Ok, check my math. basics: solar constant: 1.36 kW / m^2 earth-to-moon range: 400 km parameters from the article: frequency: 5 GHz -> wavelength: 6cm

Ok, check my math. basics: solar constant: 1.36 kW / m^2 earth-to-moon range: 400 km parameters from the article: frequency: 5 GHz -> wavelength: 6cm earth antenna array linear dimension: 200 km transmit power density: 100 W/m^2 Let's assume that the earth antenna array elements are 50m wide, and spaced out such that they cover 1% of the total 200km * 50m area, for a total antenna aperture of 1e5 square meters (10% of the SKA). Combining the stated transmit power density of 100 W/m^2 with the antenna area, we get a total transmit power of 10 MW. Throw it at Friis: power density at moon = transmit power * earth antenna area / (range^2 * wavelength^2) = 10e6 watts * 1e5 m^2 / ( 400e3^2 m^2 * 0.06^2 m^2) = 1.73 kW / m^2 Atmospheric attenuation at 5 GHz is pretty minimal. If we conservatively assume 20% loss, I think we still end up with a higher power density at a single frequency than from the sun across the entire spectrum. I don't think anything like this will ever be built, but I don't see why it is impossible. Where's the mistake?