P. Figueira, M. Oshagh, V. Zh. Adibekyan, N. C. Santos

Abstract
Aims. We re-evaluate the correlation between planetary surface gravity and stellar host activity as measured by the index log (R′_HK). This correlation, previously identified by Hartman (2010, ApJ, 717, L138), is now analyzed in light of an extended measurement dataset, roughly three times larger than the original one.
Methods. We calculated the Spearman rank correlation coefficient between the two quantities and its associated p-value. The correlation coefficient was calculated for both the full dataset and the star-planet pairs that follow the conditions proposed by Hartman (2010). To do so, we considered effective temperatures both as collected from the literature and from the SWEET-Cat catalog, which provides a more homogeneous and accurate effective temperature determination.
Results. The analysis delivers significant correlation coefficients, but with a lower value than those obtained by Hartman (2010). The two datasets are compatible, and we show that a correlation coefficient as high as previously published can arise naturally from a small-number statistics analysis of the current dataset. The correlation is recovered for star-planet pairs selected using the different conditions proposed by Hartman (2010). Remarkably, the usage of SWEET-Cat temperatures led to higher correlation coefficient values. We highlight and discuss the role of the correlation betwen different parameters such as effective temperature and activity index. Several additional effects on top of those discussed previously were considered, but none fully explains the detected correlation. In light of the complex issue discussed here, we encourage the different follow-up teams to publish their activity index values in the form of a log (R′_HK) index so that a comparison across stars and instruments can be pursued.

Keywords
planetary systems - methods: data analysis - methods: statistical

Astronomy & Astrophysics
Volume 572, Number of pages A51_1
2014 November

>> DOI