5c5 < the CMS detector as a function of $p_T$ which primarily result from calorimeter --- > CMS detector as a function of $p_T$ which primarily result from calorimeter 18c18 < coverage, the largest $p_T$ range, and at a given $p_T$, the highest statistics. --- > coverage, the largest $p_T$ range, and at a given $p_T$, the highest highest statistics. 23,28c23,26 < will be derived for a flavor mixture given by either dijet, < photon-jet or Z-jet events. In a photon-jet sample with $30 will be derived for a flavor mixture given by either dijet or > photon-jet events. In photon-jet sample with $30 only $\approx 90\%$ of the jets balancing the photons are either a $q$ or a $\bar q$ jet where as at low > $p_T$, dijets are mainly gluon jets. The relative contribution of quark and gluon jets 37,38c35,36 < Two different procedures have been used to correct the calorimeter jets < to particle jets at the Tevatron. The CDF technique\cite{bib:CDFJES} relies on the --- > Two different procedure have been used to correct the calorimeter jets > to particle jets at Tevatron. The CDF technique\cite{bib:CDFJES} relies on the 42,43c40,41 < the same technique using photon-jet events. In the former case, an accurately simulated detector response is < the primary source of calibration, whereas in the latter case, simulation is used only to determine the residual corrections --- > the same technique using photon-jet events. In former case, an accurately simulated detector response is > the primary source of calibration where as in latter case, simulation is used only to determine the residual corrections 46,49c44,46 < Although data-driven photon-jet or Z-jet balancing techniques may require less effort, an accurate simulation of the detector is < essential for physics analysis, especially the ones relying on a precise simulation of (rare) signal < events which must be separated from a large background. Thus all effort should be made to < ensure that the detector response --- > Although data-driven photon-jet balancing technique may require less effort, an accurate simulation of the detector is > essential for physics analysis, especially the ones relying on precise simulation of (rare) signal > events which must be separated from a large background. Thus all effort should be made to ensure that detector response 51c48 < data accurately represent the real data. --- > data accurate represent that real data. 53,59c50,55 < At CMS, the use of these techniques will evolve over time. At the start of data taking, we will use Monte Carlo < based corrections derived from simulated dijet data. As the photon-jet and Z-jet data are accumulated and understood, these data-driven < based corrections will replace the Monte Carlo corrections if the overall uncertainty on the < jet scale is smaller. As mentioned above the corrections < derived from photon/Z-jet events are dominated by quark-initiated jets. These corrections will be converted to dijet mixture of quark and gluon jets based < on the Monte Carlo information. At some later stage, after the detector simulation has been tuned to $pp$ collision data, the simulated data will be used to < improve upon or replace the data based corrections, especially at high $p_T$ values where the photon event statistics is not --- > At CMS, the use of these techniques will evolve over time. At the start of data taking, we will used Monte Carlo > based corrections derived from simulated dijet data. As the photon-jet data is accumulated and understood, the photon-jet > based corrections will replace the Monte Carlo corrections if overall uncertainty on jet scale is smaller. As mentioned above the corrections > derived from photon-jet are dominated by quark-initiated jets. These corrections will be converted to dijet mixture of quark and gluon jets based > on the Monte Carlo information. At some later stage, after detector simulation has been tuned to $pp$ collision data, the simulated data will be used to > improve upon or replace the data based corrections, especially at high $p_T$ values where photon event statistics is not 63,66c59,62 < In this scheme, the jet energy scale is obtained by comparing the < hadronic event with the photons measured in the electromagnetic calorimeter or < a $Z$ boson reconstructed from either electrons or muons. The decay into two muons has additionally the advantage, that the