 (left)B decays and are most constraining. fn/fp = 0.7 (right).tribution, The lower such panels that the have resultingδ = 15 sum keV is and againfn/f thep = same0.7 size (left as and right). In the lower right panel we have − − Depending on the details of the SUSY spectrum, constraints taken the sodium quenchingthe standard factor model to be one.QNa If=0 this.43. strip In thewere other to open, panels the we cross take thefrom standard the rare value decayQNat=0.bH3. Notice+ could the eventually become com- upper left panel usesallowed a different cross scale section for σn is. approximatelyIt is clear that employing a factor of the 2 higher, IVDM and iDM mechanisms significantly→ weakens the constraints from null searches, and allows for a small region of agreement when the sodiumpetitive. quenching We find factor that for is WIMPsvaried within in the 5-15 GeV range, the σ < 1 10 41 cm2, and the Tevatron constraints on Higgs 42 2 n − scattering cross section must be smaller than 5 10− cm . a reasonable range. ∼ × × production would start to be relevant. Thus it appears a MSSM neutralino is in tension with the Finally, we comment on the more model-dependent flavor data from CoGeNT. To explain the observed rates in these suggest an excessphysics in the oxygen implications. band Forof 32b eventssγ, without over cancellation,of the data such taken bydetectors CRESST would could require have alocal significant overdensity in the DM of a → an estimated backgroundlarge values of 8.7 of events.tan β Lightwould DM-nuclei require chargedimpact Higgs on masses the preferredfactor region of 6 to in hit our the scenario. edge of the window. We leave for future scattering can offercloser an explanation to 300 GeV of [38]. this In excess, principle, and thethere is theThe possibility CRESST of experimentwork a discussion also offers of the the potentialeffect of a to thermal relic history on DM mass and DM-nucleonlarge canceling cross-section contributions. required However, take thistest requires the inelastic a large naturethe allowed of the parameterDM because space of of the the pres- low mass MSSM window, values close to thosecontribution required fromfor an squark/gaugino explanation of diagrams the ence (e.g. ofwith both light tungstenbut and it is oxygeninteresting in tothe note detector. that that For region near the CoGeNT DAMA and CoGeNT observations. An approved analysis stops and charginos). Such a delicate cancelationan elastic would DM be explanationwindow of gives the rise oxygen to approximately band events the a correct relic density. surprising, and might well show up elsewhere depending on how it were implemented (e.g., non-minimal flavor violation). We thank Tim Cohen and Dan Phalen for discussions. Dark Matter and the Baryon Asymmetry

In standard picture, DM abundance set by thermal freeze-out Γann ! H

What if instead set by baryon density?

Experimentally, ΩDM 5Ωb Gelmini, Hall, Lin, Barr, Kaplan, ≈ Kitano, Low, Farrar, Zaharijas, Find mechanism n n Fujii, Yanagida DM ≈ b m 5m DM ≈ p

Wednesday, June 1, 2011 Many Examples of Asymmetric DM

Integrate out heavy state Effective operators: X

c c c

Energy W = Xu d d N X M 1 GeV p ∼ Standard Model Dark Matter

Inaccessibility Luty, Kaplan, KZ

Wednesday, June 1, 2011 Sub-Weakly Interacting Massive Particles

10-39 plored WIMP parameter space, and cuts into the region DAMA/Na where supersymmetric WIMP dark matter is accessible 10-40 ] 2 by the LHC [17]. Moreover, the new result challenges CoGeNT DAMA/I the interpretation of the DAMA [19] and CoGeNT [18] 10-41 CDMS results as being due to light mass WIMPs.

10-42 We gratefully acknowledge support from NSF, DOE, EDELWEISS SNF, Volkswagen Foundation, FCT, R´egion des Pays de

10-43 la Loire, STCSM, DFG, and the Weizmann Institute of Science. We are grateful to LNGS for hosting and sup- XENON100 (2010) WIMP-Nucleon Cross Section [cm 10-44 porting XENON. XENON100 (2011) Buchmueller et al.

