Forward-Models of Cosmic Dawn: connecting 21cm simulations to the real world

FORWARD will enable a new generation of transformational probes of Cosmic Dawn (CD) to provide meaningful inference on key questions of galaxy formation. The CD contains some of the most sought-after insights into galaxy formation. Though these secrets remain locked away by a host of formidable real-world challenges, upcoming radio telescopes are optimistic. Interferometers such as HERA and future SKA, along with global experiments such as EDGES-3 are honed to observe the faint 21cm emission from neutral Hydrogen, redshifted to 50-150 MHz, in complementary ways. The dominant limitation of current experiments are spectral systematics that couple with overwhelmingly bright foregrounds. As we enter the regime in which the most extreme physical models begin to be ruled out by observations, meaningful inference of key parameters is most susceptible to inaccuracies in accounting for these systematics. Prior efforts to provide parameter constraints from CD have focused on forecasts, and treated systematics crudely and optimistically. FORWARD will synthesise the popular semi-numerical simulation code 21cmFAST and a radically overhauled parameter-inference engine, 21cmMC-2, with a series of new pluggable modules designed to efficiently forward-model the response of realistic interferometers, enabling the most direct comparison of theory to data. By applying this novel machinery jointly to observations from HERA and EDGES-3, we will (i) provide meaningful constraints on the parameters of galaxy formation during CD, (ii) build confidence in the reported detections, and (iii) enable future joint constraints from the SKA & EDGES-3. The framework developed will follow a legacy of code excellence, providing a bedrock for state-of-the-art developments over the coming years.