PI: Shikhar Mittal
The period when the first stars began to form, known as cosmic dawn, remains one of the least explored epochs in the history of the Universe. A powerful probe of this era is the 21-cm signal, a faint radio emission from neutral hydrogen that provides a unique window into both cosmic dawn and the even earlier dark ages. Ongoing ground-based experiments such as REACH (PI: E. de Lera Acedo, Cambridge) are actively searching for this signal, while proposed space-based missions such as CosmoCube (PI: E. de Lera Acedo, Cambridge) aim to extend these efforts to earlier times. Fully exploiting the scientific potential of these experiments requires accurate theoretical modelling of the 21-cm signal across a wide range of astrophysical scenarios.
The 21-cm signal encodes the complex interplay between the first stars and their electromagnetic radiation, particularly ultraviolet (UV) and X-ray emission, which shape both the amplitude and spectral structure of the signal. Interpreting forthcoming data therefore demands fast, flexible, and computationally efficient tools capable of exploring large, high-dimensional parameter spaces.
To meet this need, I have developed ECHO21 , a user-friendly, open-source Python package designed to model the 21-cm signal from epochs before the formation of the first stars through to the present day. ECHO21 simultaneously solves for the thermal evolution of the intergalactic medium and the ionization history of the Universe, enabling rapid computation of the 21-cm signal. Using DiRAC CPU resources, we generated a large suite of 15,625 distinct 21-cm signals, spanning a broad range of astrophysical parameters, in approximately 10 minutes using 50 CPUs (Figure 1). This demonstrates that ECHO21 is well suited for navigating a large-scale parameter and hence for statistical inference required by upcoming observations.
While ECHO21 already provides a fast and flexible modelling framework, the code is under active development. In particular, the current version does not include full radiative-transfer effects of UV photons. Under the present DiRAC CPU allocation, we are performing high-resolution radiative-transfer calculations to obtain resolution-converged summaries of UV coupling (Rani, Mittal, et al, in prep.). These results will be incorporated into future versions of ECHO21 , significantly improving the physical fidelity of the model while retaining its computational efficiency.
Figure 1: 21-cm signal ( 𝑇 21 ) from dark ages to the end of reionization generated using ECHO21 .
This article is partly based on the results published as S. Mittal et al. RASTI, 5, rzag001 (2026) . The code is available at https://github.com/shikharmittal04/echo21 and the documentation at https://echo21.readthedocs.io/en/latest/index.html