Performing retrieval on a mock data set, where the “ground truth” is known (see the text for input parameters). We focus on performing retrieval with the UB model (see Table 1). The columns represent the different wavelength coverages: 0.7–5 μm with 0.01 μm resolution (first column), HR 8799b-like data (second column) and HR 8799e-like data (third column). The rows represent different assumptions for R and g: fixed R and g (first row), fixed R but g is a fitting parameter (second row), fixed g but R is a fitting parameter (third row), both R and g are fitting parameters (fourth row). The labels “CO,” “CO2,” “CH4,” and “H2O” refer to the mixing ratios of carbon monoxide, carbon dioxide, methane, and water, respectively. ﹩{T}_{\mathrm{int}}﹩ is in units of K. The labels “﹩{\kappa }_{0}﹩” and “gravity” are for ﹩\mathrm{log}{\kappa }_{0}﹩ and ﹩\mathrm{log}g﹩ in mks and cgs units, respectively.