reconstruction of the Z~boson relies only on the tracking and muon system and not on the calorimeters. < The photon-jet events have larger statistics but suffer from a large --- > In this scheme, jet energy scale obtained by comparing the > hadronic event with the photons measured in the EM calorimeter or > a $Z$ boson reconstructed from either electrons or muons. The > photon-jet events have larger statistics but suffer from a large 68,82c64,66 < identification requirements. The contamination depends on the details of the photon identification requirements. < The $Z$-jet sample is hardly contaminated but has smaller statistics. < \begin{figure}[hbtp] < \begin{center} < \resizebox{8cm}{!}{ < \includegraphics{ZjetStatisticsAfterCuts.eps} < } < \caption{ < A comparison of the expected number of events after cuts for the photon-jet and Z-jet samples after selection cuts for an integrated luminosity of $1fb^{-1}$. The photon-jet sample is presented with the expected QCD background contribution after cuts. The signal to background ratio is about one for the region of low transverse momentum and increases with $p_T$. For the Z-jet sample, only the Z-boson decaying into two central muons is considered which is expected to be hardly contaminated by any background.} < < \label{Fig:ZGammaStatAfterCuts} < \end{center} < \end{figure} < < The QCD background to the photon-jet sample is about two orders of magnitudes larger than signal. Therefore, several selection criteria have to be applied in order to prepare a clean photon-jet sample. After a current set of selection cuts, the signal to background ratio in this sample is about one for the region of small transverse momenta of the photon. For large transverse momenta, the background decreases compared to the signal. In contrast to the photon-jet sample, the expected number of Z-jet events is smaller, especially for small transverse momenta of the Z~boson. The expected number of events presented in Fig.~\ref{Fig:ZGammaStatAfterCuts} is for the decay of the Z~Boson into two central muons as in this selection, the reconstruction of the Z boson depends only on the tracking and muon system and does not use calorimeter information. --- > identification requirements. The contamination depends on the details of photon identification requirements. > The $Z$-jet sample is less contaminated but has smaller statistics. Actual difference in the statistics will > depend on the pre-scale used during data taking. Below we discuss the photon-jet samples. 84d67 < In addition, the actual difference in statistics will depend on the pre-scale used during data taking. Below we discuss the photon-jet samples and two of the commonly used photon-jet balancing techniques as well as the transverse momentum method for the Z-jet samples with the Z boson decaying into muons. 111c94 < --- > Two of the commonly used photon-jet balancing techniques are described below. 125,137c108 < \begin{figure}[hbtp] < \begin{center} < \resizebox{15cm}{!}{{\includegraphics{ZmumuResponseAll.eps}{\includegraphics{ZmumuResponseQuark.eps}}{\includegraphics{ZmumuResponseGluon.eps}}}} < \caption{ < Ratio of calorimeter and particle jet $p_T$ (iterative cone algorithm with a cone size of $R=0.5$) to the transverse momentum of the Z~boson, reconstructed either from reconstructed or generator muons. The plot on the left hand side shows the ratio for the generated mixture of jets coming from quarks and gluons for a transverse momentum of the Z~boson between $80 In the absence of any radiation, the $p_T$$^{\gamma}$= $p_T$$^{parton}$ 139,144c110,113 < $p_T^{\gamma/z}$ is accurately known. In practice, because of initial state < radiation, the mean of the $p_T^{parton}/p_T^{\gamma/z}$ distribution may not be 1.0. < < The contamination from < the dijet events in the photon-jet data, where one jet < mimics a photon may modify the distributions in Fig.~\ref{Fig:PhotonResponse}. --- > $p_T^{\gamma}$ is accurately known. In practice, because of initial state > radiation, the mean of the $p_T^{parton}/p_T^{\gamma}$ distribution may not be 1.0. The contamination from > the dijet events, where one jet > mimics a photon may modify this distribution. ======================================================================= 147,148c116,117 < all events (black) are shown in Fig.