-45 10 6 7 8 910 20 30 40 50 100 200 300 400 1000 WIMP Mass [GeV/c2] ∗ Electronic address: [email protected] XENON † Electronic address: [email protected] FIG. 5: Spin-independent elastic WIMP-nucleon cross-section [1] G. Steigman and M. S. Turner, Nucl. Phys. B253, 375 σ as function of WIMP mass m . The new XENON100 limit χ (1985); G. Jungman, M. Kamionkowski, and K. Griest, at 90% CL, as derived with the Profile Likelihood method Phys. Rept. 267, 195 (1996). taking into account all relevant systematic uncertainties, is Wednesday, June 1, 2011 [2] N. Jarosik et al., Astrophys. J. Suppl. 192, 14 (2011); shown as the thick (blue) line together with the 1σ and 2σ K. Nakamura et al. (Particle Data Group), J. Phys. G37, sensitivity of this run (shaded blue band). The limits from 075021 (2010). XENON100 (2010) [7] (thin, black), EDELWEISS [6] (dotted, [3] M. W. Goodman and E. Witten, Phys. Rev. D31, 3059 orange), and CDMS [5] (dashed, orange, recalculated with (1985). v = 544 km/s, v = 220 km/s) are also shown. Expecta- esc 0 [4] J. D. Lewin and P. F. Smith, Astropart. Phys. 6, 87 tions from CMSSM are indicated at 68% and 95% CL (shaded (1996). gray) [17], as well as the 90% CL areas favored by CoGeNT [5] Z. Ahmed et al. (CDMS), Science 327, 1619 (2010). (green) [18] and DAMA (light red, without channeling) [19]. [6] E. Armengaud et al. (EDELWEISS) (2011), arXiv:1103.4070. [7] E. Aprile et al. (XENON100), Phys. Rev. Lett. 105, 3 and a density of ρχ =0.3 GeV/cm . The S1 energy res- 131302 (2010). olution, governed by Poisson fluctuations, is taken into [8] E. Aprile et al. (XENON100) (2011), arXiv:1103.5831. account. Uncertainties in the energy scale as indicated in [9] E. Aprile et al., Phys. Rev. C79, 045807 (2009). Fig. 1 as well as uncertainties in vesc are profiled out and [10] E. Aprile et al. (XENON100) Phys. Rev. D83, 082001 incorporated into the limit. The resulting 90% confidence (2011). [11] E. Aprile and T. Doke, Rev. Mod. Phys. 82, 2053 (2010). level (CL) limit is shown in Fig. 5 and has a minimum 45 2 2 [12] G. Plante et al. (2011), arXiv:1104.2587. σ =7.0 10− cm at a WIMP mass of m = 50 GeV/c . × χ [13] F. Arneodo et al., Nucl. Instrum. Meth. A449, 147 The impact of eff data below 3 keVnr is negligible at (2000); D. Akimov et al., Phys. Lett. B524, 245 (2002); 2L mχ = 10 GeV/c . The sensitivity is the expected limit in R. Bernabei et al., Eur. Phys. J. direct C3, 11 (2001). absence of a signal above background and is also shown E. Aprile et al., Phys. Rev. D72, 072006 (2005). V. Che- in Fig. 5 as 1σ and 2σ region. Due to the presence of pel et al., Astropart. Phys. 26, 58 (2006). A. Manzur et al., Phys. Rev. C81, 025808 (2010). two events around 30 keVnr, the limit at higher mχ is weaker than expected. This limit is consistent with the [14] E. Aprile et al., Phys. Rev. Lett. 97, 081302 (2006). [15] E. Aprile et al. (XENON100) (2011), arXiv:1103.0303. one from the standard analysis, which calculates the limit [16] S. Yellin, Phys. Rev. D66, 032005 (2002). based only on events in the WIMP search region with an [17] O. Buchmueller et al. (2011), arXiv:1102.4585. acceptance-corrected exposure, weighted with the spec- [18] C. E. Aalseth et al. (CoGeNT), Phys. Rev. Lett. 106, 2 trum of a mχ = 100 GeV/c WIMP, of 1471 kg days. 131301 (2011). This result excludes a large fraction of previously× unex- [19] C. Savage et al., JCAP 0904, 010 (2009). Sub-Weakly Interacting Massive Particles χ(!)

36 2 σ 10− cm n ∼ N

Energy χ Weak interactions Z boson ? Standard Model Dark Matter Inaccessibility

Wednesday, June 1, 2011 Sub-Weakly Interacting Massive Particles χ(!)

45 46 2 σ 10− − cm n ∼ N χ Higgs boson

M 1 GeV p ∼ ? Standard Model Dark Matter

Wednesday, June 1, 2011 Hunt for Dark Matter

• Indirect Detection • “Discovery” of weak scale dark matter

PAMELA

Wednesday, June 1, 2011 Indirect Detection Meade, Papucci, Strumia, Volansky dS!Γ dS!Γ DM DM # e$e!, Einasto profile DM DM # Μ$Μ!, Einasto profile DM DM # Τ$Τ!, Einasto profile dS!Γ 10!20 10!20 10!20

GC!Γ GC!Γ 10!22 GR!Γ 10!22 GR!Γ 10!22 GC!Γ sec sec IC sec Ν

! ! Ν ! 3 IC 3 3 IC

cm cm cm GR!Γ in in in v v v Σ 10!24 PAMELA and FERMI Σ 10!24 PAMELA and FERMI Σ 10!24 PAMELA and FERMI GC!radio GC!radio GC!radio

10!26 10!26 10!26 • Recent Launch 10of2 AMS103 104 102 103 104 102 103 104 DM mass in GeV DM mass in GeV DM mass in GeV

!Γ 3 dS!Γ dSGC!Γ • Improved constraints$ ! GC!Γ $ ! $ ! DMGCDM!radio# e e , isothermal profile DMGCDM!radio# Μ Μ , isothermal profile DMGCDM!radio# Τ Τ , isothermal profile hilation rate) to take on the value that best matches the data. In dSGC!!ΓΓ GR!Γ GR!Γ each case, we find10 excellent!20 fits to the observations, although 10!20 10!20 on propagationvery high annihilation rates model are required. We quantify this by GR!Γ introducing a boost factor (BF) which denotes the enhance- ment of the annihilation rate relative to a smooth halo with a Ν Ν 3 local dark matter density of 0.3 GeV/cm . We find that boost IC --> better 10!22 10!22 10!22

factors in the rangesec of approximately 400 to 2000 are required sec sec ! IC ! !

3 3 3 IC to produce the positron fraction observed by PAMELA. Al- though such valuescm are considerably larger than those gener- cm cm in in in

