Sparticles in motion: Analyzing compressed SUSY scenarios with a new method of event reconstruction

The observation of light superpartners from a supersymmetric extension to the Standard Model is an intensely sought-after experimental outcome, providing an explanation for the stabilization of the electroweak scale and indicating the existence of new particles which could be consistent with dark matter phenomenology. For compressed scenarios, where sparticle spectra mass splittings are small and decay products carry low momenta, dedicated techniques are required in all searches for supersymmetry. In this paper we suggest an approach for these analyses based on the concept of recursive jigsaw reconstruction, decomposing each event into a basis of complementary observables, for cases where strong initial state radiation has sufficient transverse momentum to elicit the recoil of any final state sparticles. We introduce a collection of kinematic observables which can be used to probe compressed scenarios, in particular exploiting the correlation between missing momentum and that of radiative jets. As an example, we study squark and gluino production, focusing on mass-splittings between parent superparticles and their lightest decay products between 25 and 200 GeV, in hadronic final states where there is an ambiguity in the provenance of reconstructed jets.

Figure 1

A simplified decay tree diagram for analyzing compressed signal topologies in events with an ISR system.

Figure 2

The distribution of $R_{\mathrm{ISR}}$ for the production and decay of ${\tilde{\chi }}_2^0{\tilde{\chi }}_2^0\to Z({\ell }^{+}{\ell }^{-}){\tilde{\chi }}_1^{0}h(\gamma \gamma ){\tilde{\chi }}_1^0$ at the 13 TeV LHC. (Top) $R_{\mathrm{ISR}}$ for $m_{{\tilde{\chi }}_2^0}=500\mathrm{GeV}$ and varying $m_{{\tilde{\chi }}_1^0}$. (Bottom) $R_{\mathrm{ISR}}$ as a function of $p_{\mathbf{ISR},T}^{\mathbf{CM}}$ for $m_{{\tilde{\chi }}_2^0}=500\mathrm{GeV}$, $m_{{\tilde{\chi }}_1^0}=450\mathrm{GeV}$. Simulated events are generated and analyzed using the RestFrames software package [9].

Figure 3

Distribution of $p_{\mathbf{ISR},T}^{\mathbf{CM}}$ for events that have passed the preselection requirements in Table 1. A high $p_T$ jet system, identified with ISR, is required to build the decay tree.

Figure 4

Distribution of the $p_{\mathbf{ISR},T}^{\mathbf{CM}}$ as a function of $R_{\mathrm{ISR}}$ for (from left to right) $\mathrm{boson}+\text{jets}$ and $\mathrm{top}+\mathrm{X}$ backgrounds, gluino and squark pair-production signal samples.

Figure 5

The number of jets with minimum $p_T>20\mathrm{GeV}$ assigned to the $\mathbf{V}$ frame, $N_{\text{jet}}^{\mathbf{V}}$, after application of the $p_{\mathbf{ISR},T}^{\mathbf{CM}}$ and $R_{\mathrm{ISR}}$ selections described in Table 1. Gluino signals tends to have a larger $N_{\text{jet}}^{\mathbf{V}}$ compared to SM backgrounds.

Figure 6

Distribution of number of jets with momentum greater than 20 GeV, assigned to the Visible system “$\mathbf{V}$” as a function of $R_{\mathrm{ISR}}$ for the $\mathrm{boson}+\text{jets}$ (upper left), di-boson (upper right), $\mathrm{top}+\mathrm{X}$ (lower left) and gluino signal (lower right) samples.

Figure 7

Distribution of transverse mass of the “sparticle” system “$\mathbf{S}$” ($M_T^{\mathbf{S}}$) as a function of $R_{\mathrm{ISR}}$ for the $\mathrm{boson}+\text{jets}$ (upper left), di-boson (upper right), $\mathrm{top}+\mathrm{X}$ (lower left) and gluino signal (lower right) samples.

Figure 8

Distribution of $\mathrm{\Delta }{\varphi }_{\mathbf{ISR},\mathbf{I}}$ for SM backgrounds compared with the squark pair signal samples. All curves are shown after applying the relevant criteria from Table 1. Although both signal and background distributions tend towards $\pi$ the signal has a much stronger tendency to do so.

Figure 9

The distribution of $R_{\mathrm{ISR}}$ for example gluino signals and backgrounds after the application of requirements on $p_{\mathbf{ISR},T}^{\mathbf{CM}}$ ($>1000\mathrm{GeV}$), $M_T^{\mathbf{S}}$ ($>100\mathrm{GeV}$), $N_{\text{jet}}^{\mathbf{V}}\ge 3$, and $\mathrm{\Delta }{\varphi }_{\mathbf{ISR},\mathbf{I}}(>3.0)$. We observe that selecting events with large values of $R_{\mathrm{ISR}}$ provides excellent discrimination between signals and backgrounds.

Figure 10

Projected exclusion and discovery reach for gluino pair production (upper plot) squark pair production (lower plot) in the compressed regions with $25\mathrm{GeV}\le \mathrm{\Delta }m\le 200\mathrm{GeV}$.