Hunt for new phenomena using large jet multiplicities and missing transverse momentum with ATLAS in 4.7 fb^-1 of sqrt(s) = 7 TeV proton-proton collisions

Results are presented of a search for new particles decaying to large numbers of jets in association with missing transverse momentum, using 4.7 fb^-1 of pp collision data at sqrt(s) = 7 TeV collected by the ATLAS experiment at the Large Hadron Collider in 2011. The event selection requires missing transverse momentum, no isolated electrons or muons, and from>=6 to>=9 jets. No evidence is found for physics beyond the Standard Model. The results are interpreted in the context of a MSUGRA/CMSSM supersymmetric model, where, for large universal scalar mass m_0, gluino masses smaller than 840 GeV are excluded at the 95% confidence level, extending previously published limits. Within a simplified model containing only a gluino octet and a neutralino, gluino masses smaller than 870 GeV are similarly excluded for neutralino masses below 100 GeV.


Introduction
Many extensions of the Standard Model of particle physics predict the presence of TeVscale strongly interacting particles that decay to lighter, weakly interacting descendants. Any such weakly interacting particles that are massive and stable can contribute to the dark matter content of the universe. The strongly interacting parents would be produced in the proton-proton interactions at the Large Hadron Collider (LHC), and such events would be characterized by significant missing transverse momentum E miss T from the unobserved weakly interacting daughters, and jets from emissions of quarks and/or gluons.
In the context of R-parity conserving [1][2][3][4][5] supersymmetry [5][6][7][8][9][10], the strongly interacting parent particles are the squarksq and gluinosg, they are produced in pairs, and the lightest supersymmetric particles can provide the stable dark matter candidates [11,12]. Jets are produced from a variety of sources: from quark emission in supersymmetric cascade decays, production of heavy Standard Model particles (W , Z or t) which then decay hadronically, or from QCD radiation. Examples of particular phenomenological interest include models where squarks are significantly heavier than gluinos. In such models the gluino pair production and decay process g +g → t +t +χ 0 1 + t +t +χ 0 1 can dominate, producing large jet multiplicities when the resulting top quarks decay hadronically. In the context of MSUGRA/CMSSM models, a variety of different cascade decays, including thegg initiated process above, can lead to large jet multiplicities.
A previous ATLAS search in high jet multiplicity final states [13] examined data taken during the first half of 2011, corresponding to an integrated luminosity of 1.34 fb −1 . This paper extends the analysis to the complete ATLAS 2011 pp data set, corresponding to 4.7 fb −1 , and includes improvements in the analysis and event selection that further increase sensitivity to models of interest.
Events are selected with large jet multiplicities ranging from ≥ 6 to ≥ 9 jets, in association with significant E miss T . Events containing high transverse momentum (p T ) electrons or muons are vetoed in order to reduce backgrounds from (semi-leptonically) decaying top quarks or W bosons. Other complementary searches have been performed by the ATLAS collaboration in final states with E miss T and one or more leptons [14,15]. Further searches have been carried out by ATLAS using events with at least two, three or four jets [16], or with at least two b-tagged jets [17]. Searches have also been performed by the CMS collaboration, including a recent analysis in fully hadronic final states [18].

The ATLAS detector and data samples
The ATLAS experiment [19] is a multi-purpose particle physics detector with a forwardbackward symmetric cylindrical geometry and nearly 4π coverage in solid angle. 1 The layout of the detector is dominated by four superconducting magnet systems, which comprise a thin solenoid surrounding inner tracking detectors, and a barrel and two end-cap toroids supporting a large muon spectrometer. The calorimeters are of particular importance to this analysis. In the pseudorapidity region |η| < 3.2, high-granularity liquid-argon (LAr) electromagnetic (EM) sampling calorimeters are used. An iron/scintillator-tile calorimeter provides hadronic coverage for |η| < 1.7. The end-cap and forward regions, spanning 1.5 < |η| < 4.9, are instrumented with LAr calorimetry for both EM and hadronic measurements.