determinationally predicted basedv onof the results of N-body simulations [26], v v Σ ! Σ ! Σ ! such simulations10 do24 not have thePAMELA resolutionand toFERMI study the small 10 24 PAMELA and FERMI 10 24 PAMELA and FERMI scale structure of the galactic halo distribution and rely on ex- particle physicstrapolations to estimate annihilation boost factors. For this reason, we do not dismiss the possibility that small scale inho- mogeneities in the10 dark!26 matter distribution could significantly 10!26 10!26 2 3 4 2 3 4 2 3 4 boost the annihilation10 rate. 10 10 FIG. 3: The contribution10 to the cosmic10 ray antiproton-to-pr10 oton ra- 10 10 10 In Fig. 2, we show the cosmic ray electron plus positron Wednesday, June 1, 2011 DM mass in GeV tio from annihilating Kaluza-KleinDM mass darkin matterGeV as normalized to the DM mass in GeV spectrum, with contributions from the same dark matter model PAMELA and ATIC signals (and as in Figs. 1 and 2). The upper two and parameters shown in Fig. 1, compared to the measure- curves denote the results using propagation model A, and clearly ex- ments of ATIC [4]. Here, we have attempted to account ceed the antiproton content measured by PAMELA [28]. The lower for possible systematic errors (which are not quantified in two lines denote the results for propagation model B and predict considerably fewer antiprotons, safely evading this constraint. Results Ref. [4]) byFigure including 5% 6: systematicDirect errors DM (in addition annihilation to are shown for dark. matter We masses compare of 600 and 800 GeV.the region favored by PAMELA (green the statistical errors shown) in the calculation of the χ2, and by allowingbands) for an overall and shift by in the PAMELA, normalization/expos FERMIure and HESS observations (red ellipeses) with HESS observations of up to 25%. Although very limited information is available Fig. 3, we show the contribution to the cosmic ray antiproton- regarding theof systematic the Galatic errors involved Center in these observat [19ions,] (blueto-proton continuous ratio from KK dark line), matter, for Galactic the parametersRidge (anni- [20] (blue dot-dashed), and we feel that we have made reasonable estimates of these quan- hilation rates, masses, and propagation models) used in Figs. 1 tities. spherical dwarfes [21, 22] (blueand dashed), 2. The upper two FERMI lines in this figure observations correspond to the re- in the ‘10◦ 20◦’ region and of Based on the result shown in Figs. 1 and 2, we conclude sults found using propagation model A, and clearly exceed ÷ that KK darkobservations matter is capable of of providing the a Galactic good fit to the Centerthe ratio measured at radio-frequencies by PAMELA [28]. The lowerν two= lines, 408 GHz [44] (dashed red lines) combined observations of PAMELA14 and ATIC. In particu- in contrast, denote the results using propagation model B and lar, we findand acceptable at ν values10 for the χHz2 perby degree VLT of free- [45]are (upper consistent with purple PAMELA’s lines, antiproton when measurement. present, We for equipartition and constant dom for each mass and propagation∼ model we have consid- thus conclude that in order for KK dark matter to produce the ered (althoughmagnetic propagation model field). B provides See somewhat discussion bet- PAMELA in the and text ATIC signals for remarks without overproducing regarding antipro- the validity of the constraints. ter fits than model A). We would like to emphasize that the tons, a propagation+ model with a rather narrow diffusion+ zone + mass of theWe dark consideredmatter particle in this DM model annihilations is quite con- (L ∼ 1 intokpc) muste bee adopted.− (left We remindcolumn), the readerµ that suchµ− (middle), τ τ − (right), unity strained by relic abundance considerations, and could not eas- a propagation model is completely consistent with all current ily have beenboost very different and fromSommerfeld the values we have factors consid- andcosmic theray data. NFW (upper row), Einasto (middle), isothermal (lower) ered here.DM In particular, density to obtain profiles a thermal relic in abundanc the Milkye Other Way. constraints on dark matter17 annihilations in the Milky of KK dark matter, masses in the approximate range of 600 Way include those obtained by gamma ray, radio, and mi- to 900 GeV are generally required [15, 18]. Furthermore, KK crowave observations of the Galactic Center region. If an masses lighter than about 300-400 GeV are inconsistent with NFW-like halo profile is adopted and the annihilation rate electroweak precision measurements [27]. throughout the halo is enhanced by a common boost factor Before we can conclude that KK dark matter in this model (normalized to the PAMELA and ATIC signals), these con- is successful in producing the signals of PAMELA and ATIC, straints are likely to be exceeded [29, 30]. As baryonic effects we must consider constraints from measurements of cos- are anticipated to modify the dark matter distribution within mic ray antiprotons, gamma rays, and synchrotron emis- the inner kiloparsecs (or parsecs) of the galaxy, however, it is sion. In particular, dark matter annihilations which pro- difficult to arrive at any strong conclusions. In particular, the duce mostly quarks or gauge bosons are typically expected boost factors which enhances the dark matter annihilation rate to produce more cosmic ray antiprotons than are observed could potentially vary with location, and could be consider- by PAMELA [28] if the overall annihilation rate is normal- ably lower in the inner kiloparsecs of the Milky Way than in ized to generate the PAMELA and ATIC positron/electron ex- the vicinity of the Solar System. For example, if large boost cesses. This problem is somewhat mitigated in the case of KK factors exist, it would appear to imply the presence of a very dark matter, however, as most of the annihilations proceed large number of small sub-halos within the Milky Way. In the to charged leptons (which efficiently generate electrons and regions of the galaxy with the most stars (ie. the inner galaxy), positrons, while not contributing to the antiproton flux). In tidal interactions are the most likely to destroy a large fraction Watch (and Work) List

Top AFB B physics anomalies W + jets

CoGeNT/DAMA anomalies MiniBooNe Anomaly XENON and LUX below Cosmic Radiation Anomaly Higgs pole MINOS ν / ν¯ discrepancy AMS and cosmic positrons

Wednesday, June 1, 2011 !"#$%"&"'(!)*('%+ Energy9*:;+!*3!<=',"--!<*$*%*>?7 Frontier!$;

