Einstein Probe: X-ray Nondetection Limits

• Einstein Probe X-ray nondetection refers to cases where FXT follow-up observations yield upper limits on transient fluxes based on sensitivity thresholds derived from detailed background simulations.
• The methodology employs Monte Carlo simulations to separate FOV and instrumental background contributions, with open-filter configurations achieving 5σ limits around 5.9×10⁻¹⁴ erg cm⁻² s⁻¹ in the 0.5–2 keV band.
• These nondetection constraints guide the estimation of transient source luminosities and help refine models for time-domain astrophysical phenomena.

The Einstein Probe (EP) is a space X-ray imaging mission dedicated to time-domain astrophysics, equipped with two distinct scientific payloads: the wide-field X-ray telescope (WXT) and the Follow-up X-ray Telescope (FXT). The FXT units utilize Wolter-I type mirrors and pn-CCD detectors to enable deep pointed observations of transient X-ray sources discovered by the WXT in the 0.3–10 keV band. Accurate estimation of in-orbit background and corresponding point-source sensitivity is essential for interpreting nondetections and setting upper limits on variable or transient X-ray fluxes. Background levels and associated flux thresholds have been derived through detailed Monte Carlo simulations, providing a quantitative basis for evaluating Einstein Probe X-ray nondetection phenomena (Zhang et al., 2021).

1. Instrumentation and Observational Context

Einstein Probe's FXTs are positioned in a low-Earth orbit with altitude 600 km and inclination 29°, designed to follow up WXT triggers with exposures typically of 1.5 ks (25 minutes). Each FXT module covers the 0.3–10 keV energy range using Wolter-I optics with pn-CCDs as focal-plane detectors (FPDs). The FPD imaging area is a 384×384 pixel square (28.8 mm × 28.8 mm), and a circular focal-spot region of $θ = 30''$ is defined to encompass approximately 90% of the FXT point-spread function (PSF), serving as the standard extraction aperture for point-source analysis. Both the absolute background and sensitivity estimates incorporate the operational configuration, including the use of open, thin, medium, or thick filter-wheel positions.

2. In-Orbit Background Components

The FXT's total background is classified into two main sources: field-of-view (FOV) background and instrumental background.

• FOV Background: Funnelled events that pass through the Wolter-I optics, predominantly comprising cosmic photon background (the combination of the cosmic X-ray background, CXB, and Galactic soft X-ray background) reflected by the mirror, as well as low-energy protons near the geomagnetic equator. The open-filter FOV background is approximately $R_{FOV} ≈ 1.61$ counts s$^{-1}$ integrated across 0.5–10 keV. Below $\sim2$ keV, this component overwhelms the instrumental background.
• Instrumental Background: Originates from interactions of high-energy particles or photons with the telescope structure and shielding, leading to secondaries that hit the detector outside the optical path. It includes primary cosmic-ray protons/electrons/positrons, albedo (secondary) protons and electrons/positrons from Earth’s atmosphere, and albedo gamma rays. The uniform instrumental background rate is $R_{instr} ≈ 3.1 × 10^{-2}$ counts s$^{-1}$ keV$^{-1}$, corresponding to $3.7×10^{-3}$ counts s$^{-1}$ keV$^{-1}$ cm$R_{FOV} ≈ 1.61$0 on the normalized detector area.

A summary of focal-spot background rates for a single FXT module (open-filter configuration) is presented below:

Energy Band	Instrumental (cts s$R_{FOV} ≈ 1.61$1)	FOV Cosmic photons (cts s$R_{FOV} ≈ 1.61$2)
0.5–2 keV	$R_{FOV} ≈ 1.61$3	$R_{FOV} ≈ 1.61$4
2–10 keV	$R_{FOV} ≈ 1.61$5	$R_{FOV} ≈ 1.61$6
0.5–10 keV	$R_{FOV} ≈ 1.61$7	$R_{FOV} ≈ 1.61$8

Low-energy protons in the FOV near the geomagnetic equator ($R_{FOV} ≈ 1.61$9 counts s$^{-1}$0 keV$^{-1}$1) can be excluded from analysis using quality screening and filter positions, and are thus omitted from sensitivity estimation.

3. Sensitivity Derivation and Detection Thresholds

Point-source sensitivity calculations depend fundamentally on the exposure time ($^{-1}$2), on-axis effective area ($^{-1}$3), and the total background rate ($^{-1}$4) in the extraction region. For a significance threshold $^{-1}$5 (commonly $^{-1}$6 for 5σ detection), the sensitivity $^{-1}$7 for a given energy $^{-1}$8 and background systematic uncertainty $^{-1}$9 is governed by:

$\sim2$0

In the ideal case with negligible systematic uncertainty ($\sim2$1), this reduces to:

$\sim2$2

These expressions yield the minimum detectable flux in counts cm$\sim2$3 s$\sim2$4 keV$\sim2$5 for specified exposure and background conditions. The background normalization applies standard "grade" screening (discarding multi-pixel split events, typically those spreading over ≥4 pixels).