The data sample used in this analysis was taken during April -October 2011 with the LHC operating at a proton-proton centre-of-mass energy of √ s = 7 TeV. Application of beam, detector and data-quality requirements resulted in a corresponding integrated luminosity of 4.7±0.2 fb −1 [20]. The analysis makes use of dedicated multi-jet triggers that required either at least four jets with p T > 45 GeV or at least five jets with p T > 30 GeV, 1 ATLAS uses a right-handed coordinate system with its origin at the nominal interaction point in the centre of the detector and the z-axis along the beam pipe. Cylindrical coordinates (r, φ) are used in the transverse plane, φ being the azimuthal angle around the beam pipe. The pseudorapidity η is defined in terms of the polar angle θ by η = − ln tan(θ/2).
where the energy is measured at the electromagnetic scale 2 and the jets must have |η| < 3.2. In all cases the trigger efficiency was greater than 98% for events satisfying the offline jet multiplicity selections described in Section 4.

Object reconstruction
The jet, lepton and missing transverse momentum definitions are based closely on those of Ref.
[13], with small updates to account for evolving accelerator and detector conditions. Jet candidates are reconstructed using the anti-k t jet clustering algorithm [21, 22] with radius parameter of 0.4. The inputs to this algorithm are clusters of calorimeter cells seeded by cells with energy significantly above the noise level. Jet momenta are reconstructed by performing a four-vector sum over these topological clusters of calorimeter cells, treating each as an (E, p) four-vector with zero mass. The jet energies are corrected for the effects of calorimeter non-compensation and inhomogeneities by using p T -and η-dependent calibration factors based on Monte Carlo (MC) simulations validated with extensive test-beam and collision-data studies [23]. Only jet candidates with p T > 20 GeV and |η| < 4.9 are retained. Further corrections are applied to any jet falling in problematic areas of the calorimeter. The event is rejected if, for any jet, this additional correction leads to a contribution to E miss T that is greater than both 10 GeV and 0.1 E miss T . These criteria, along with selections against non-collision background and calorimeter noise, lead to a loss of signal efficiency of ∼8% for the models considered. When identification of jets containing heavy flavour quarks is required, either to make measurements in control regions or for cross checks, a tagging algorithm exploiting both impact parameter and secondary vertex information is used. Jets are tagged for |η| < 2.5 and the parameters of the algorithm are chosen such that 70% of b-jets and ∼1% of light flavour or gluon jets, are selected in tt events in Monte Carlo simulation [24]. Jets initiated by charm quarks are tagged with about 20% efficiency.
Electron candidates are required to have p T > 20 GeV and |η| < 2.47, and to satisfy the 'medium' electron shower shape and track selection criteria of Ref. [14]. Muon candidates are required to have p T > 10 GeV and |η| < 2.4. Additional requirements are applied to muons when defining leptonic control regions. In this case muons must have longitudinal and transverse impact parameters within 1 mm and 0.2 mm of the primary vertex, respectively, and the sum of the transverse momenta of other tracks within a cone of ∆R = 0.2 around the muon must be less than 1.8 GeV, where ∆R = (∆η) 2 + (∆φ) 2 .
The measurement of the missing transverse momentum two-vector p miss T and its magnitude (conventionally denoted E miss T ) is then based on the transverse momenta of all electron and muon candidates, all jets with |η| < 4.5 which are not also electron candidates, and all calorimeter clusters with |η| < 4.5 not associated to such objects [25].
Signal region 7j55 8j55 9j55 6j80 7j80 8j80 Number of isolated leptons (e, µ) = 0 Following the steps above, overlaps between candidate jets with |η| < 2.8 and leptons are resolved as follows. First, any such jet candidate lying within a distance ∆R = 0.2 of an electron is discarded, then any lepton candidate remaining within a distance ∆R = 0.4 of such a jet candidate is discarded. Thereafter, all jet candidates with |η| > 2.8 are discarded, and the remaining electron, muon and jet candidates are retained as reconstructed objects.

Event selection
Following the object reconstruction described in Section 3, events are discarded if they contain any jet failing quality criteria designed to suppress detector noise and non-collision backgrounds, or if they lack a reconstructed primary vertex with five or more associated tracks.