!" q˜ ,"-- g˜ A! B*#$%&'(-*7"&6*%D A*BE*F$G )*+& 45& A! B B • SearchH I:@$*J:4@;9'5(=*$-&",:&"'(*J>*-'#K"(;* for Higgs and supersymmetry #"($:9*$?5:&"'(-*&'*;$&*&6$*8:@$*%9'J:J"#"&>*8'9*:*q˜ H*#G"3'(! 13'9"34&5.,3#%&(.5.*3'#%(&7')&)&15##(.4&(.5.*3'#% g˜ 2'+;%<+!&*2I takesH D'%*J:4@;9'5(=*$-&",:&$=*7"&6*4'(&9'#*9$;"'( front stage "$J7! " K'<+!L!)MJ " N!O!)N

!"#$%&!' +34#5-"'(*%#'&*-6'7(*8'9 &*2!<="+!A!B! :*%6$(',$('#';"4:#*<00<*-4$(:9"' @3(!>%;"3*! 76"46*:##'7-*-#$%&'(- "(*&6$*=$4:>*46:"(- #@++7! <.0?5:9@1;#5"('2*A*<.46"B2*A*<.-#$%&'(2 #@++7! A*<.46"C*2 'C:%;("3>!@%%! !"#$%&'(&)&*#+,-.((./&01(#234&,)-3'*5.(6& #*('%+!>"D"3>! )%/&)&2)7#-)85.&(*.%)-'#&01")-/.-4,)-3'*5.(2 # E!/F-!$G",

18 • Likely,-./0.1/,, to push supersymmetric )*2"*3(!4567!1/,, states up just below a TeV

Wednesday, June 1, 2011 !"#$%"&"'(!)*('%+ Energy9*:;+!*3!<=',"--!<*$*%*>?7 Frontier!$;

18 • Likely,-./0.1/,, to push supersymmetric )*2"*3(!4567!1/,, states up just below a TeV

Wednesday, June 1, 2011 Top AFB

deltay_not_qdeltay_t_lephad [0][0][3][1][1][0]C

deltay_not_qdeltay_t_lephad [0][0][3][2][1][0]C Entries 9615115 Entries 9613620 Mean 0.00342 Mean -0.01103 RMS 0.7666 Semileptonic RMS 0.7731

Adata = -0.048 ± 0.039 Adata = 0.067 ± 0.040 data (positive) data (negative) A = 0.008 ± 0.004 A = -0.015 ± 0.004 tt+bkg tt + bkg tt+bkg Events 250 tt + bkg Events 250 A = 0.011± 0.003 A = -0.015 ± 0.003 bkg tt bkg tt A =0.110 0.039 A = -0.004 0.012 FB bkg ± Abkg = -0.014 ± 0.012 200 200 ±

150 150

100 100

50 50

0 0 8 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3

recon_hadTop_rapidity [0][0][3][2][1][0]C

Entries 9613620

Mean 0.01133 recon_hadTop_rapidity [0][0][3][1][1][0]C

Entries 9615115 RMS 0.556

Mean -0.01399 RMS 0.5552 y -y y -y Fully Leptonic l h l h A = 0.076 ± 0.039 A = -0.070 ± 0.040 data (negative) data data (positive) data !l+ - !l- CDF II Preliminary A = -0.024 ± 0.004 A = 0.013 ± 0.004 250 tt + bkg tt+bkg 250 tt + bkg tt+bkg Events Events -1 bkg A = -0.011 ± 0.003 bkg A = 0.006 ± 0.003 L dt = 5.1 fb tt tt & Data A = -0.073 0.012 A = 0.035 0.012 bkg ± bkg 60 ± tt CDF200 5.3 fb^-1 200 ± 1 # error Fake Events DY 150 150 40 Z% $$ WW/WZ/ZZ

100 100 20 50 50

0 0 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 01.5 2 2.5 y y -2 0 2 h h " ! AFB =0.14 0.05 l FIG.Wednesday, 3: Distributions June 1, 2011 of ∆y (top) and y (bottom) for events with± negativeFIG. 3: The leptons rapidity (left) between and positive the positively leptons and (right). negatively charged leptons in data and in the predicted signal and background lh h simulations. The forward backward asymmetry in data is Afb =0.14 ± 0.05.

V. MEASUREMENT OF THE INCLUSIVE tions for events with negative leptons (left) and pos- ytop-ytbar in Lab (e-e) CDF II Preliminary ASYMMETRIES itive leptons (right). An indication of asymmetry is -1 also observed in this figure: t quarks are dominant in L dt = 5.1 fb ¯ Data We now turn to the rapidity distributions in the data. the forward (proton) direction and the t quarks in the & backward (¯p) direction. The measured asymmetries are tt The inclusive distributions of the ∆ylh and yh variables + ± 1 # error A = 0.070 0.04015 and A− =0.076 0.039, again are shown in Fig. 2, compared to the standard pythia h h Fake equal and− opposite± Events within uncertainties. ± tt¯ + background prediction. These distributions contain DY the full sample of both lepton signs and should be sym- Z% $$ metric. The legend on the top right shows the asym- 10 WW/WZ/ZZ metries in all components. The data agrees well with The sign reversal of the asymmetry under interchange tt¯+bkg prediction in both variables, and, in particular, of the lepton charge (or, in our formalism, under inter- the asymmetries are consistent with zero. change of t and t¯) is consistent with CP conservation. A forward-backward asymmetry becomes apparent With larger samples and5 improved precision, the com- when the sample is separated by charge. The top row parison of the charge separated distributions will pro- of Fig. 3 shows the ∆y distributions for events with neg- vide a strict test of CP conservation in tt¯ production. If ative leptons (left) and positive leptons(right). We find we assume CP conservation we can calculate the total + A =0.067 0.040 and A− = 0.048 0.039, where asymmetry in each frame0 using Eqs. (1) and (2). The lh ± lh − ± -2 -1 0 1 2 the uncertainties are statistical only. With limited sig- distributions of these variables are shown in Fig. 4. The lab t¯t nificance, the asymmetries are equal in magnitude and asymmetry in the tt¯ frame is A =0.057 0.028, and in " yt opposite in sign. the laboratory frame is Ap¯p =0.073 0.028,± where both ± The bottom plots of Fig. 3 shows the yh distribu- uncertainties are statistical.ee FIG. 4: Aobs =0.270 ± 0.112(stat.) (Pred.: −0.010 ± 0.070). K-S probability is calculated to be 2.5 %. Top AFB