~\ref{Fig:PhotonResponse}(left). From this figure, we see that for the events at the peak, < the parton energy is equal to the photon energy. Thus the jet energy scale at parton level can be determined by positioning --- > all events (black) are shown in Fig.~\ref{Fig:PhotonResponse}(left). From this figure, we see that for the events at peak, > parton energy is equal to the photon energy. Thus the jet energy scale at parton level can be determined by positioning 154,156c123 < showering/hadronization models such as {\sc pythia}. < < The ratio of particle (calorimeter) jet energy to the photon --- > showering/hadronization models such as {\sc pythia}. The ratio of particle (calorimeter) jet energy to the photon 158c125 < is shown in solid (dashed) histograms in Fig.~\ref{Fig:PhotonResponse}(middle). The mean of the ratio --- > is shown in solid (dashed) histograms in Fig.~\ref{Fig:PhotonResponse}(middle). The mean of ratio 163,166c130 < gluon initiated particle jet is also $\sim 10\%$ lower than the calorimeter response to a quark initiated particle jet. < < A similar behavior is observed for the Z-jet sample with a transverse momentum between $80 gluon initiated particle jet is also $\sim 10\%$ lower than calorimeter response to a quark initiated particle jet. 168,169c132 < profile of the jet and hence the leakage outside the cone at $\sqrt{s}=1.96$ TeV reasonably < well~\cite{bib:CDF-JetShapes} but this will be have to be verified at $\sqrt{s}=14$ TeV. --- > profile of the jet and hence the leakage outside the cone at $\sqrt{s}=1.96$ TeV reasonably well~\cite{bib:CDF-JetShapes} but will be have to be verified at $\sqrt{s}=14$ TeV. 172,174c135,136 < In an event, the transverse momentum of the photon ($\vec{p}_{T\,\gamma}$) should balance the < transverse momentum of the recoiling system ($\vec{p}_{T\,recoil}$). < The recoil system includes --- > An event, the transverse momentum of the photon ($\vec{p}_{T\,\gamma}$) should balance the > transverse momentum of recoiling system ($\vec{p}_{T\,recoil}$). The recoil system includes 190c152 < Assuming the response of the calorimeter is given by --- > Assuming the response of calorimeter is given by 202,203c164 < The missing transverse energy in an event can be written\footnote{It is assumed that all < particles in the particle jet balancing the photon go into the --- > The missing transverse energy in an event can be written\footnote{It is assumed all particles in the particle jet balancing the photon go into 210,212c171,172 < Assuming that $|\vec{E}_T^{miss \rm soft}|=0$ and $\vec{P}_T^{\rm particle\ jet}= < -\vec{P}_T^{\rm photon\ jet}$, the $MPF$ method is equivalent( identical) < to the $p_T$ balance method. --- > Assuming that $|\vec{E}_T^{miss \rm soft}|=0$ and $\vec{P}_T^{\rm particle\ jet}= -\vec{P}_T^{\rm photon\ jet}$, $MPF$ method is equivalent( identical) > to $p_T$ balance method. 215,216c175,176 < requiring one and only jet in the event and and requiring that this jet is in the opposite direction to the photon. We will use these cuts and, below, < we replace $R_{recoil}$ by $R_{Jet,MPF}$. Using these assumptions, Eq.~\ref{eqn:mpf} can be written as --- > requiring one and only jet in the event and and requiring that this jet is in opposite direction to the photon. We will use these cuts and, below, > we replace $R_{recoil}$ by $R_{Jet,MPF}$. Using these assumption, Equation~\ref{eqn:mpf} can be written as 225,226c185 < jet. However, the correction factors for R=0.5 and R=0.7 are expected to be different as the < particles in an annulus between --- > jet. However, the correction factors for R=0.5 and R=0.7 are expected to be different as the particles in annulus between 229c188 < the measured $R_{Jet,MPF}$ has to be corrected for the finite size of the jet. This correction will be determined from --- > the measured $R_{Jet,MPF}$ has to corrected for the finite size of the jet. This correction will be determined from 236,241c195,199 < are shown in Fig.~\ref{Fig:DeltaPtMPF} for MidPoint jets with R=0.5 in the $|\eta|<1.3$ region. < The $p_T$ balance response is lower than $p_T^{CaloJet}/p_T^{ParticleJet}$ because the denominator in < $p_T$-balance is equivalent to the parent parton which in general has more energy than particle < jet. From these plots one can estimate that $k_{MPF}$ is $\sim 3\%$ for a 100 GeV jet. Please note that this $k_{MPF}$ implicitly < corrects for the calorimeter out-of-cluster showering {\it i.e.