For events containing no isolated electrons or muons, six non-exclusive signal regions (SRs) are defined as shown in Table 1. The first three require at least seven, eight or nine jets, respectively, with p T > 55 GeV; the latter three require at least six, seven or eight jets, respectively, with p T > 80 GeV. The final selection variable is E miss T / √ H T , the ratio of the magnitude of the missing transverse momentum to the square root of the scalar sum H T of the transverse momenta of all jets with p T > 40 GeV and |η| < 2.8. This ratio is closely related to the significance of the missing transverse momentum relative to the resolution due to stochastic variations in the measured jet energies [25]. The value of E miss T / √ H T is required to be larger than 4 GeV 1/2 for all signal regions. A previous ATLAS analysis of similar final states [13] required jets to be separated by ∆R > 0.6 to ensure that the trigger efficiency was on its plateau. It has since been demonstrated that the requirement of an offline jet multiplicity at least one larger than that used in the trigger is sufficient to achieve a 98% trigger efficiency. Investigations on the enlarged data sample, in comparison to the previous incarnation of the strategy used here, allow various improvements to be made; in particular, the requirement on jet-jet separation is modified so as to increase the acceptance for signal models of interest by a factor two to five, without introducing any significant trigger inefficiency.
The dominant backgrounds are multi-jet production, including purely strong interaction processes and fully hadronic decays of tt; semi-and fully-leptonic decays of tt; and leptonically decaying W or Z bosons produced in association with jets. Non-fully-hadronic tt, and W and Z are collectively referred to as 'leptonic' backgrounds. Contributions from gauge boson pair and single top quark production are negligible. The determination of the multi-jet and 'leptonic' backgrounds is described in Sections 6 and 7, respectively.   All Monte Carlo samples employ a detector simulation [43] based on GEANT4 [44] and are reconstructed with the same algorithms as the data.

Multi-jet backgrounds
The dominant background at intermediate values of E miss T is multi-jet production including purely strong interaction processes and fully hadronic decays of tt. These processes are not reliably predicted with existing Monte Carlo calculations, and so their contributions must be determined from collision data. Indeed, the selection cuts have been designed such that multi-jet processes can be determined reliably from supporting measurements.
The method for determining the multi-jet background from data is motivated by the following considerations. In events dominated by jet activity, including hadronic decays of top quarks and gauge bosons, the E miss T resolution is approximately proportional to √ H T , and is almost independent of the jet multiplicity. The distribution of the ratio E miss T / √ H T has a shape that is almost invariant under changes in the jet multiplicity, as shown in Figure 1. The multi-jet backgrounds therefore can be determined using control regions with lower E miss T / √ H T and/or lower jet multiplicity than the signal regions. 5 The control 5 Residual variations in the shape of the E miss T / √ HT are later used to quantify the systematic uncertainty regions are assumed to be dominated by Standard Model processes, an assumption that is corroborated by the agreement of multi-jet cross section measurements with up to six jets [45] with Standard Model predictions.
As an example, the estimation of the background expected in the 8j55 signal region is obtained as follows. A template describing the shape of the E miss T / √ H T distribution is obtained from those events that contain exactly six jets, using the same 55 GeV p T threshold as the target signal region. That six-jet E miss T / √ H T template is normalized to the number of eight-jet events observed in the region E miss T / √ H T < 1.5 GeV 1/2 after subtraction of the 'leptonic' background expectation. The normalized template then provides a prediction for the multi-jet background for the 8j55 signal region for which E miss T / √ H T > 4 GeV 1/2 . A similar procedure is used for each of the signal regions, and can be summarized as follows. For each jet p T threshold p < ∈ {55 GeV, 80 GeV}, control regions are defined for different numbers n jet of jets found above p < . The number of events N p<,n jet (s min , s max ) for which E miss T / √ H T (in units of GeV 1/2 ) lies between s min and s max is determined, and the predicted 'leptonic' contributions L p<,n jet (s min , s max ) subtracted Transfer factors The multi-jet prediction for the signal region is found from the product of the T p<,n jet , with the same p < as the signal region and n jet = 6 when p < = 55 GeV (n jet = 5 when p < = 80 GeV) times the number of events (after subtracting the expected contribution from 'leptonic' background sources) satisfying signal region jet multiplicity requirements but with E miss T / √ H T < 1.5 GeV 1/2 .