• Mass Dependence 17 we assign a systematic uncertainty of 0.035 for this effect. tt Additional systematic uncertainties are evaluated in A tt parton - level a manner similar to the inclusive case. These uncertain- CDF data 5.3 fb-1 ties are estimated by repeating the analysis while varying 0.4 tt NLO QCD the model assumptions within their known uncertainties for background normalization and shape, the amount of initial- and final-state radiation (ISR/FSR) in pythia, the calorimeter jet energy scale (JES), the model of fi- 0.2 nal state color connection, and parton distribution func- tions (PDF). Table XII shows the expected size of all systematic uncertainties. The physics model dependence dominates. 0.0

TABLE XIII: Asymmetry At¯t at high and low mass compared ! 0.2 to prediction. 2 2 selection Mtt¯ < 450 GeV/c Mtt¯ 450 GeV/c 2 ≥ 450 GeV/c M tt data 0.016 0.034 0.210 0.049 tt¯+bkg +0− .012 ± 0.006 0.030 ± 0.007 (mc@nlo) ± ± FIG. 14: Parton-level asymmetry in ∆y at high and low mass data signal 0.022 0.039 0.017 0.266 0.053 0.032 compared to mcfm prediction. The shaded region represents tt¯ +0− .015 ± 0.006± 0.043 ± 0.009 ± the total uncertainty in each bin. (mc@nlo) ± Wednesday, June 1, 2011± data parton 0.116 0.146 0.047 0.475 0.101 0.049 mcfm +0− .040 ± 0.006± 0.088 ± 0.013 ± ± ±

*"" +,-./0 Table XIII compares the low and high mass asymme- )"" try to predictions for the data level, the background sub- ("" tracted signal-level, and the fully corrected parton-level. '"" The MC predictions include the 15% theoretical uncertainty. At low mass, within uncertainties, the asymmetry &"" at all correction levels agrees with predictions consistent %"" with zero. At high mass, combining statistical and sys- $"" tematic uncertainties in quadrature, the asymmetries at all levels exceed the predictions by more than three stan- !"" dard deviations. The parton-level comparison is summa- #"" 2 rized in Fig. 14. For Mtt¯ 450 GeV/c , the parton-level ¯ " asymmetry at in the tt¯ rest≥ frame is Att =0.475 0.114 " #" !" $" %" &" '" (" )" *" #"" ! (stat+sys), compared with the MCFM prediction± of χ At¯t =0.088 0.013. ± FIG. 15: Distribution of tt¯ reconstruction χ2. Black crosses are data, histogram is sig+bkg prediction.The last bin on the right contains all events with χ2 > 100.

VIII. CROSS-CHECKS OF THE MASS DEPENDENT ASYMMETRY A. Lepton Type The large and unexpected asymmetry at high mass de- mands a broader study of related eﬀects in the tt¯ data. All of our simulated models predict asymmetries that We look for anomalies that could be evidence of a false are independent of the lepton type: pythia predicts positive, along with correlations that could reveal more asymmetries that are consistent with zero, and the Octet about a true positive. In order to avoid any assumptions models predict asymmetries that are consistent with each related to the background subtraction, we make compar- other. The data are shown in Table XIV. At high mass, isons at the data level, appealing when necessary to the both lepton types show positive asymmetries consistent full tt¯ + bkg simulation models. within errors. Models to Generate

• s-channel or t-channel

q t g t q t • s-channel: axigluon Ferrario and Rodrigo M M M ¯ • t-channel: ﬂavor q t¯ q¯ t¯ g t¯

violating vector or (a) s-channel qq¯ (b) s-channel gg (c) t-channel qq¯

scalar Jung, Murayama, Pierce, WellsFigure 1: Tree level tt¯ production diagram with mediator M exchange. Shu, Tait, Wang Ligeti, Schmaltz,fully Tavares leptonic tt¯ events, which has recently been discussed in a CDF note [49]. Lastly, we Grinstein, Kagan, Trott, Zupan discuss the LHC reach for discovering such states, based on the analysis of [48].

II. MODELS

The Leading Order (LO) SM tree-level amplitude for tt¯ production does not generate a Wednesday, June 1, 2011 forward-backward asymmetry. In the SM, a small positive top forward-backward asymmetry is generated through interference between a one-loop box diagram and a LO tree level 3 diagram, A (M ¯ < 450GeV) = 0.040 0.006, A (M ¯ > 450 GeV) = 0.088 0.013. FB tt ± FB tt ± Since the SM contribution is generated at NLO, if there is an additional LO tree-level contribution from new physics, it can easily dominate. Such LO diagrams are of the form of those in Fig. (1). They can be either s-channel (Fig. (1a) and (1b)) or t-channel (Fig. 1c). s-channel mediators couple directly to light ﬂavors and gluons, and therefore the mediator masses must be large enough to evade dijet

resonance search constraints [11, 17]. To maximize the contribution to AFB, such a model must have a big axial coupling. On the other hand, t-channel models should have large flavor violation between the light and the top generations, as can be seen in Fig. (1c). Large flavor violation is experimentally allowed even for low mass mediators, M, as long as new couplings between light generations and left-handed quarks is suppressed; then strong limits on flavor violation and from dijet

3 Interference between initial state gluon radiation and ﬁnal state gluon radiation makes a very small negative contribution to the asymmetry.