} the energy deposited outside the < calorimeter cluster by the particles included in the particle jet. --- > are shown in Fig.~\ref{Fig:DeltaPtMPF} for MidPoint jets with R=0.5 in $|\eta|<1.3$ region. > The $p_T$ balance response is lower than $p_T^{CaloJet}/p_T^{ParticleJet}$ as because the denominator in > $p_T$-balance is equivalent to the parent parton which in general has more energy than particle jet. From these plots one can estimate that $k_{MPF}$ is $\sim 3\%$ for 100 GeV jet. Please note that this $k_{MPF}$ implicitly > corrects for the calorimeter out-of-cluster showering {\it i.e.} the energy deposited out side the > calorimeter cluster by the particles included in particle jet. 247c205 < \subsubsection{Comparison of the methods} --- > \subsubsection{Comparison of two methods} 250,251c208 < % \resizebox{8cm}{!}{ < \resizebox{15cm}{!}{ --- > \resizebox{8cm}{!}{ 253,254d209 < \includegraphics{GraphIC5CaloJetResponse.eps} < 257,258c212,214 < \caption{ < Three measures of jet response are shown as a function of transverse momentum in a photon+jet Monte Carlo sample: the MC truth based response CaloJet/GenJet, a $p_T$ balance response CaloJet/Photon, and the MPF response. In addition, response from a Z-jet Monte Carlo sample with the Z~boson decaying into muons is shown as a function of transverse momentum of the Z~boson.} --- > \caption{Three measures of jet response are shown as a function of transverse momentum in a photon+jet > Monte Carlo sample. The MC truth based response CaloJet/GenJet, a $p_T$ balance response > CaloJet/Photon, and the MPF response.} 265,268c221,223 < The ${\it MPF}$ technique depends on the accurate measurement of the missing transverse energy which may not be < available at the start of the run. In any case, we plan to use both methods and compare the results. A comparison of the corrections < using Spring07 samples is shown in Fig.~\ref{Fig:DeltaPtMPF} as a function of $p_T^{\gamma}$. For < simplicity we have used a pure photon-jet sample. For the photon, we use the $p_T^{\gamma}$ of the generated photon. In addition, the response determined using the Z-jet samples is shown in this plot. For this signal, the CSA07 datasets with the Z decaying into muons have been used. --- > The ${\it MPF}$ technique depends on the accurate measurement of the missing transverse energy which may be not be > available at that start of the run. In any case, we plan to use both methods and compare the results. A comparison of the corrections > using Spring07 samples is shown in Fig.~\ref{Fig:DeltaPtMPF} as a function of $p_T^{\gamma}$. For simplicity (and avoid serious work), we have used a pure photon-jet sample. For photon, we use the $p_T^{\gamma}$ of the generated photon. 276c231 < Current JES corrections are derived from simulated data. A particle jet is matched to the nearest calorimeter jet in --- > Current JES corrections are derived from simulated data. A particle jet is matched to nearest calorimeter jet in 282c237 < calorimeter jet Lorentz vector. The calorimeter response and the correction factors for CMSSW 1\_3\_1 as a function of calorimeter jet $p_T$ for Iterative cone of R=0.5 and R-0.7 are shown in Fig.~\ref{fig:MCJetResponse}. --- > calorimeter jet 4 vector. The calorimeter response and the correction factors for CMSSW 1\_3\_1 as a function of calorimeter jet $p_T$ for Iterative cone of R=0.5 and R-0.7 are shown in Fig.~\ref{fig:MCJetResponse}. 289c244 < \caption{(left) Simulated calorimeter response to jets versus particle jet $p_T$ for iterative cone jets with $R=0.5,0.7$, (right) --- > \caption{(left) Simulated calorimeter response to jets verses particle jet $p_T$ for iterative cone jets with $R=0.5,0.7$, (right) 296,299c251,253 < As discussed above, the calorimeter response to gluon jets is lower than the response to quark < jets, since for the same < $P_{T}^{GenJet}$ the energy spectrum of particles in a gluon jet is softer. The photon-jet events are dominated < by $qg\rightarrow q\gamma$ events whereas dijet events, at low $p_T$, mainly arise from $gg\rightarrow gg$ scattering. Also for the Z-jet events, the balanced jets are mostly comming from a $q$ or a $\bar q$. --- > As discussed above the calorimeter response to gluon jets is lower than the response to quark jet as for same > $P_{T}^{GenJet}$, the energy spectrum of particles in a gluon jet is softer. The photon-jet events are dominated > by $qg\rightarrow q\gamma$ events where as dijet events, at low $p_T$, mainly arise from $gg\rightarrow gg$ scattering. 