The effect of stellar limb darkening values on the accuracy of the planet radii derived from photometric transit observations

Context. The radius of an exoplanet is one of its most important parameters. Studies of planetary interiors and their evolution require 1% precision in the radius determination. Transiting exoplanets offer a unique oppurtunity to measure the radius of exoplanets in stellar units. These radius measurements and their precision are strongly affected by our knowledge of limb darkening. Aims. We study how the precision of the exoplanet radius determination is affected by our present knowledge of limb darkening in two cases: when we fix the limb darkening coefficients and when we adjust them. We also investigate the effects of spots in one-colour photometry. Methods. We study the effect of limb darkening on the planetary radius determination both via analytical expressions and by numerical experiments. We also compare some of the existing limb darkening tables. When stellar spots affect the fit, we replace the limb darkening coefficients, calculated for the unspotted cases, with effective limb darkening coefficients to describe the effect of the spots. Results. There are two important cases. (1) When one fixes the limb darkening values according to some theoretical predictions, the inconsistencies of the tables do not allow us to reach accuracy in the planetary radius of better than 1-10% (depending on the impact parameter) if the host star's surface effective temperature is higher than 5000 K. Below 5000 K the radius ratio determination may contain even 20% error. (2) When one allows adjustment of the limb darkening coefficients, the a/R-s ratio, the planet-to-stellar radius ratio, and the impact parameter can be determined with sufficient accuracy (<1%), if the signal-to-noise ratio is high enough. However, the presence of stellar spots and faculae can destroy the agreement between the limb darkening tables and the fitted limb darkening coefficients, but this does not affect the precision of the planet radius determination. We also find that it is necessary to fit the contamination factor, too. Conclusions. We conclude that the present inconsistencies of theoretical stellar limb darkening tables suggests one should not fix the limb darkening coefficients. When one allows them to be adjusted, then the planet radius, impact parameter, and the a/Rs can be obtained with the required precision.