5 To Watch

q t˜ g t˜ g ˜ g M • Top-ﬂavor t

q violating M t˜ M ¯ q q resonancesq t˜ M t˜ q M

• (M -> tj)(a) tt¯ production (b) t-channel (c) s-channel (d) u-channel Gresham, Kim, KZ 350 (M=φa only) 7 300 6 "

• SingleFigure 1:Top Tree level tt¯and single M production diagrams involving the mediator M and thedata coupling 250

of 5 GeV 1 $

50 a #

˜ ˜ ˜ ¯ fb gM . The top quark, t, is t = t when M200= W !,ZH! , and t = t when M = φ (triplet or sextet).4 1 fb ! j t ! 150

• (M -> jj) M 3 given d # Top forward-backward asymmetry generating models of type (ii) discussed abovebin have Σ 100 # 2 2 • d Χ interactions of the form gMtq¯ 50where M is the mediator, q is a light quark, and1 g is

order 1. Thus the production of0 M through qg Mt as in Fig. 1 is expected0 to be 200 300 →400 500 600 700 200 300 400 500 600 700 ! Mt j GeV M GeV substantial. t j 400 3.5 ! " Wednesday, June 1, 2011 ! " For mediators with mass mM >mt, this implies M can decay through M tq˜ 3.0, where

• " 300 → data 2.5 0.5 t˜ t t¯ tj˜ t˜tj˜ j of = or . Therefore, a resonance# should exists in events, where is a jet formed 1 $ fb !

fb 2.0 j t 1 !

from the light quark q. Θ 200 1.5 cos given d # bin

Σ 1.0 To avoid constraints from same sign top pair production, we assume that M# is not

d 100 2

• Χ self-conjugate, and then the signature is a top-jet (tj) or anti-top-jet (tj¯ ) resonance0.5 in 0 0.0 tt¯ plus jet events. $1.0 $0.5 0.0 0.5 1.0 $1.0 $0.5 0.0 0.5 1.0 cos Θ! ! t j cos Θt j Due to baryon number conservation, the ﬁnal state light quark baryon number must • 200 GeV Resonance match that of the initial state quark. In a pp machine (as opposed to pp¯), which has quark collisions dominantly over anti-quark collisions, the resonance will be dominantly Figure 3: Diﬀerential cross section in fb versus invariant mass (left top) or cos θtj¯ (left bottom)

either tj or tj¯ , depending on the baryon number2 of the mediator, BM = 2/3 or 1 and corresponding χ per bin (right) for the background± only hypothesis given 1 fb− of data BM = 0, respectively. (right) and a 200 GeV W " resonance with coupling gR = 1. The normalized background is shown

t Therefore, in contrast to otherin LHC solid search blue, studies and the for “measured” models related (Standard to the A ModelFB anomaly, + signal) diﬀerential cross-section is shown which have focused on the tt¯ oras dijet a red invariant dashed-dotted mass distributions line. The overall [30, background 33, 34],2 here normalization we was ﬁxed by matching to the “measured” cross section between invariant masses from 300 to 700 GeV. 2 For generic colored resonance search through QCD interations, see [35].

production cross-section of σ = 40 pb for mM = 200 GeV, changing to σ = 4 pb for mM = 3 600 GeV, can be constrained at the 3σ level. For example, this corresponds to a reach in ¯ coupling of W " to dRtR at the level of gR = 1 for mW ! = 200 GeV, weakening to gR =1.75

for mW ! = 600 GeV, assuming a 100% branching ratio to top-jet. For a resonance coupling to u quarks in the initial state and a production cross-section of σ = 27 pb for m = 200 GeV, changing to σ 3 4 pb for m = 600 GeV, can be M ≈ − M 1 constrained at the 3σ level with 1 fb− . For example, this corresponds to a reach in coupling

for a color triplet at the level of gφ =0.8 for mφ = 200 GeV, weakening to gφ =1.3 for

12 Early LHC

• Large couplings to top --> LHC7 search with 1 fb^-1!

ZH ' ZH ' 0.6 300 GeV, g!1.4 300 GeV, g!1.4 0.8 400 GeV, g!1.6 400 GeV, g!1.6 ! 400 GeV, g 1.8 400 GeV, g!1.8 0.4 600 GeV, g!2.4 0.6 600 GeV, g!2.4 800 GeV, g!3.2 800 GeV, g!3.2 800 GeV, g!3.4 800 GeV, g!3.4 t t

t 0.2 t 0.4 FB FB A A

0.0 0.2

"0.2 0.0

0 450 0 1

mt t GeV "y W' W' Gresham,0.6 Kim, KZ 200 GeV, g!1.4 ! " 200 GeV, g!1.4 0.8 300 GeV, g!1.8 300 GeV, g!1.8 400 GeV, g!2.4 400 GeV, g!2.4 0.4 400 GeV, g!2.6 0.6 400 GeV, g!2.6 600 GeV, g!3.4 600 GeV, g!3.4 600 GeV, g!3.6 Wednesday, June 1, 2011 600 GeV, g!3.6 t t

t 0.2 t 0.4 FB FB A A

0.0 0.2

"0.2 0.0

0 450 0 1

mt t GeV "y

! "

Figure 5: AFB for ZH! ,W! models with couplings as indicated in the legend (with g = gR,gL = 0), with the hatched regions corresponding to 1σ errors based on the limited statistics of the sample.

