# How is an exoplanet characterised as “Earth-Like”?

Which features or characteristics are taken into account to characterize an exoplanet as "Earth-Like"?

There is no unique definition of Earth-like and it depends on what characteristics are important in the context of discussion.

The most simple one is to just compare radii and masses. That is observationally sensible as radius is directly deducable from transits (assuming stellar size known) and mass is often an accessible properties for transiting planets when you have access to complementary radial-velocity data. It's also a good starting point if you want to add further restrictions. The habitable planet catalogue uses this definition (considering only planets in the habitable zones around their stars, using a simple equilibrium temperature definition)

If your quest is actually for a 2nd Earth not only in surface gravity but other characteristics, one approach is outlined by the Earth similarity index (ESI). While it is not directly a measure for habitability either, it is a good proxy by comparing more essential parameters with Earth like mass, gravity, surface temperature. For a selection of parameters you define this ESI as

$$\mathrm{ESI} = \prod_{i=1}^n \left(1 - \frac{x_i - x_{i,o}}{x_i + x_{i,o} }\right)^{w_i/n}$$

where $$x_{i,o}$$ is the Earth's value of a property and $$n$$ the total number of properties. This is basically a weighted geometric mean of the property differences. It is a good idea to consider 'internal values' and 'external' ones. The Planetary Habitability Lab uses surface temperature and escape velocity as external and radius and density as internal and use this to create a nice 2D scatter plot for known planets.

• Yes and no. Transiting alone is not sufficient, you need RV measurements, too, to be able to test for its mass. Often RV is possible and available for transiting planets. I amend my answer to reflect that – planetmaker Mar 12 at 8:27
• Finding the mass of an Earth-sized planet is VERY difficult. It is actually impossible at present if it is an Earth-mass planet in an Earth-like orbit around a Sun-like star, which a lot of people would argue is exactly what you need for a planet to be considered "Earth-like". Transits give you the radius of a planet. That is the most easy thing to measure. – Rob Jeffries Mar 12 at 9:40
• You need the latest technology in spectroscopy to resolve the RV excitations of the host star for Earth-like planets. But it is meanwhile a proven technique and readily applied to all planets with transits where possible. An Earth in 1 AU distance around a sun might still be a challange but becoming accessible for RV; but an Earth around a K star in its habitable zone is closer-in and has a stronger signal. See articles.adsabs.harvard.edu/cgi-bin/… – planetmaker Mar 12 at 11:24
• See web.archive.org/web/20101020155144/http://espresso.astro.up.pt/… and eso.org/sci/facilities/lasilla/instruments/harps/science/papers/… - both are not exactly new publications (10 years) and they already set the detection limit in the range of Earth-like RV variations. – planetmaker Mar 12 at 11:37
• Please point precisely to where an Earth-like planet mass has been deduced for a star in a 300+ day orbit around a solar-type star. The most simple thing is to compare radii. Why not alter your answer to say that whilst masses can be found for some Earth-mass planets, these are found in short-period orbits. These can still be habitable, if they are close orbits around a cooler star. Not arguing with any of that. But some would argue that these are not Earh-like. – Rob Jeffries Mar 12 at 11:51