Constraining the Evolution of the HI Spin Temperature with Fast Radio Bursts

Fast radio bursts (FRBs) emit broad band radio wave radiation that may, in rare cases, encode atomic hydrogen (HI) absorption signals produced as they traverse the interstellar medium of their host galaxies. Combining such signals with high resolution HI emission maps offers a unique opportunity to probe the dynamics of neutral gas at cosmological distances through constraints of the HI excitation temperature $T_{spin}$, which characterises the balance of neutral gas phases and the underlying thermal processes within these galactic environments. While no absorption signal has been recorded in an FRB to date, we demonstrate a proof of concept with the bright (F = 35 Jy ms) and narrow (0.2 ms) FRB 20211127I detected by ASKAP. We find a 3$σ$ upper limit on the integrated optical depth in the pulse-averaged spectrum of 33 km s$^{-1}$, and, based on the HI emission observed in a 3 hr MeerKAT L-band observation, subsequently find a lower limit on $T_{spin}$ of 26 K. While this test case provides little constraining power, we find that narrow, non-repeating FRBs with fluences greater than 20/70/150 Jy ms observed with all dishes with the current MeerKAT/ASKAP/DSA telescopes can probe integrated optical depths below 5 km s$^{-1}$. Furthermore, we highlight that utilising FAST’s incredible sensitivity to stack thousands of bursts from hyperactive repeaters also provides a plausible avenue through which HI absorption, and hence $T_{spin}$, can be measured. Finally, we discuss how HI absorption can address several modern challenges in FRB science, providing a physical anchor for locating bursts within their host galaxies and helping to disentangle the host contribution to dispersion and scattering.

💡 Research Summary

The paper proposes a novel method to constrain the spin temperature ( (T_{\text{spin}}) ) of neutral hydrogen (HI) in external galaxies by exploiting the 21 cm absorption that may be imprinted on fast radio burst (FRB) signals as they pass through their host interstellar medium. The spin temperature, which governs the relative population of the two hyperfine levels of HI, is a direct probe of the balance between the cold neutral medium (CNM) and warm neutral medium (WNM) and therefore of the thermal state of the ISM, feedback processes, and galaxy evolution. Traditional measurements of (T_{\text{spin}}) require a background quasar with a known HI column density, often obtained from Ly‑α absorption or direct HI emission, both of which are rare for extragalactic sightlines. FRBs, being bright, millisecond‑duration point sources that occur at cosmological distances, provide a new class of background illuminators that can be paired with high‑resolution HI emission maps of their hosts.

The authors first lay out the theoretical framework: the Boltzmann relation between the hyperfine level populations (Eq. 1), the definition of optical depth from the observed flux decrement (Eq. 2), and the conversion from integrated optical depth to HI column density (Eq. 4). They then derive the sensitivity of an FRB absorption measurement in terms of the spectral signal‑to‑noise ratio (SNR), the telescope system equivalent flux density (SEFD), the FRB fluence, line width, and redshift (Eqs. 5‑10). A key result is Eq. 10, which expresses a 3σ upper limit on the integrated optical depth as a function of observable quantities, showing that higher fluence, lower SEFD, and narrower lines dramatically improve detectability.

To demonstrate feasibility, the paper presents a proof‑of‑concept analysis of FRB 20211127I, a bright (35 Jy ms) and narrow (0.2 ms) burst detected by ASKAP. The host galaxy was observed with MeerKAT for three hours, providing a spatially resolved HI emission spectrum at the exact FRB location. After careful baseline subtraction and accounting for strong interstellar scintillation, the authors find no statistically significant absorption feature. They place a 3σ upper limit on the integrated optical depth of 33 km s⁻¹. Combining this with the HI column density derived from the MeerKAT emission map yields a lower limit on the spin temperature of (T_{\text{spin}} > 26) K. Although this constraint is weak, it validates the methodology.

The paper then explores future prospects. Using the derived sensitivity formula, the authors show that single, non‑repeating FRBs with fluences exceeding ~20 Jy ms (MeerKAT), ~70 Jy ms (ASKAP), or ~150 Jy ms (DSA) could probe integrated optical depths below 5 km s⁻¹, providing much tighter (T_{\text{spin}}) limits. More promising is the use of the Five‑hundred‑meter Aperture Spherical Telescope (FAST) to stack thousands of bursts from hyper‑active repeaters; the resulting increase in SNR could enable detection of very shallow absorption lines even for modest‑fluence bursts. The authors discuss how such measurements would anchor FRB positions within their hosts, help disentangle host contributions to dispersion measure (DM) and scattering, and ultimately allow a redshift‑dependent census of HI spin temperature, shedding light on the evolution of the CNM/WNM balance over cosmic time.

In summary, the study establishes that FRB‑based HI absorption, when combined with high‑resolution emission data, is a viable tool for measuring the spin temperature of neutral hydrogen in distant galaxies. While the current single‑burst demonstration yields only a modest lower limit, the analysis outlines clear pathways—targeting brighter bursts, employing more sensitive facilities, and stacking repeat bursts—to achieve meaningful constraints. Such measurements will provide a new window on the thermal state of the interstellar medium, the impact of stellar and AGN feedback, and the broader context of galaxy evolution across the observable universe.