Systematic uncertainties on multi-jet backgrounds
The method is validated by determining the accuracy of predictions for regions with jet multiplicities and/or E miss T / √ H T smaller than those chosen for the signal regions. Figure 1 shows that the shape of the E miss T / √ H T distribution for p < = 55 GeV and n jet = 6 is predicted to an accuracy of better than 20% from that measured using a template with the same value of p < and n jet = 5. Similarly, the distribution for p < = 80 GeV and n jet = 5 can be predicted for all E miss T / √ H T using a template with n jet = 4. The templates are normalized for E miss T / √ H T < 1.5 GeV 1/2 , and continue to provide a good prediction of the distribution out to values of E miss T / √ H T of 4 GeV 1/2 and beyond. Additional validation regions are defined for each p < and for jet multiplicity requirements equal to those of the signal regions, but for the intermediate values of (s min , s max ) of (1.5, 2), (2, 2.5) and (2.5, 3.5). Residual inaccuracies in the predictions are used to quantify the systematic uncertainty from the closure of the method. Those uncertainties are in the range 15%-25%, depending on p < and E miss T / associated with the method, as described in Section 6.1.
The mean number of proton-proton interactions per bunch crossing µ increased during the 2011 run, reaching µ = 16 at the start of proton fills for runs late in the year. Sensitivity to those additional interactions is studied by considering the jet multiplicity as a function of µ , and of the number of reconstructed primary vertices. The consistency of the high-p T tracks within the selected jets with a common primary vertex is also investigated. The effect of additional jets from pile-up interactions is found to be significant for low-p T jets but small for jets with p T > 45 GeV, and negligible for the jet selection used for the SRs.
The presence of multiple in-time and out-of-time pp interactions also leads to a small but significant deterioration of the E miss T resolution. The effectiveness of the E miss T / √ H T template method described above is tested separately for subsets of the data with different values of the instantaneous luminosity, and hence of µ . Good agreement is found separately for each subset of the data. Since the data set used to form the template has the same pile-up conditions as that used to form the signal regions, the changing shape of the E miss T resolution is included in the data-driven determination and does not lead to any additional systematic uncertainty.
Due to the presence of neutrinos produced in the decay of hadrons containing bottom or charm quarks, events with heavy-flavour jets exhibit a different E miss T distribution. To quantify the systematic uncertainty associated with this difference, separate templates are defined for events with at least one b-tagged jet and for those with none. The sum of the predictions for events with and without b-tagged jets is compared to the flavour-blind approach, and the difference is used to characterize the systematic uncertainty from heavy flavour (10%-20%). Other systematic uncertainties account for imperfect knowledge of: the subtracted 'leptonic' contributions (10%), the potential trigger inefficiency (2%), and imperfect response of the calorimeter in problematic areas (1%).
The backgrounds from multi-jet processes are cross checked using another data-driven technique [16] which smears the energies of individual jets from low-E miss T multi-jet 'seed' events in data. Separate smearing functions are defined for b-tagged and non-b-tagged jets, with each modelling both the Gaussian core and the non-Gaussian tail of the jet response, including the loss of energy from unobserved neutrinos. The jet smearing functions are derived from GEANT4 [44] simulations [43]. The Gaussian core of the function is tuned to di-jet data, and the non-Gaussian tails are verified with data in three-jet control regions in which the p miss T can be associated with the fluctuation of a particular jet. There is agreement within uncertainties between the background predicted by this jet-smearing method and the primary method based on the shape invariance of E miss T / √ H T .

