The Event Horizon Telescope recently observed the first shadow of a black hole. Images like this can potentially be used to test or constrain theories of gravity and deepen the understanding in plasma physics at event horizon scales, which requires accurate parameter estimations. In this work, we present Deep Horizon, two convolutional deep neural networks that recover the physical parameters from images of black hole shadows. We investigate the effects of a limited telescope resolution and observations at higher frequencies. We trained two convolutional deep neural networks on a large image library of simulated mock data. The first network is a Bayesian deep neural regression network and is used to recover the viewing angle $i$, and position angle, mass accretion rate $\dot{M}$, electron heating prescription $R_{\rm high}$ and the black hole mass $M_{\rm BH}$. The second network is a classification network that recovers the black hole spin $a$. We find that with the current resolution of the Event Horizon Telescope, it is only possible to accurately recover a limited number of parameters of a static image, namely the mass and mass accretion rate. Since potential future space-based observing missions will operate at frequencies above 230 GHz, we also investigated the applicability of our network at a frequency of 690 GHz. The expected resolution of space-based missions is higher than the current resolution of the Event Horizon Telescope, and we show that Deep Horizon can accurately recover the parameters of simulated observations with a comparable resolution to such missions.

The astrophysical agenda of the 21st century requires a very sharp view of celestial objects. High angular resolution studies are essential for fundamental studies of a broad variety of astrophysical phenomena ranging from relativistic physics of black holes, their gravitational and electromagnetic imprints, violent transient processes, including those producing detectable gravitational waves, birth and evolution of planetary systems. Over the past decades, radio astronomy made huge leap in achieving ground-breaking angular resolution measured in tens of microarcseconds (one tenth of nanoradian and better). Recently a global Event Horizon Telescope (EHT) collaboration obtained first direct images of the shadow of a super-massive black hole in the nucleus of the active galaxy M87. These observations were conducted at 230 GHz. The two first generation Space Very Long Baseline Interferometry (VLBI) missions, VSOP/HALCA led by the Japan Aerospace Exploration Agency (JAXA) and RadioAstron led by the Russia Roscosmos State Corporation and Russia Academy of Sciences, achieved the highest angular resolution at frequencies from 0.3 to 22 GHz in observations conducted in the period 1997 - 2019. The next step in advancing high angular resolution radio astronomy is in combining high frequency (millimeter and sub-millimeter wavelengths) and interferometric baselines exceeding the Earth diameter. The present THEZA White Paper describes a combination which would unify technology developments in giga-/tera-hertz instrumentation and space-borne radio astronomy. The current preprint version of the THEZA White Paper is slightly re-formatted and edited comparing to the official submitted version.