Thermal Phase Curves in Hot Gas Giant Exoplanets Exhibit a Complex Dependence on Planetary Properties

We present a catalog of uniformly processed 3.6-$\mu$m and 4.5-$\mu$m band exoplanet thermal phase curves based on Infrared Array Camera observations obtained from the Spitzer Heritage Archive. The catalog includes phase curve measurements for 34 planets, 16 of which contain full orbit coverage and have detectable secondary eclipses in both channels. The data are processed in the EXCALIBUR pipeline using a uniform analysis consisting of aperture photometry and modeling of instrument effects along with the exoplanet signal. Nearest-neighbors regression with a Gaussian kernel is used to correct for instrumental systematics correlated to the star's centroid position and shape in conjunction with a novel test to avoid overfitting. These methods may have utility in addressing sub-pixel gain variations present in modern infrared detectors. We analyze the 3.6-$\mu$m and 4.5-$\mu$m phase curve properties and find a strong wavelength-dependent difference in how the properties correlate with physical parameters as well as evidence that the phase curve properties are determined by multiple physical parameters. We suggest that differences between the 3.6-$\mu$m and 4.5-$\mu$m phase curve properties are due to 3.6~$\mu$m observations probing regions of the atmosphere which could include a cloud layer. Taken together, the observed phase curve behavior suggests that different physical processes are responsible for establishing the thermal phase curve at different pressures, which are probed by different wavelengths, and that further 3D GCM modeling is required to investigate the reason for this complex dependence on planetary properties.