'Leptonic' backgrounds
Non-fully-hadronic (i.e. semi-leptonic or di-leptonic) tt, and W and Z production are collectively referred to as 'leptonic' backgrounds. The process Z → νν + jets contributes to the signal regions since it produces jets in association with E miss T . Leptonic tt and W decays contribute to the signal regions when hadronic τ decays allow them to evade the lepton tt + jets Muon kinematics p T > 20 GeV, |η| < 2.4 Muon multiplicity = 1 = 2 Jet p T , |η|, multiplicity (CR) As in Table 1. veto, with smaller contributions from events in which electrons or muons are produced but are not reconstructed. The 'leptonic' background predictions employ the Monte Carlo simulations described in Section 5. To reduce uncertainties from Monte Carlo modelling and detector response, it is desirable to normalize the background predictions to data using control regions (CR) and cross-check them against data in other validation regions (VR). These control regions and validation regions are designed to be distinct from, but kinematically close to, the signal regions. Each is designed to provide enhanced sensitivity to a particular background process.
The control and validation regions are defined as shown in Table 2. By using control regions that are kinematically similar to the signal regions, theoretical uncertainties, including those arising from the use of a leading-order (LO) generator, are reduced. The tt + jets and W + jets validation regions each require a single muon and no electrons. For the tt process the single-muon selection is primarily sensitive to the semi-leptonic decay. 6 The tt + jets validation region is further enhanced by the requirement of at least one b-tagged jet, whereas for W + jets enhancement a b-tag veto is applied. Since it is dominantly through hadronic τ decays that W and tt contribute to the signal regions, the corresponding control regions are created by recasting the muon as a (τ -)jet. For Z → νν + jets the validation regions select events from the closely related process Z → µµ + jets. The related control regions are formed from these validation regions by recasting the muons as neutrinos.
In detail, for those control regions where the Monte Carlo simulations predict at least one event for 4.7 fb −1 , the leptonic background prediction s i for each signal region from each background is calculated by multiplying the number of data events c data i found in the corresponding control region by a Monte Carlo-based factor t MC This transfer factor is defined to be the ratio of the number of MC events found in the signal region to the number of MC events found in the control region In each case, the event counts are corrected for the expected contamination by the other background processes. Whenever less than one event is predicted in the control region, the Monte Carlo prediction for the corresponding signal region is used directly, without invoking a transfer factor. For the tt + jets background, the validation region requires exactly one isolated muon, at least one b-tagged jet, and no selected electrons. The transverse mass for the muon transverse momentum p µ T and the missing transverse momentum two-vector p miss T is calculated using massless two-vectors and must satisfy 50 GeV < m T < 100 GeV. Figure 2 shows the jet multiplicity in the tt validation regions, and it is demonstrated that the Monte Carlo provides a good description of the data. The tt control regions used to calculate the background expectation differ from the validation regions as follows. Since the dominant source of background is from hadronic τ decays in the control regions, the muon is used to mimic a jet, as follows. If the muon has sufficient p T to pass the jet selection threshold p < , the jet multiplicity is incremented by one. If the muon p T is larger than 40 GeV it is added to H T . The selection variable E miss T / √ H T is then recalculated, and required to be larger than the threshold value of 4 GeV 1/2 . Distributions of the jet multiplicity in the tt control regions may also be found in Figure 2.
The W + jets validation regions and control regions are defined in a similar manner to those for tt + jets, except that a b-jet veto is used rather than a b-jet requirement (see Table 2). Figure 3 shows that the resulting jet multiplicity distributions are well described by the Monte Carlo simulations.
The Z + jets validation regions are defined (as shown in Table 2) requiring precisely two muons with invariant mass m µµ consistent with m Z . The dominant backgrounds from Z + jets arise from decays to neutrinos, so in forming the Z + jets control regions from the validation regions, the vector sum of the p T of the muons is added to the measured p miss T , to model the E miss T expected from Z → νν events. The selection variable E miss T / √ H T is then recalculated and required to be greater than 4 GeV 1/2 for events in the control region. Figure 4 shows that the resulting jet multiplicity distributions in both validation and control regions are well described by the Monte Carlo simulations.
For each of the 'leptonic' backgrounds further comparisons are made between Monte Carlo and data using the lower jet p T threshold of 45 GeV, showing agreement within uncertainties for all multiplicities (up to nine jets for tt, see Figures 2(a) and 2 (b). The Alpgen Monte Carlo predictions for Z + jets and W + jets were determined with six additional partons in the matrix element calculation, and cross checked with a calculation in which only five additional partons were produced in the matrix element -in each case with additional jets being produced in the parton shower. The two predictions are consistent with each other and with the data, providing further supporting evidence that the parton shower offers a sufficiently accurate description of the additional jets.

