13:45 〜 15:15
[PEM11-P16] Performance of correlated-k distribution method in radiative cooling of atmospheric escape simulation
キーワード：系外惑星、大気散逸
Until today, thousands close-in super-/sub-Earths are discovered. Due to their small orbits and thus their strong stellar irradiation enviroments, the hydrodynamic escape of planetary atmospheres caused by stellar UV is expected to highly affect their evolution and surface environments. One of the key processes in the escaping atmospheres is radiative cooling, since the process removes thermal energy from atmospheres heated up by the high energy photons from its host stars and then significantly affects the thermal structure and reduce the escape rate. While molecules can be dominant coolants in the escaping atmospheres, especially in heavy-element-enriched ones, there is a numerical difficulty in considering all radiative line of molecules. This is because the number of radiative lines of molecules is generally too large (e.g., the number of H2O lines is about half a million) to be handled in hydrodynamic simulations. Due to this reason, hydrodynamic simulations have taken a limited number of molecules’ lines into account with an emission line/band selection (e.g., Johnstone 2020, Yoshida and Kuramoto 2020).
In an attempt to a numerical way to consider all molecule lines in hydrodynamic simulations, in this study, we investigate how efficiently correlated-k method (Liou1980) works with different wavelength resolutions (R). This method is a grouping method of radiative lines and derives the absorption or emission probability density in wavelength bands which successfully reproduces the result of more sophisticated/accurate line-by-line calculations despite the small computational costs. In our assessment, we consider Mars-size planets with H2-rich transonic atmospheres containing H2O or CO as radiative species and having different thermal structures and different fractions of the radiative species. Then, we found that the error of the radiative cooling rate calculation with correlated-k method comparing to line-by-line method is within the order of ten percent for R>3000. Also, we found the difference of calculated escape rates between correlated-k method and a line selection scheme is within the order of ten percent for R>100 and the difference of calculated temperature profiles between them is within the order of ten percent for R=100. As conservative one, we found correlated-k method with R=1000 provides the values of escape rate and temperature profiles which differ from correlated-k method with R=3000 by just only the 10 percent and 1 percent, respectively. In addition, we discuss in what condition of planetary atmosphere correlated-k method efficiently works in hydrodynamic simulations.
In an attempt to a numerical way to consider all molecule lines in hydrodynamic simulations, in this study, we investigate how efficiently correlated-k method (Liou1980) works with different wavelength resolutions (R). This method is a grouping method of radiative lines and derives the absorption or emission probability density in wavelength bands which successfully reproduces the result of more sophisticated/accurate line-by-line calculations despite the small computational costs. In our assessment, we consider Mars-size planets with H2-rich transonic atmospheres containing H2O or CO as radiative species and having different thermal structures and different fractions of the radiative species. Then, we found that the error of the radiative cooling rate calculation with correlated-k method comparing to line-by-line method is within the order of ten percent for R>3000. Also, we found the difference of calculated escape rates between correlated-k method and a line selection scheme is within the order of ten percent for R>100 and the difference of calculated temperature profiles between them is within the order of ten percent for R=100. As conservative one, we found correlated-k method with R=1000 provides the values of escape rate and temperature profiles which differ from correlated-k method with R=3000 by just only the 10 percent and 1 percent, respectively. In addition, we discuss in what condition of planetary atmosphere correlated-k method efficiently works in hydrodynamic simulations.