Properties of an Earth-like planet orbiting a Sun-like star: Earth observed by the EPOXI mission.

NASA's EPOXI mission observed the disc-integrated Earth and Moon to test techniques for reconnoitering extrasolar terrestrial planets, using the Deep Impact flyby spacecraft to observe Earth at the beginning and end of Northern Hemisphere spring, 2008, from a range of ∼1/6 to 1/3 AU. These observations furnish high-precision and high-cadence empirical photometry and spectroscopy of Earth, suitable as "ground truth" for numerically simulating realistic observational scenarios for an Earth-like exoplanet with finite signal-to-noise ratio. Earth was observed at near-equatorial sub-spacecraft latitude on 18-19 March, 28-29 May, and 4-5 June (UT), in the range of 372-4540 nm wavelength with low visible resolving power (λ/Δλ=5-13) and moderate IR resolving power (λ/Δλ=215-730). Spectrophotometry in seven filters yields light curves at ∼372-948 nm filter-averaged wavelength, modulated by Earth's rotation with peak-to-peak amplitude of ≤20%. The spatially resolved Sun glint is a minor contributor to disc-integrated reflectance. Spectroscopy at 1100-4540 nm reveals gaseous water and carbon dioxide, with minor features of molecular oxygen, methane, and nitrous oxide. One-day changes in global cloud cover resulted in differences between the light curve beginning and end of ≤5%. The light curve of a lunar transit of Earth on 29 May is color-dependent due to the Moon's red spectrum partially occulting Earth's relatively blue spectrum. The "vegetation red edge" spectral contrast observed between two long-wavelength visible/near-IR bands is ambiguous, not clearly distinguishing between the verdant Earth diluted by cloud cover versus the desolate mineral regolith of the Moon. Spectrophotometry in at least one other comparison band at short wavelength is required to distinguish between Earth-like and Moon-like surfaces in reconnaissance observations. However, measurements at 850 nm alone, the high-reflectance side of the red edge, could be sufficient to establish periodicity in the light curve and deduce Earth's diurnal period and the existence of fixed surface units.