Systematic uncertainties on 'leptonic' backgrounds
The use of control regions is effective in reducing uncertainties from Monte Carlo modelling and detector response. When predictions are taken directly from the Monte Carlo, the 'leptonic' background determinations are subject to systematic uncertainties from Monte Carlo modelling of: the jet energy scale (JES, 40%), the jet energy resolution (JER, 4%), the number of multiple proton-proton interactions (3%), the b-tagging efficiency (5% for tt), the muon trigger and reconstruction efficiency and the muon momentum scale. The numbers in parentheses indicate the typical values of the SR event yield uncertainties prior to the partial cancellations that result from the use of control regions.
The JES and JER uncertainties are calculated using a combination of data-driven and Monte Carlo techniques [23], using the complete 2011 ATLAS data set. The calculation accounts for the variation in the uncertainty with jet p T and η, and that due to nearby jets. The Monte Carlo simulations model the multiple proton-proton interactions with a varying value of µ which is well matched to that in the data. The residual uncertainty from pile-up interactions is determined by reweighting the Monte Carlo samples so that µ is increased or decreased by 10%. The uncertainty in the integrated luminosity is 3.9% [20].
When transfer factors are used to connect control regions to signal regions, the effects of these uncertainties largely cancel in the ratio. For example, the impact of the jet energy scale uncertainty is reduced to ≈ 6%. Figure 5 shows the E miss T / √ H T distributions after applying the jet selections for the six different signal regions (see Table 1) prior to the final E miss T / √ H T > 4 GeV 1/2 requirement. Figure 6 shows the jet multiplicity distributions for the two different jet p T thresholds after the final E miss T / √ H T requirement. It should be noted that the signal regions are not exclusive. For example, in Figure 5, all plots contain the same event at E miss T / √ H T ∼ 11 GeV 1/2 . The 'leptonic' backgrounds shown in the figures are those calculated from the Monte Carlo simulation, using the MC calculation of the cross section and normalized to 4.7 fb −1 . The number of events observed in each of the six signal regions, as well as their Standard Model background expectations are shown in Table 3. Good agreement is observed between SM expectations and the data for all six signal regions.  Table 3. sources other than the Standard Model, the corresponding upper limit σ 95% BSM,max × A × on the cross section times efficiency within acceptance (which equals the limit on the observed number of signal events divided by the luminosity) and the p-value for the Standard-Modelonly hypothesis (p SM ).

Results, interpretation and limits
In the absence of significant discrepancies, limits are set in the context of two supersymmetric (SUSY) models. The first is the tan β = 10, A 0 = 0 and µ > 0 slice of the MSUGRA/CMSSM parameter space. The second is a simplified SUSY model with only a gluino octet and a neutralinoχ 0 1 within kinematic reach. Theoretical uncertainties on the SUSY signals are estimated as described in Section 5. Combined experimental systematic uncertainties on the signal yield from jet energy scale, resolution, and event cleaning are approximately 25%. Acceptance times efficiency values are tabulated elsewhere [49].
The limit for each signal region is obtained by comparing the observed event count with that expected from Standard Model background plus SUSY signal processes, taking into account all uncertainties on the Standard Model expectation, including those which are correlated between signal and background (for instance jet energy scale uncertainties) and all but theoretical cross section uncertainties (PDF and scale) on the signal expectation. The combined exclusion regions are obtained using the CL s prescription [50], taking the signal region with the best expected limit at each point in parameter space. The 95% confidence level (CL) exclusion in the tan β = 10, A 0 = 0 and µ > 0 slice of MSUGRA/CMSSM is shown in Figure 7. The ±1 σ band surrounding the expected limit shows the variation anticipated from statistical fluctuations and systematic uncertainties on SM and signal processes. The uncertainties on the supersymmetric signal cross section from PDFs and higher-order terms are calculated as described in Section 5, and the resulting signal cross section uncertainty is represented by ±1σ lines on either side of the observed limit. 7 The analysis substantially extends the previous exclusion limits [13,16,17] for m 0 > 500 GeV. For large m 0 , the analysis becomes independent of the squark mass, and the lower bound on the gluino mass is extended to almost 840 GeV for large mq. 8 In the simplified model gluinos are pair-produced and decay with unit probability to t +t +χ 0 1 . In this context, the 95% CL exclusion bound on the gluino mass is 870 GeV for neutralino masses up to 100 GeV.

