(abridged) The Super-Earth candidate GL 581 d is the first potentially habitable extrasolar planet. Therefore, GL 581 d is used to illustrate a hypothetical detailed spectroscopic characterization of such planets. Atmospheric profiles from 1D radiative-convective model scenarios of GL 581 d were used to calculate high-resolution synthetic spectra. From the spectra, signal-to-noise ratios were calculated for a telescope such as the planned James Webb Space Telescope. The presence of the model atmospheres could be clearly inferred from the calculated synthetic spectra due to strong water and carbon dioxide absorption bands. Surface temperatures could be inferred for model scenarios with optically thin spectral windows. Dense, CO2-rich scenarios did not allow for the characterization of surface temperatures and to assess habitability. Degeneracies between CO2 concentration and surface pressure further complicated the interpretation of the calculated spectra, hence the determination of atmospheric conditions. Still, inferring approximative CO2 concentrations and surface pressures would be possible. In practice, detecting atmospheric signals is challenging. The SNR for a single transit was only larger than unity in some near-IR bands for transmission spectroscopy. Most interestingly, the false-positive detection of biomarker candidates such as methane and ozone could be possible in low resolution spectra due to the presence of CO2 absorption bands which overlap with biomarker spectral bands. This can be avoided however by observing all main CO2 IR bands instead of concentrating on, e.g., the 4.3 or 15 micron bands only. Furthermore, a masking of ozone signatures by CO2 absorption bands is shown to be possible. Simulations imply that such a false-negative detection of ozone would be possible even for rather large ozone concentrations of up to 1E-5.Read less <