Wave-roughened ocean model pushes exoplanet glint detection down to 120 degrees

Phase angle and angular separation pull against each other. For a circular orbit seen edge on, a planet's projected distance from its star goes as the sine of the phase angle, and sin 120 degrees is 0.87 against 0.77 for sin 130 degrees, so working at 120 degrees leaves the planet about 13 percent further out from the glare an instrument has to suppress ^[20]. Blocking the parent star's light is one of the hardest parts of the Habitable Worlds Observatory design, and the ten degrees is described in the phys.org account as breathing room for its designers ^[15].

The lit fraction moves the same direction. Illumination of a disk at phase angle alpha is (1 + cos alpha)/2, which is a quarter at 120 degrees and about 18 percent at 130 ^[21].

What sets the boundary at 120 degrees is colour. Starlight that skims an ocean at these angles takes a long slant path out through the atmosphere, and Rayleigh scattering strips most of the blue before it leaves, so the flash reaches the telescope reddened ^[12]. Sand, rock and soil are Lambertian: they bounce light equally in all directions and produce no such feature ^[6].

This is a simulation study, not a detection. Cornish and Robinson, both at the University of Arizona, modified rfast and added the Cox-Munk ocean model so that wind speed and wave ripples set how the surface reflects ^[10]. The account of their work describes the noise thresholds only as "relatively high" and does not quantify them ^[22]. No liquid ocean has been confirmed on another planet, though dozens of the more than 5,500 known exoplanets sit in their stars' habitable zones ^[3][2][1].

The underlying physics has been measured twice, both times inside the solar system. Cassini caught a specular glint off Titan's hydrocarbon lakes in 2009 ^[8], and decades earlier a team including Carl Sagan used a Galileo flyby to spot glint off Earth's own oceans ^[9].

The cloud assumption is where I would push hardest. The simulations used 50 percent cover for an Earth-like world, and there are far cloudier planets even nearby ^[16]. Thick cloud can blot the ocean out entirely; high-altitude cirrus can scatter starlight toward the telescope and imitate a glint flash where no ocean exists ^[17]. The paper points to one way out, distinguishing gas from liquid absorption bands, which puts a spectral test behind any brightness rise ^[18].

So the useful place for this result is a requirements document. It says which phase angles a coronagraph and an observing schedule have to reach, and it says the thing to look for is a reddened crescent ^[11][13]. The authors also note how little data exists on Earth in its own crescent phase, and call collecting some as training data crucial if the tool is to help HWO scientists interpret what they see ^[19].