4. Quantitative Sensitivity and Nondetection Limits

For typical FXT follow-up exposures of 25 minutes and a Crab-like spectrum ($\sim2$6, $\sim2$7 cm$\sim2$8), the 5σ flux limits have been computed for both open and thick filter states, and for both statistical and systematic uncertainty-dominated cases. Numerical thresholds are as follows:

Energy Band	Filter	5σ Limit (σ_sys=0) [erg cm$\sim2$9 s$R_{instr} ≈ 3.1 × 10^{-2}$0]	5σ Limit (σ_sys=10%) [erg cm$R_{instr} ≈ 3.1 × 10^{-2}$1 s$R_{instr} ≈ 3.1 × 10^{-2}$2]	μCrab Equivalent (σ_sys=0)	μCrab Equivalent (σ_sys=10%)
0.5–2 keV	Open	$R_{instr} ≈ 3.1 × 10^{-2}$3	$R_{instr} ≈ 3.1 × 10^{-2}$4	5.1–9.7	13.6–19.9
0.5–2 keV	Thick	$R_{instr} ≈ 3.1 × 10^{-2}$5	$R_{instr} ≈ 3.1 × 10^{-2}$6
2–10 keV	Open	$R_{instr} ≈ 3.1 × 10^{-2}$7	$R_{instr} ≈ 3.1 × 10^{-2}$8	28.0–37.1	52–69
2–10 keV	Thick	$R_{instr} ≈ 3.1 × 10^{-2}$9	$^{-1}$0

At exposure times $^{-1}$1 background-limited regime, $^{-1}$2, denoting that sensitivity improves with the square root of integration time barring systematic limitations. For 3σ nondetections, the corresponding open-filter 0.5–2 keV limit in 25 min is $^{-1}$3 erg cm$^{-1}$4 s$^{-1}$5 ($^{-1}$6 μCrab).

Nondetection at the prescribed sensitivity threshold translates into an upper limit on the source flux and, given an assumed distance, the corresponding X-ray luminosity. For example, a $^{-1}$7 erg cm$^{-1}$8 s$^{-1}$9 upper limit in the 0.5–2 keV band leads to $^{-1}$0 erg s$^{-1}$1 at 10 kpc.

5. Methodological Dependencies and Caveats

Several factors critically impact the sensitivity assessment and the interpretation of nondetections:

• Extraction Region Size: The standard focal-spot aperture is set to $^{-1}$2, corresponding to $^{-1}$3 encircled energy. Varying the aperture modifies the background in direct proportion to the area and, thus, $^{-1}$4.
• Sky Direction and Background Variance: The soft Galactic X-ray background (0.5–2 keV) exhibits sky-dependent variability up to a factor of $^{-1}$5. Sensitivity at high Galactic latitudes improves by $^{-1}$6–50%.
• Orbital Background Modulation: Modulations driven by geomagnetic latitude, solar cycle, and South Atlantic Anomaly (SAA) crossings can alter cosmic-ray and albedo fluxes by factors of 2–3, affecting the net background and hence the achievable sensitivity.
• Systematics: Systematic uncertainties in background subtraction, effective area calibration, and PSF/vignetting modeling contribute to sensitivity degradation. Explicitly, a 10% systematic uncertainty on the background increases the flux limits by approximately a factor of 2.
• Source Spectrum Assumptions: Deviations from the assumed Crab-like power-law influence the effective area folding and optimal energy band, shifting the numerical $^{-1}$7 by $^{-1}$8–50%.

A plausible implication is that actual nondetection limits must account for both statistical noise and systematics derived from real-time observational context, necessitating conservative interpretation in reporting upper limits.

6. Interpretation and Impact of FXT Nondetections

For exposures that yield no statistically significant source at the pre-specified $^{-1}$9 threshold, the nondetection upper limits derived from the formalism above define strict flux (and, with distance, luminosity) constraints for both steady and transient X-ray phenomena. For short-duration flares with $3.7×10^{-3}$0, $3.7×10^{-3}$1 is replaced by $3.7×10^{-3}$2 in the calculations. These limits, particularly in the 0.5–2 keV and 2–10 keV bands, are critical for ruling out emission models or constraining source energetics in time-domain astrophysical studies.

The background structure of FXT, with FOV background dominating at $3.7×10^{-3}$3 keV and instrumental background dominating above, defines the sensitivity floor for detection and thus the quantitative basis for upper limits on X-ray nondetections relevant to both Galactic and extragalactic transients (Zhang et al., 2021).