What of using a selective wavelength radiating material - ie emits 'heat' at a wavelength that is clear of atmospheric absorption wavelengths, so the material is cooler than all other local materials/surfaces ( if such a phenoma is possible ) - then water vapour at a relative humidity , might be cooled to, or below its dewpoint on the surface of the radiator material ?