Direct detection of terrestrial exoplanets: Comparing the potential for space and ground telescopes

Telescopes of various different designs are potentially capable of detecting extrasolar terrestrial planets. We analyze here in a consistent way the limiting sensitivities set by photon noise from the background underlying the planet signal, which may be of thermal, zodiacal or stellar origin. The strength of the unsuppressed stellar halo is itself set by photon noise in wavefront measurement. While optical telescopes have potentially higher limiting sensitivity, thermal detection is more secure, At 11 μm wavelength, the planet/star contrast is 1000 times more favorable than in the optical. Together with the longer wavelength, this leads to a 500 times more relaxed tolerance for star suppression, one that can be met by a fast servo based on the bright star flux sensed at shorter wavelengths. Either Darwin or a 100 m ground telescope should be capable of thermal detection of the earth in a solar system twin at 10 pc at 5 to 10σ in 24 hr. At optical wavelengths, the limiting sensitivity for space telescopes is set at the 10-30σ level by photon noise in the zodiacal background. Reaching this limit, as do the deep fields of the Hubble Space Telescope, will require extreme coronagraphic suppression of the bright star at 0.1 arcsec separation. The ∼1 m-scale Fourier components of the wavefront would need to have stable amplitude ≤2 picometers, a severe challenge. On the ground, fast atmospheric correction at the photon noise limit will leave residual Fourier amplitudes of 20-60 pm, for a halo background 100-1000 times zodiacal. But given larger apertures and stronger fluxes, optical sensitivity can still be high, provided the photon noise limit of short halo exposures can be maintained in a long-term average. If this challenge can be met, detection in 24 hr would be at the 5σ level for a 20 m Antarctic telescope, ∼50σ for the 100 m OWL. If a terrestrial planet were detected at 10 pc, a spectrum that could reveal water and oxygen would be of great interest. Thermal features can be accessed only from space, where broad spectral cover is possible. Optical spectroscopy could be undertaken with a 100 m telescope on the ground. An Antarctic location would give high sensitivity to water, and the added benefit of thermal imaging with high sensitivity and resolution.