302c256,258 < using $p_T$ balancing method after the response has been corrected for the parton energy not included in the particle jet. It can be compared to the response directly measured in a photon-jet sample (green curve). The purple curve shows the response measured in the dijet sample. --- > using $p_T$ balancing method after the response has been corrected for the parton energy not included in particle jet. It > can be compared to response directly measured in photon-jet sample (green curve). The purple curve shows the response measured in > dijet sample. 304c260 < In order to provide the corrections for the same flavor mixture as the MCJet corrections, the corrections derived from the photon-jet --- > In order to provide the corrections for same flavor mixture as the MCJet corrections, the corrections derived from the photon-jet 307c263,266 < The current calorimeter response simulated in {\sc geant4} program describes the CMS test beam data reasonably well but shows a small discrepancy --- > > > > The current calorimeter response simulated in {\sc geant4} program and describes the CMS test beam data reasonably well but shows a small discrepancy 309,311c268,269 < from what is observed in test beam data due to the presence of the magnetic field and the extra < material in front of the calorimeter. These differences < can be understood and modeled so that an accurate ($<3\%$) jet energy scale can be determined. --- > from what is observed in test beam data due to the presence of the magnetic field and the extra material in front of calorimeter. These difference > can be understood and modeled so that accurate ($<3\%$) jet energy scale can be determined. 318,320c276,280 < \caption{ < Comparison of the calorimeter response as measured from a data driven technique using a photon-jet with the response measured from a Monte Carlo based technique in a dijet sample. The blue curve shows the response measured using the $p_T$ balancing method after the response has been corrected for the parton energy not included in the particle jet. It can be compared to response directly measured in the photon-jet sample. The purple curve shows the response measured in the dijet sample. The difference between the purple and green curves is a reflection of the different flavor composition of the two samples. < } --- > \caption{Comparison of the calorimeter response as measured from data driven technique from photon-jet sample with > the response measured from Monte Carlo based technique from dijet sample. The blue curve shows the response measured > using $p_T$ balancing method after the response has been corrected for the parton energy not included in particle jet. It > can be compared to response directly measured in photon-jet sample. The purple curve shows the response measured in > dijet sample. The difference between purple and green curves is reflection of different flavor composition of the two samples.} 330,331c290 < The procedure is applied on top of the zero-suppression or/and offset corrections discussed in < section~\ref{sec:offset}. --- > The procedure is applied on top of the zero-suppression or/and offset corrections~\ref{sec:offset}. 336c295 < dependence of $\mu$ on $E_{track}$ is fitted with an ad-hoc function $F(E^{track})$. --- > dependence $\mu$ on $E_{track}$ is fitted with an ad-hoc function $F(E^{track})$. 340c299 < the value $E^{track} - \mu$ is added to $E_{jet}$. The momenta of the tracks that reach the calorimeter surface out --- > value $E^{track} - \mu$ is added to $E_{jet}$. The momenta of the tracks that reach the calorimeter surface out 346,348c305,306 < where $\mu_{i} = F(E^{track}_{i})$ for tbe $i$th track in tbe cone at the calorimeter surface and < $\mu_{i} = 0$ for tracks out < of tbe cone at the calorimeter surface. --- > where $\mu_{i} = F(E^{track}_{i})$ for i-track in cone at the calorimeter surface and $\mu_{i} = 0$ for track out > of cone at the calorimeter surface. 351c309 < occupancy or coarse granularity. The example of zero-suppression and jet plus track corrections, done in sequence, --- > occupancy or coarse granularity. The example of zero-suppression and jet plus track corrections, done in the sequence, 379d336 <