21CMMC with a 3D light-cone: The impact of the co-evolution approximation on the astrophysics of reionization and cosmic dawn

We extend 21CMMC, a Monte Carlo Markov Chain sampler of 3D reionization simulations, to perform parameter estimation directly on 3D light-cones of the cosmic 21 cm signal. This brings theoretical analysis closer to the tomographic 21 cm observations achievable with next generation interferometers like the Hydrogen Epoch of Reionization Array and the Square Kilometre Array. Parameter recovery can therefore account for modes that evolve with redshift/ frequency. Additionally, simulated data can be more easily corrupted to resemble real data. Using the light-cone version of 21CMMC, we quantify the biases in the recovered astrophysical parameters if we use the 21 cm power spectrum from the co-evolution approximation to fit a 3D light-cone mock observation. While ignoring the light-cone effect under most assumptions will not significantly bias the recovered astrophysical parameters, it can lead to an underestimation of the associated uncertainty. However, significant biases (~few - 10\u3c3) can occur if the 21 cm signal evolves rapidly (i.e. the epochs of reionization and heating overlap significantly), and (i) foreground removal is very efficient, allowing large physical scales (k 72 0.1Mpc-1) to be used in the analysis or (ii) theoretical modelling is accurate to within ~10 per cent in the power spectrum amplitude