Summary
A search for new physics is presented using final states containing large jet multiplicities in association with missing transverse momentum. The search uses the full 2011 pp LHC data set taken at √ s = 7 TeV, collected with the ATLAS detector, which corresponds to an integrated luminosity of 4.7 fb −1 .
Six non-exclusive signal regions are defined. The first three require at least seven, eight or nine jets, with p T > 55 GeV; the latter three require at least six, seven or eight jets, with p T > 80 GeV. In all cases the events are required to satisfy E miss T / √ H T > 4 GeV 1/2 , and to contain no isolated high-p T electrons or muons. Investigations on the enlarged data sample have resulted in improvements compared to a previous measurement using a similar strategy. In particular, inclusion of events with smaller jet-jet separation increases the acceptance for signal models of interest by a factor two to five, without without significantly increasing the systematic uncertainty.
The Standard Model multi-jet background is determined using a template-based method that exploits the invariance of E miss T / √ H T under changes in jet multiplicity, cross-checked with a jet-smearing method that uses well reconstructed multi-jet seed events from data. The other significant backgrounds -tt + jets, W + jets and Z + jets -are determined using a combination of data-driven and Monte Carlo-based methods.
In each of the six signal regions, agreement is found between the Standard Model prediction and the data. In the absence of significant discrepancies, the results are interpreted as limits in the context of R-parity conserving supersymmetry. Exclusion limits are shown for MSUGRA/CMSSM, for which, for large m 0 , gluino masses smaller than 840 GeV are excluded at the 95% confidence level. For a simplified supersymmetric model in which both of the pair-produced gluinos decay via the processg → t +t +χ  Refs. [13, 16, 17] the effect of the signal cross section uncertainty was folded into the displayed limits and so was not shown separately. 8 Limits on sparticle masses quoted in the text are those from the lower edge of the 1 σ signal cross section band rather than the central value of the observed limit, so can be considered conservative.                    Figure 6. The distribution of jet multiplicity for jets with p T > 55 GeV (a) and those with p T > 80 GeV (b). Only events with E miss T / √ H T > 4 GeV 1/2 are shown.  Figure 7. Combined 95% CL exclusion curves for the tan β = 10, A 0 = 0 and µ > 0 slice of MSUGRA/CMSSM (a) and for the simplified gluino-neutralino model (b). The dashed grey and solid red lines show the 95% CL expected and observed limits respectively, including all uncertainties except the theoretical signal cross section uncertainty (PDF and scale). The shaded yellow band around the expected limit shows its ±1σ range. The ±1σ lines around the observed limit represent the result produced when moving the signal cross section by ±1σ (as defined by the PDF and scale uncertainties). The contours on the MSUGRA/CMSSM model show values of the mass of the gluino and the mean mass of the squarks in the first two generations. Exclusion limits are also shown from previous ATLAS searches with ≥2, 3 or 4 jets plus E miss T [16], multi-jets plus E miss T [13] or with same-sign dileptons [46] and from LEP [47] in (a). The lower plot shows limits from ATLAS searches with same-sign dileptons [46] or with one-lepton plus b-jet [48].
We thank CERN for the very successful operation of the LHC, as well as the support staff from our institutions without whom ATLAS could not be operated efficiently.
[24] ATLAS Collaboration, Commissioning of the ATLAS high-performance b-tagging algorithms in the 7 TeV collision data. ATLAS-CONF-2011-102. [48] ATLAS Collaboration, Search for supersymmetry in pp collisions at √ s = 7 TeV in final states with missing transverse momentum and b-jets with the ATLAS detector. ATLAS-CONF-2012-003.

A Event displays
A display of an event that passes the 9j55 and 7j80 signal region selections can be found in Figure 8. A display of an event that passes all signal region selections can be found in Figure 9.