An outline of the paper: Photometric variability of the disk integrated infrared emission of the Earth. by Gómez-Leal, Pallé and Selsis
What could we infer from a high-accuracy light curve of an Earth-like planet in the thermal infrared? Gómez-Leal et al. use satellite measurements to reconstruct the global infrared light-curve of the Earth, and then studying it as an exoplanet.
Earlier studies have considered the same issue (*), but in the case of measuring reflected sunlight in the visible. They find that with good time series with two colours, and provided the rotation period of the planet is known independently, one can reconstruct a good first-order map of the position of landmasses and oceans on an Earth-like planet.
Now, Gómez-Leal et al. consider the potential of a stupendously accurate time series of the global infrared flux of Earth over a whole year. What they find is that the diurnal and seasonal cycles can be recovered very well. The amplitudes of these cycles are very dependent on the angle at which the planet is seen: when seen pole-on, the day-night cycle is muted, the seasonal cycle is stronger in the Northern hemisphere (because of the scarcity of continents in the South).
Beyond this, the variability gives an indication of the inhomogeneity of the planet, both in terms of surface features and cloud coverage.
The main features in the daily thermal light curve can be associated to known regions. For instance the Sahara and other tropical deserts become very hot in the summer. No surprise there. Less intuitive is the fact that the coldest spots are the polar regions and … Indonesia! This is because, from space in the infrared, we see the opaque tops of large tropical storms. These reach the highest part of the troposphere where the temperature drops to minus 50 Celsius or so.
Whether we could invert this information for another planet to recover the landmass, ocean and cloud information is another matter. It seems challenging, when you think that what makes Indonesia special is the “fractal” arrangement of islands and sea that combines the evaporation from an equatorial ocean with the heat from land masses.
Incidentally, the authors conclude by pointing out that, in the case of the Earth, the phases of the Moon dominate the infrared lightcurve for a distant observer. Alien planetologists unaware of the presence of our heavy satellite would thus probably infer a 28-day rotation period for our planet, with a huge continent on one side and a splendid equatorial archipelago on the other. If their data is good enough, they may detect a slight 24-hour variation, that they may attribute to the phases of a small close-in moon.(*) Cowan et al. [2009, ApJ 700, 915] and Fujii & Kawahara [2012, arXiv1204.3504]
[Feature Image: Thermal image of the Earth on July 1st, 2001, from Fig. 1 of the paper.]
Alien Maps of an Ocean-bearing world, Cowan et al. 2009 (ApJ 700, 915)
Mapping Earth-analogs from Photometric Variability: Spin-Orbit Tomography for Planets in Inclined Orbits, Fujii & Kawahara 2012 (sumitted to ApJ)