The contribution to AFB includes both t-channel ZH! ,W! exchange and single ZH! ,W! production. The red bars are the CDF observation with 1σ errors, while the blue bars indicate the NLO SM contribution from [1], which has not been included in the LO contribution calculated via MadGraph

and PYTHIA. The last bin includes all events with m ¯ > 450 GeV and ∆y > 1, respectively. tt | |

A similar analysis is carried out for triplets, sextets and axigluons in Fig. (6). While triplet and sextet models can marginally reproduce the asymmetries at LO, the rise between the low and high invariant mass bins and low and high rapidity bins is not as pronounced for

the triplets and sextets as for the W ! and ZH! . The reason for this in the sextet and triplet cases can, for example, be easily extracted from the analytical expressions, Eqs. (5), (6), (10). At high invariant mass, the scattering amplitude is dominated by the squared term. There uˆ uˆ in Eq. (10), and the eﬀect of the on the asymmetry vanishes. By contrast, the t ! φ Asq 2 2 2 2 Z! ,W! are dominated byu ˆ /tˆ (1 + c ) /(1 c ) which retains a contribution to the H t M ∼ θ − θ asymmetry at high invariant mass. The axigluon models tend to signiﬁcantly underproduce the asymmetry in the high in-

14 B-physics Anomalies

Tevatron like-sign muons

b 3 + + a = (8.5 2.8) 10− b¯b µ µ X sl − ± × →

B s mixing in ∆Γ s and S ψφ

Less significant:

measurement of sin 2β in B ψ K d → and penguin dominated b sqq¯ → • Maybe there is something ﬂavorful brewing?

Wednesday, June 1, 2011 Watch (and Work) List

Top AFB B physics anomalies W + jets

CoGeNT/DAMA anomalies MiniBooNe Anomaly XENON and LUX below Cosmic Radiation Anomaly Higgs pole MINOS ν / ν¯ discrepancy AMS and cosmic positrons

Wednesday, June 1, 2011 ! ! IntensityMiniBooNE Frontier e and e Data

! Mode • Neutrino physics anomalies • MiniBooNE MiniBooNE !e and !e Data

5.66E20 POT

! Mode ! Mode

5.66E20 POT

! ModeWednesday, June 1, 2011 3

2 2 2 MB# ∆m41 |Ue4||Uµ4| ∆m51 |Ue5||Uµ5| δ/πχ/dof KARMEN 1 3+2 0.47 0.128 0.165 0.87 0.138 0.148 1.64 110.1/130 10 NOMAD LSND/MiniBooNe Antineutrino mode MB results Full Energy Range 1+3+1 0.47 0.129 0.154 0.87 0.142 0.163 0.35 106.1/130

2 ]

2 Table II: Parameter values and χ at the global best ﬁt •! Results for 5.66E20 POT points for 3+2 and 1+3+1 oscillations (∆m2’s in eV2).

•! Maximum likelihood [eV fit in0 2 41 10 LSND + MB#$ Kopp, Maltoni, Schwetz simple 2 neutrino modelm 1 0.3

" 90, 99% CL LSND •! Null excluded at 99.5% with disappearance 0.4 respect to the two neutrino 0.8 MiniBooNE

[%] 0.3 0.2 oscillation fit (anti-neutrinos) 0.6 0.2 LSND

-1 P 10 E>475 MeV 0.1 99% CL (2 dof) 0.4 0.1 MiniBooNE -4 -3 -2 -1 10 10 10 10 excess events 0.2 (neutrinos) 0 2 sin 2!SBL 0 0.3 0.6 0.9 1.2 1.5 3 0.3 0.6 0.9 1.2 1.5 3 Figure 3: Global constraints on sterile neutrinos in the 3+1 ECCQE [GeV] ECCQE [GeV] model. We show the allowed regions at 90% and 99% CL from # # a combined analysis of the LSND [3] and MiniBooNE anti- neutrino [4] signals (filled regions), as well as the constraints Figure 4: Predicted spectra for MiniBooNE data and the from the null results of KARMEN [20], NOMAD [21] and transition probability for LSND (inset). Solid histograms re- fer to the 3+2 global best fit point (Tab. II), dashed his- MiniBooNE neutrinoMiniBooNE [19] appearance searches (blue contour). 20 The limit from disappearance experiments (green contours) tograms correspond to the best fit of appearance data only includes data from CDHS [22], atmospheric neutrinos, and (LSND, MiniBooNE ν/ν¯, KARMEN, NOMAD). For Mini- from the SBLHard reactor experiments.to fit disappearance For the latter we compare BooNE experiments we fit only data above with 475 MeV. 3 + 2 light the results for the new anti-neutrino flux prediction from [5] (solid) and the previous ones [6] (dashed). The region to thesteriles right of the curves is excluded at 99% CL. Wednesday, June 1, 2011 for new (old) reactor fluxes. Hence we conclude that the 3+1 scenario does not provide a satisfactory description of the data despite the new hint coming from reactors. atmospheric neutrinos. Technical details of our analysis Let us move now to the 3+2 model, where SBL exper- can be found in [8, 10] and references therein. iments depend on the seven parameters listed in Tab. II. In the 3+1 scheme the SBL experiments depend on In addition to the two mass-squared differences and the 2 the three parameters ∆m41, |Ue4|, and |Uµ4|. Since moduli of the mixing matrix elements, also a physical ∗ ∗ only one mass-scale is relevant in this case it is not complex phase enters, δ ≡ arg(Uµ4Ue4Uµ5Ue5). This possible to obtain CP violation. Therefore, oscillations phase leads to CP violation in SBL oscillations [8, 24], involving one sterile neutrino are not capable of rec- allowing to reconcile differing neutrino and anti-neutrino onciling the different results for neutrino (MiniBooNE) results from MiniBooNE/LSND. Tab. II shows the para- and anti-neutrino (LSND and MiniBooNE) appearance meter values at the global best fit point and the corre- 2 searches. Fig. 3 compares the allowed regions from LSND sponding χ value. Changing from the previous to the 2 and MiniBooNE anti-neutrino data to the constraints new reactor flux calculations the χ decreases by 10.6 from the other experiments in the 3+1 model. Note units, indicating a significant improvement of the descrip- that, even though reactor analyses using the new flux tion of the data, see also upper panel of Fig. 2. From that prediction prefer non-zero Ue4, no closed regions ap- figure follows also that going from 3+1 to 3+2 leads to pear for the disappearance bound (solid curve), since a significant improvement of the fit with the new reactor 2 2 2 sin 2θSBL =4|Ue4| |Uµ4| can still become zero if fluxes, which was not the case with the old ones. The 2 Uµ4 = 0. We find that the parameter region favored by χ improves by 11.2 units, which means that 3+1 is dis- LSND and MiniBooNE anti-neutrino data is ruled out by favoured at the 97.6% CL (4 dof) with respect to 3+2, 2 other experiments, except for a tiny overlap of the three compared to ∆χ =6.3 (82% CL) for old fluxes. 2 2 99% CL contours around ∆m41 ≈ 1 eV . Note that in In Fig. 1 we show the prediction for the Bugey spectra this region the constraint from disappearance data does at the global best fit point as dashed curves. Clearly they not change significantly due to the new reactor flux pre- are very similar to the best fit of reactor data only. Fig. 4 dictions. Using the PG test from [23] we find a compat- shows the predicted spectra for MiniBooNE neutrino and ibility of the LSND+MiniBooNE(¯ν) signal with the rest anti-neutrino data, as well as the LSNDν ¯µ → ν¯e transi- −5 2 of the data only of about 10 , with χPG = 21.5(24.2) tion probability. Again we find an acceptable fit to the CosmicMotivation…. Anomalies Cosmology Fits for the Number of Sterile Neutrinos (J. Hamann, et. al. arXiv:1006.5276)Hamann et al.

