First generation space-based optical coronagraphic telescopes will obtain images of cool gas and ice giant exoplanets around nearby stars .
The albedo spectra of exoplanets lying at planet-star separations larger than about 1 AU –where an exoplanet can be resolved from its parent star–are dominated by reflected light to beyond 1 \mu m and are punctuated by molecular absorption features .
Here we consider how exoplanet albedo spectra and colors vary as a function of planet-star separation , metallicity , mass , and observed phase for Jupiter and Neptune analogs from 0.35 to 1 \mu m. We model Jupiter analogs with 1 \times and 3 \times the solar abundance of heavy elements , and Neptune analogs with 10 \times and 30 \times solar abundance of heavy elements .
Our model planets orbit a solar analog parent star at separations of 0.8 AU , 2 AU , 5 AU , and 10 AU .
We use a radiative-convective model to compute temperature-pressure profiles .
The giant exoplanets are found to be cloud-free at 0.8 AU , possess H _ { 2 } O clouds at 2 AU , and have both NH _ { 3 } and H _ { 2 } O clouds at 5 AU and 10 AU .
For each model planet we compute moderate resolution ( R = \lambda / \Delta \lambda \sim 800 ) albedo spectra as a function of phase .
We also consider low-resolution spectra and colors that are more consistent with early direct imaging capabilities .
As expected , the presence and vertical structure of clouds strongly influence the albedo spectra since cloud particles not only affect optical depth but also have highly directional scattering properties .
Observations at different phases also probe different volumes of atmosphere as the source-observer geometry changes .
Because the images of the planets themselves will be unresolved , their phase will not necessarily be immediately obvious , and multiple observations will be needed to discriminate between the effects of planet-star separation , metallicity , and phase on the observed albedo spectra .
We consider the range of these combined effects on spectra and colors .
For example , we find that the spectral influence of clouds depends more on planet-star separation and hence atmospheric temperature than metallicity , and it is easier to discriminate between cloudy 1 \times and 3 \times Jupiters than between 10 \times and 30 \times Neptunes .
In addition to alkalis and methane , our Jupiter models show H _ { 2 } O absorption features near 0.94 \mu m. While solar system giant planets are well separated by their broad-band colors , we find that arbitrary giant exoplanets can have a large range of possible colors and that color alone can not be relied upon to characterize planet types .
We also predict that giant exoplanets receiving greater insolation than Jupiter will exhibit higher equator to pole temperature gradients than are found on Jupiter and thus may exhibit differing atmospheric dynamics .
These results are useful for future interpretation of direct imaging exoplanet observations as well as for deriving requirements and designing filters for optical direct imaging instrumentation .