2009MNRAS.392..181K

As the number of known exoplanets continues to grow, the question as to whether such bodies harbour satellite systems has become one of increasing interest. In this paper, we explore the transit timing effects that should be detectable due to an exomoon and predict a new observable. We first consider transit time variation (TTV), where we update the model to include the effects of orbital eccentricity. We draw two key conclusions.

(i) In order to maintain Hill stability, the orbital frequency of the exomoon will always be higher than the sampling frequency. Therefore, the period of the exomoon cannot be reliably determined from TTV, only a set of harmonic frequencies.

(ii) The TTV amplitude is ~= M_Sa_S where M_S is the exomoon mass and a_Sis the semimajor axis of the moon's orbit. Therefore, M_S and a_Scannot be separately determined.

We go on to predict a new observable due to exomoons - transit duration variation (TDV). We derive the TDV amplitude and conclude that its amplitude is not only detectable, but the TDV signal will also provide two robust advantages.

(i) The TDV amplitude is ~=M_Sa^–1/2_S and therefore the ratio of TDV to TTV allows for M_Sand a_S to be separately determined.

(ii) TDV has a π/2 phase difference to the TTV signal, making it an excellent complementary technique.