3 + N • Current data: s mv = 0

• To watch: Planck 3 + N experiment s Ns m+ s3 = 0 ms = 0

Bashinsky, Friedland, KZ Wednesday, June 1, 2011 FIG. 16: Summary of the bounds on the numbers of free-streaming and coupled neutrinos from the current data (left) and 4 the simulated Planck data (right). For the case of the current data, the Large Red Galaxy (LRG) dataset from SDSS yields a region that differs significantly from what is obtained with the Main SDSS dataset. The combined fit is also shown. The sensitivity of Planck to the numbers of both freely streaming and coupled neutrinos will be a dramatic improvement over that of all present-day experiments combined (compare “Planck only” and “All current” on the right). If Planck’s data is combined with today’s data, further improvement on the constraints are expected.

APPENDIX: ON RESONANT PRODUCTION logarithm similar to that in the above equation (“the Coulomb logarithm”) is regulated by plasma screening In this section we perform a more rigorous derivation effects. 2 of the results of Sect. VI. The rate for the process needs to be compared to T × 2 Consider a “test” neutrino traveling through a neu- T /Mpl (“momentum exchange of order T by the time trino gas. We need to know how far it travels before ap- when the temperature equals T ”). We get that neutrinos preciably changing the direction of its momentum, i.e., are streaming freely at recombination when given the initial momentum of order T , when the parti- 1/4 1/4 g ! (Trec/Mpl) × (16π/ ln(Trec/mφ)) . (A.3) cle’s momentum in the transverse direction becomes order T . The rate for this process needs to be compared to The 1/4 power makes the coefficient (the second term) of 2 the expansion rate of the universe, given by ∼ T /Mpl in order one for a wide range of mφ, hence for the purpose of the era of radiation domination. this estimate we can drop it. The order of magnitude is Let us consider the scattering of two neutrinos, νν → set by the ratio (Trec/Mpl). For the temperature O(eV ) νν, and start with the t-channel exchange of φ. In this one thus finds the neutrinos are not coupled by the t- case, for light φ (mφ # T ) the cross-section is strongly channel scalar exchange if forward peaked (the well-known property of Ruther- −7 ford scattering). The situation is well-known in plasma g ! 10 . (A.4) physics. The build-up of transverse momentum happens The 1/4 power also means that when the t channel pro- mostly as a result of many small-angle scattering events, cess is dominant it is reasonable to treat neutrinos as rather than a single large-angle event. The multiple scat- tightly coupled on all scales, not just the sound horizon tering events result in a random walk process for the at the CMB decoupling. Indeed, even considering scales transverse momentum pT , so that the rate of change of 100 times smaller than the sound horizon only changes 2 pT is given by the bound on g by about a factor of 3.

2 If mφ >T, the situation is different. The momentum dpT dσ 2 ∼ d(cos θ) nvrelp . (A.1) exchange occurs via large-angle scattering and the cross dt d cos θ θ 2 4 ! section is set by T /mφ, so that 2 2 2 3 1/4 Taking vrel ∼ 1, pθ ∼ T θ , n ∼ T and dσ/d cos θ ∼ g " (Trec/Mpl) (mφ/Trec). (A.5) 4 2 2 2 2 −2 g /(16π)T (T θ + mφ) , we get The interaction decouples as the temperature drops, a dp2 g4 characteristic feature of nonrenormalizable interactions. T ∼ T 3 ln(T/m ). (A.2) dt 16π φ Indeed, at temperature T # mφ the φ field can be inte- grated out, resulting in an effective 4-fermion vertex. The small-angle behavior of the cross section is regulated We now consider νν → νν scattering via the s-channel by the mass of the scalar field. In plasma physics, the process. In this channel, an important possibility is the Summary

• Next 2-3 years promise to be action- packed! • Look to all frontiers for progress • With current path, likely to re-shape theory • Need “outside-the-box” theoretical ideas

Wednesday, June 1, 2011